Educational Prompt Engineering Self-Efficacy Scale (Ed-PESS): Scale Development and Psychometric Validation

Educational Prompt Engineering Self-Efficacy Scale (Ed-PESS): Scale Development and Psychometric Validation | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Educational Prompt Engineering Self-Efficacy Scale (Ed-PESS): Scale Development and Psychometric Validation Fatih KARATAS, Recep GÜR, Barış ERİÇOK, Fatma Başarır, Lokman TANRIKULU This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9003817/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Pre-service teachers' confidence in designing pedagogically purposeful prompts for generative AI tools remains an underexamined construct in teacher education. This study developed and validated the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS) to address the absence of a psychometrically sound instrument measuring this domain. Scale development followed Boateng et al.'s three-phase framework, grounded in a systematic literature review and expert review procedures. Two independent samples of pre-service teachers were recruited: an exploratory factor analysis (EFA) sample (n = 300) and a confirmatory factor analysis (CFA) sample (n = 305). EFA using principal axis factoring with Promax rotation produced a four-factor structure explaining 64.60% of total variance. CFA with Yuan-Bentler robust correction confirmed acceptable model fit. The final 26-item scale comprises four dimensions: Basic Prompt Engineering, Pedagogical Prompt Engineering, Ethical Prompting Practice, and Continuous Professional Development. Internal consistency was high across all subscales. The Ed-PESS provides a valid, reliable instrument for assessing prompt engineering self-efficacy in pre-service teacher education. It supports formative curriculum evaluation and targeted intervention design. prompt engineering educational prompt engineering prompt engineering self-efficacy AI scale development Figures Figure 1 Figure 2 Figure 3 1.Introduction The integration of artificial intelligence (AI) into educational settings is driving a profound transformation across diverse areas, from curriculum design to assessment practices, and from evolving teacher roles to issues of accessibility and inclusivity. With the widespread adoption of AI applications, fundamental aspects of teaching including lesson planning, content development, personalized feedback, and interactive learning experiences are being reshaped. This transformation extends beyond mere technological innovation. It represents a paradigm shift in how teaching and learning are conceptualized and enacted (Serra & Oliveira, 2025 ). Within this context, AI integration is increasingly recognized as a strategic priority for educational institutions (Pratschke, 2024 ), gaining growing attention at the institutional level in higher education (Lee & Palmer, 2025 ). Despite this growing integration, AI's pedagogical value depends not on the technology itself but on the quality of user AI interaction. Arocha ( 2025 ) theoretically demonstrates that the educational outputs of AI systems are sensitive to the structure of user inputs, while Selwyn et al. ( 2025 ) emphasize that these tools require deliberate guidance from users to generate meaningful outputs. Empirical findings also support this view. Users without systematic training tend to use AI as a search engine and accept outputs without critical evaluation (Sigot & Tassoti, 2025 ). In this context, the creation of appropriate prompts plays a decisive role in obtaining targeted educational outputs from AI tools (ElSayary et al., 2025 ; Park & Choo, 2024 ). In its most basic definition, a prompt is a text-based instruction or input provided by the user to an AI system to generate a response or complete a specific task (Arocha, 2025 ; Correia et al., 2025 ). Crosthwaite et al. ( 2025 ) indicates that effective prompts are aligned with learning objectives, contextually rich, and well structured, whereas ineffective prompts consist of vague or overly simplistic commands. This suggests that prompt quality directly determines pedagogical outcomes (Davidson & Karell, 2025 ). Consequently, the ability to craft effective prompts has emerged as a critical competency for AI use in education, and the conceptual framework for purposefully structuring this interaction is referred to as educational prompt engineering (Agirdag, 2025 ). In the literature, prompt engineering is defined within a rigorous computational and semantic framework as the systematic structuring and refinement of user inputs to obtain desired outputs from AI systems (Lee & Palmer, 2025 ; Pratschke, 2024 ). Although prompt engineering is closely related to AI literacy, which encompasses general competencies such as understanding and evaluating AI technologies (Long & Magerko, 2020 ; Ng et al., 2021 ), it represents a distinct, practice-oriented application of this broader competence, enabling teachers to take concrete actions toward their pedagogical goals and obtain purpose-aligned outputs from AI tools (Celik et al., 2026 ; Kang et al., 2025 ). While the concept was initially developed in technical contexts, the need for domain-specific strategies has led to the emergence of a new competency area termed educational prompt engineering, which intersects with pedagogical expertise in education. This emergence establishes the basis for investigating prompt engineering as a systematically teachable competency. Prompt engineering is a teachable and developable skill; therefore, its pedagogy can be empirically investigated (Lee & Palmer, 2025 ). Recent empirical research demonstrates that structured prompt engineering training significantly influences learners' prompting strategies and learning outcomes. Woo et al. ( 2025 a) found that students used prompt engineering not merely as a technical operation but as a tool supporting cognitive processes such as overcoming writer's block and developing story plots. In the context of teacher education, Sigot and Tassoti ( 2025 ) observed that through a structured prompting framework, students transitioned from simple search-engine-like usage to strategic and iterative prompting practices. Similarly, Hwang et al. ( 2025 ) reported that students without guidance focused on superficial linguistic features and struggled to construct goal-aligned prompts. Mutanga et al. ( 2025 ) also reported that most students exhibited usage patterns relying on superficial commands. Moreover, AI-supported instructional practices and prompt engineering interventions have been shown to strengthen task-specific self-efficacy beliefs (Andewi et al., 2025 ; Huang et al., 2024 ; Woo et al., 2025 ), support self-regulated learning (Mzwri & Turcsányi-Szabó, 2025 ), foster creativity and critical thinking (Kabeer et al., 2025 ), and reduce technology anxiety (Davila-Moran et al., 2025 ). Taken together, these findings suggest that prompt engineering skills do not develop spontaneously but require intentional, pedagogically grounded instruction. This highlights the importance of systematically integrating such training into teacher education curricula. While systematically integrating prompt engineering training into teacher education curricula is essential, the effectiveness of such training also depends on how confident teachers feel in applying these skills in instructional contexts. This points to the need for a construct that captures teachers' confidence in designing and refining pedagogically grounded prompts. Given that these competencies and beliefs are predominantly formed during pre-service education, teacher education curricula should move beyond technical usage to also address the development of educational prompt engineering self-efficacy (Moorhouse et al., 2025 ; Serra & Oliveira, 2025 ). In a systematic review, Qian ( 2025 ) demonstrates that prompt engineering in education is positioned as both a technical skill and a pedagogical strategy. Accordingly, scholars argue that pedagogically productive interaction with AI requires the holistic integration of pedagogical content knowledge, AI knowledge, and prompting skills (Correia et al., 2025 ; Moorhouse et al., 2025 ). Critical scholarship also indicates that educational prompt engineering should be understood not merely as a technical skill but within a broader framework encompassing pedagogical purposes and contextual sensitivity (Agirdag, 2025 ; Arocha, 2025 ; Selwyn et al., 2025 ). However, competency alone does not guarantee effective practice; teachers must also believe in their ability to apply these skills in pedagogical contexts. Identifying and fostering this self-efficacy is therefore essential for curriculum design, as it determines whether acquired competencies translate into meaningful classroom practice. Accordingly, pre-service teachers' educational prompt engineering self-efficacy, that is, their perceived capability to apply prompt design skills in instructional tasks such as lesson planning, material development, and assessment, emerges as a key construct warranting empirical attention (Crosthwaite et al., 2025 ). Furthermore, drawing on the theoretical framework that conceptualizes self-efficacy as task- and context-specific (Bandura, 2006 ), prompt engineering self-efficacy can be defined as an individual's subjective belief in their ability to create pedagogically meaningful prompts when interacting with AI systems, to integrate these prompts into teaching and learning processes, and to sustain this interaction within an evaluation–refinement cycle. Importantly, this belief extends beyond technical competence and is closely associated with how teachers position themselves in relation to AI. This imperative, however, first requires a clear conceptualization of prompt engineering self-efficacy as a distinct construct within teacher education. Van Zoost ( 2025 ) suggests that prompt engineering serves as a technology of the self, moving teachers beyond mere users to a space where they negotiate their pedagogical decisions and professional identities. Empirical evidence further supports this perspective. Biberman-Shalev ( 2025 ) found that pre-service teachers, through the process of prompt refinement, recognized the personalization limitations of AI and rediscovered the value of their own pedagogical roles. Beyond this identity dimension, these perspectives also imply that teachers with stronger self-efficacy beliefs are more likely to engage in purposeful and pedagogically aligned interactions with AI tools. Taken together, these findings suggest that educational prompt engineering self-efficacy reflects not merely a technical skill but a professional capacity that mediates between knowledge and practice, shaping how teachers position themselves as pedagogically informed agents in AI-integrated environments. These conceptual and empirical considerations carry significant implications for teacher education. For future teachers, prompt engineering self-efficacy is becoming a foundational area of professional belief, as fundamental as confidence in preparing lesson plans or developing assessment instruments. Within the professional roles that pre-service teachers will assume, traditional teaching competencies such as lesson planning, assessment, and instructional material development are expected to be carried out in integration with AI tools (ElSayary et al., 2025 ; Park & Choo, 2024 ). This integration, however, is not straightforward. As Pratschke ( 2024 ) noted, experiences from Khan Academy have shown that while having AI generate a lesson plan is easy, having it generate a "good" lesson plan is a demanding process requiring mastery of pedagogical principles. This suggests that pre-service teachers need to develop not only technical familiarity with AI tools but also confidence in directing them toward pedagogical purposes. Regarding how such confidence can be cultivated, Hsu ( 2025 ) proposes that within a constructivist prompting framework, it can be fostered through principles such as meaningful prompting and metacognitive analysis. From a broader institutional perspective, Correia et al. ( 2025 ) directly define teachers' evolving roles as "prompt engineers" and relate this identity to UNESCO's AI competency framework. According to the authors, effective prompt design enables teachers to maintain their professional autonomy as ethical decision makers and pedagogical process designers, underscoring the importance of pre-service teachers acquiring these competencies during their initial training. In the specific context of teacher education, other studies examined the use of prompt engineering strategies within the Intelligent TPACK framework (Celik et al., 2026 ), the mediating role of disciplinary literacy (Kang et al., 2025 ), and the relationship with pedagogical creative thinking (Tümen Akyıldız, 2026 ). While these studies reveal the relationship between prompt engineering and pedagogical practices, they have predominantly been framed around general AI self-efficacy or perceptions toward technology, rather than directly focusing on a distinct, task- and context-specific self-efficacy construct unique to prompt engineering in pedagogical settings. As a learnable and developable capacity, this construct is also directly measurable, which in turn necessitates valid and reliable instruments for its systematic assessment. 1.1.Need for Educational Prompt Engineering Self-Efficacy Scale (Ed-PESS) The growth of generative AI in educational contexts has led to the development of various measurement instruments targeting related constructs. The Artificial Intelligence Literacy Scale (AILS), comprising 12 items across four dimensions (awareness, usage, evaluation, and ethics) provides a validated framework for measuring general AI literacy (Wang et al., 2023 ). Similarly, the AI Self-Efficacy Scale developed by Wang and Chuang ( 2024 ) assesses individuals' perceived confidence in using generative AI through a 22-item instrument with four dimensions: anthropomorphism, assistance, technological skill, and comfort with AI. In teacher education contexts specifically, Chiu et al. ( 2025 ) introduced the Teacher Artificial Intelligence Competence Self-Efficacy Scale (TAICS), a 24-item instrument measuring six dimensions including AI assessments, AI knowledge, AI ethics, AI pedagogy, professional engagement, and human-centered education. The Intelligent-TPACK framework (Celik, 2023 ) extends the traditional TPACK model to address AI-specific pedagogical knowledge, encompassing Intelligent-TK, Intelligent-TCK, Intelligent-TPK, and ethical dimensions. Additionally, digital competence frameworks such as the Students' Digital Competence Scale (SDiCoS; Tzafilkou et al., 2022 ) and the Digital Competency Scale for Teachers (Gümüş & Kukul, 2023 ) have emerged to assess technology integration capabilities. Despite these contributions, significant limitations constrain their applicability to measuring prompt engineering or prompt engineering self-efficacy among pre-service teachers. First, most instruments were developed before the emergence of generative AI and prompt engineering as distinct constructs, focusing instead on general technology acceptance and use (Laupichler et al., 2023 ). Second, existing measures emphasize general AI interaction capabilities rather than the specific belief structures underlying confidence in prompt design for pedagogical purposes. Third, none of these instruments capture the iterative, dialogic nature of prompt engineering confidence, which requires educators to engage in cycles of prompt formulation, output evaluation, and refinement to achieve instructionally appropriate responses (Celik et al., 2026 ; Tümen Akyıldız, 2026 ). Besides these studies, most recently and most closely related to the present study, Gibreel and Arpaci ( 2025 ) developed the Prompt Engineering Competence Scale (PECS), a unidimensional nine-item instrument measuring users' general proficiency in prompt engineering. The PECS represents a significant contribution by establishing prompt engineering as a measurable construct. However, critical distinctions limit its applicability to education. First, the PECS measures competence rather than self-efficacy, yet self-efficacy beliefs are more predictive of behavioral engagement than objective skill levels (Bandura, 2006 ). Second, the instrument was developed with general AI users, lacking pedagogical specificity essential for understanding how teachers design prompts to achieve instructional objectives. Consequently, a teacher or teacher candidate may demonstrate high PECS scores while lacking confidence in applying prompt engineering to educational purposes. Additionally, the PECS, while addressing prompt engineering competence, measures actual performance rather than self-efficacy beliefs and lacks pedagogical specificity relevant to teacher preparation contexts. These limitations reveal a clear methodological gap, specifically the absence of a psychometrically validated instrument designed to measure educational prompt engineering self-efficacy for pre-service teachers. The present study addresses this gap through the development and validation of the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS). While educational policy increasingly requires AI integration in teacher preparation (Kang et al., 2025 ; Tümen Akyıldız, 2026 ), no valid instrument exists to determine whether teacher candidates believe they can effectively engineer prompts to bring out pedagogically appropriate AI responses. The consequences of this problem are far-reaching, as teacher training curricula cannot systematically identify students with low prompt engineering self-efficacy, evaluate the impact of instructional interventions, or detect prompt-related anxiety that may slow down effective AI utilization. Self-efficacy theory (Bandura, 2006 ) provides the theoretical foundation for addressing this gap. Self-efficacy, defined by Bandura as belief in one's ability to achieve desired goals, encompasses competence and confidence in successfully performing tasks. Building on the research gap highlighted in the literature, the purpose of this study is to develop the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS) and to provide evidence of its validity and reliability. To ensure theoretical validity, the instrument’s development was grounded in a systematic literature review and thematic analysis, with every item derived from and supported by direct evidence from the literature. Accordingly, the study aimed to obtain evidence regarding the construct validity and reliability of the scale within the framework of current psychometric standards. To this end, the following research questions were addressed: What is the factor structure of the Ed-PESS based on exploratory factor analysis? To what extent does confirmatory factor analysis support the factor structure of the Ed-PESS? What is the evidence for the reliability and convergent validity of the Ed-PESS and its subscales? 2. Methods 2.1 Research Design This study employed an instrument development and validation design to develop the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS). The scale development process was guided by Boateng et al.’s ( 2018 ) three-phase, nine-step framework: item development (domain identification and item generation; content validity), scale development (question pre-testing, survey administration and sampling, item reduction, and latent factor extraction), and scale evaluation (tests of dimensionality, reliability, and validity). In Phase 1, the construct domain was defined and an initial item pool was created based on a systematic review of the prompt engineering and AI in teacher education literature. Expert review and pilot testing with the target group were then used to improve item relevance and clarity. In Phase 2, the draft scale was administered to Sample 1 of pre-service teachers, where item reduction procedures and exploratory factor analysis were used to identify the latent structure. In Phase 3, the retained items were tested in an independent Sample 2 (n = 305) using confirmatory factor analysis. Because the data were not normally distributed, maximum likelihood estimation with a Yuan and Bentler robust correction was used. Reliability was examined using Cronbach’s alpha and McDonald’s omega, alongside internal structure validity indices (e.g., factor loadings and AVE). 2.2. Participants Given the inherent requirements of the scale development process, this study was conducted with two independent participant groups. In the first phase, data were collected from 327 pre-service teachers to examine the construct validity of the 32-item draft scale through EFA. Following univariate (z-score, ± 3) and multivariate (Mahalanobis distance) outlier screening, the final EFA sample comprised n₁ = 300 participants. For the CFA phase, a separate sample of 341 pre-service teachers was recruited independently; after removing five univariate and 31 multivariate outliers, the final CFA sample comprised n₂ = 305 participants. A participant-to-item ratio exceeding 10:1 in both groups provides evidence for sampling adequacy. Detailed demographic characteristics of both groups are presented in Table 1 . Table 1 Demographic Characteristics of Study Group 1 (EFA) and Study Group 2 (CFA) Variable Category Study Group 1 (EFA, n₁ = 300) Study Group 2 (CFA, n₂ = 305) Gender Female 219 (73.0%) 224 (73.4%) Male 81 (27.0%) 81 (26.6%) Age 17–18 — 19 66 (21.6%) 20 56 (18.7%) 74 (24.3%) 21 79 (26.3%) 22 and above Range 18–40 17–38 Year of Study 1st Year 101 (33.1%) 2nd Year 83 (27.7%) 127 (41.6%) 3rd Year 113 (37.7%) 4th Year Academic Program (Top Programs) English Language Teaching 24.0% 9.2% German Language Teaching 16.0% — Social Studies Teaching — 9.5% Primary AI Usage Purpose Homework & Presentation Prep 69 (23.0%) 31.1% AI Usage Duration 1–2 years 36.4% 30.5% AI Training Status Received prior AI training 37 (12.3%) 20 (6.6%) No training; willing to receive 181 (60.3%) 140 (45.9%) No training; not interested 79 (26.3%) 141 (46.2%) Note. EFA = Exploratory Factor Analysis; CFA = Confirmatory Factor Analysis; HBV = Hacı Bektaş Veli. Percentages may not sum to 100 due to rounding. Dashes (—) indicate data not separately reported for that group. As seen in Table 1 , several patterns in the demographic profiles of both groups warrant methodological consideration. The pronounced gender imbalance, with female participants comprising approximately 73% of both samples, reflects the broader feminization of teacher education in Turkey rather than a sampling artifact. Both groups were predominantly composed of first- and second-year students with limited AI training experience; fewer than 13% in Group 1 and fewer than 7% in Group 2 had received any formal AI instruction. This near-absence of structured AI training, combined with the fact that nearly half of Group 2 reported no training and no intention to pursue it, suggests that the scale was validated under ecologically representative conditions for the current state of AI integration in Turkish pre-service teacher education. At the same time, it raises the question of whether self-efficacy scores may be systematically attenuated by limited prior exposure, a potential confound that longitudinal or intervention-based designs should address. 2.3 Instrument Development and Procedure Data were collected via an online survey administered to pre-service teachers between May and November 2025; participation was voluntary, anonymous, and preceded by informed consent, with no incentives offered. Following a systematic literature review and researcher discussions, an initial pool of 38 items was drafted in accordance with 5 Likert-type scale development conventions, with ambiguous expressions deliberately avoided. Items were reviewed by three Measurement and Evaluation specialists, two Computer Education and Instructional Technology specialists, and two Turkish Language specialists for language, content, and psychometric quality; applying Lawshe's (1975) CVR criterion of .80 at p < .05, six items falling below this threshold were removed, and the remaining items were revised to yield a 32-item draft instrument comprising four subscales: Basic Prompt Engineering (9 items), Pedagogical Prompt Engineering (7 items), Ethical Prompting Practice (9 items), and Continuous Professional Development (7 items). The study was conducted with two independent samples: responses collected from 327 pre-service teachers were screened for univariate and multivariate outliers, yielding a final EFA sample of n₁ = 300 (retention rate ≈ 91.7%); a separate sample of 341 participants was subsequently recruited for CFA, with analogous screening procedures producing a final sample of n₂ = 305 (retention rate ≈ 89.4%). 2.4. Data Analysis Data analysis was carried out in three stages: data screening, construct validity analyses, and reliability analyses. Prior to conducting the analyses, the normality assumption of the dataset was evaluated using Mardia's coefficients. Additionally, z-scores (± 3 threshold) were employed for univariate outlier detection, and Mahalanobis distance values were used for multivariate outlier identification. To provide evidence of construct validity and to determine the factor structure of the scale, EFA was performed as the initial analytical step. The suitability of the dataset for factorization was assessed using the Kaiser-Meyer-Olkin (KMO) coefficient and Bartlett's Test of Sphericity. The KMO value was found to be .942, indicating excellent sampling adequacy, and Bartlett's test results were statistically significant, χ²(325) = 6173, p < .001. Principal Axis Factoring was selected as the extraction method, and given the presence of inter-factor correlations (r = .507 to .645), Promax rotation, an oblique rotation method, was employed. The number of factors was determined by examining eigenvalues greater than 1 and inspecting the scree plot. CFA was subsequently applied to confirm the structure identified through EFA. Because the data did not conform to a multivariate normal distribution, the Maximum Likelihood (ML) estimation method was used in conjunction with Yuan-Bentler (robust) correction. Model fit was evaluated using the following indices: Robust χ²/df, RMSEA (Root Mean Square Error of Approximation), SRMR (Standardized Root Mean Square Residual), CFI (Comparative Fit Index), and TLI (Tucker-Lewis Index). To provide evidence of reliability, Cronbach's alpha (α), McDonald's omega (ω), and the composite reliability (CR) coefficient were computed to assess the internal consistency of both the total scale and its subscales. Average Variance Extracted (AVE) values were examined as evidence of convergent validity. All analyses were performed using the JAMOVI statistical software. 2.5. Measures: Item Generation and Development The item development process followed a theoretically rigorous approach supported by a systematic literature review to identify the construct's dimensional structure. Subsequently, thematic analysis of core papers established the framework for item generation, ensuring that each item was based on specific empirical or theoretical evidence. This literature-grounded methodology ensures initial content validity. Following Boateng et al.'s ( 2018 ) scale development framework, the systematic search was conducted in “Web of Science” and “Scopus” databases to establish the theoretical foundation for item development. The review identified core papers addressing prompt engineering in educational contexts, from which principal themes were extracted through thematic analysis. Based on this synthesis, educational prompt engineering self-efficacy was operationally defined as a multidimensional construct comprising pre-service teachers' confidence in designing, implementing, and evaluating prompts for generative AI tools in pedagogically appropriate ways. This conceptualization integrates Bandura's (2006) self-efficacy theory with domain-specific competencies identified in the literature. The systematic literature review revealed four recurring themes that formed the theoretical architecture for item development: (1) technical prompt construction skills, (2) pedagogical integration of AI outputs, (3) ethical considerations in educational AI use, and (4) ongoing professional learning for AI integration. These themes were marked as candidate dimensions, with each theme's basic elements serving as the framework for item generation. Items were written as first-person self-efficacy statements (e.g., "I can...") using clear language appropriate for pre-service teachers, following established guidelines (Boateng et al., 2018 ). Each item was supported by direct citations from the literature to ensure theoretical validity (see Appendix 1 for all items and literature evidence). An initial pool of 32 items was generated, distributed across the four thematically-derived dimensions: Basic Prompt Engineering (9 items), Pedagogical Prompt Engineering (7 items), Ethical Prompting Practice (9 items), and Continuous Professional Development (7 items). In addition to the Appendix, Table 2 provides the first two sample items and literature support for each dimension. Table 2 Sample Items and Literature Support for Each Dimension Dimension Scale Items Direct Quote (Evidence) Dimension 1: Basic Prompt Engineering (9 items), 1. I can explain how AI tools generate content. 1. "AI systems generate text (outputs) by making statistically informed predictions based on the patterns they have learned and responding to prompts (inputs) entered by users (Park & Choo, 2024 )." 2. "The session included demonstrations of interactions with ChatGPT-4, showcasing its text generation and text-to-image functionalities. Particular emphasis was placed on the importance of crafting effective prompts, as well as demonstrating how refining prompts could elicit more accurate and contextually appropriate responses from the AI (Biberman-Shalev, 2025 )." 3. "ChatGPT works by learning from a large amount of text data, which helps it understand grammar, vocabulary, and the meaning of words in different contexts. This learning process of ChatGPT is analogous to training ChatGPT's brain to understand language (Lee et al., 2024 )." 2. I can select the AI tool (e.g., ChatGPT, Claude, Gemini) that best fits my needs. 1. "When prompt engineering, you will start by choosing a model. Prompts might need to be optimized for your specific model, regardless of whether you use Gemini language models in Vertex AI, GPT, Claude, or an open source model like Gemma or LLaMA (Boonstra, 2025 )." 2. "The most widely used large language models (LLMs) today include ChatGPT, Gemini, Claude, Mistral, and Llama, with ChatGPT remaining the most dominant overall (Mutanga et al., 2025 )." 3. "I preferred ChaGPT4.0. It is expensive, but it gives you value for money. I used my bursary and subscribed to this tool (Van Wyk, 2025 )." Dimension 2: Pedagogical Prompt Engineering (7 items), 1. I can design prompts that generate content aligned with my learning objectives. 1. "Goal-oriented and task-specific frameworks focus on designing prompts to fulfill clearly defined instructional or functional objectives. These frameworks align prompt structures with desired outcomes, such as eliciting student reasoning or performing a diagnostic task (Qian, 2025 )." 2. "For teachers to use GenAI tools effectively, they must craft prompts that encourage the AI to engage substantively with instructional goals and deliver responses aligned with intended learning outcomes (Celik et al., 2026 )." 3. "Continuous refinement through iterative prompt writing ensures alignment with learning objectives and facilitates real-time lesson adaptation (Carl et al., 2024 )." 4. "Each instruction must be 'initiated with a clear and precise directive' and maintain alignment with the learning objectives and educational goals (Lee & Palmer, 2025 )." 2. I can develop prompts that generate differentiated content suitable for students' needs. 1. "Adaptive learning is an instructional approach in which learning pathways are adjusted to align with learners' needs, prior knowledge, and interests... instruction is tailored to students' progress and characteristics (Celik et al., 2026 )." 2. "Crucially, the task emphasized designing adaptive or personalized instructional strategies by encouraging pre-service teachers to request support from the AI that considered pupil diversity, engagement, and differentiated learning needs (Celik et al., 2026 )." 3. "Real-time AI assistance in classrooms enhances lesson delivery and enables personalized learning for diverse student needs (Carl et al., 2024 )." 4. "Prompts emphasizing contextual clarity aid differentiation by adjusting outputs to learner needs (Qian, 2025 )." Dimension 3: Ethical Prompting Practice (9 items), 1. I can develop classroom rules for prompt creation (e.g., 'do not use your name') to ensure privacy and data security. 1. "Educate students about data privacy, including how their data is used by AI systems and ways to protect their digital footprint (Walter, 2024 )." 2. "One practical approach is to develop a personal checklist for appropriate and ethical uses of AI (Park & Choo, 2024 )." 3. "It should be clear how the expectations of the school look like so that students know exactly what they are allowed and what they are not allowed to do (Walter, 2024 )." 2. I can protect the confidentiality of student data while creating prompts. 1. "Providers of LLM-based GenAI tools, such as OpenAI, use prompting data to train their AI models, which may raise concerns regarding data privacy and protection... To mitigate this, learners should avoid providing personal or identifiable information when interacting with LLM-based GenAI (Hsu, 2025 )." 2. "AI systems frequently handle private user data without following open procedures for its storage or use... many AI applications gather personal data without express consent (Tümen Akyıldız, 2026 )." 3. "Ethical AI development is essential, focusing on transparency, unbiased content, and privacy-respecting practices (Walter, 2024 )." Dimension 4: Continuous Professional Development (7 items). 1. I can follow current developments in AI tools and prompt engineering. 1. "Regular training and workshops for educators will ensure they stay updated with the latest AI technology advancements." (Walter, 2024 ) 2. "Provide workshops for career guidance that emphasize adaptability and the importance of continuous learning in an AI-evolving job landscape. Teach an agile mindset and provide sources to learn the newest developments." (Walter, 2024 ) 3. "The rapid evolution of AI technologies can render existing frameworks quickly outdated. As new AI capabilities evolve, educators may find themselves needing to continuously adapt their teaching strategies, frameworks and assessment processes, or risk passively developing potential gaps in knowledge and skills." (Lee & Palmer, 2025 ) 2. I strive to use current developments in AI/prompt engineering in my teaching practices. 1. "Several teachers indicated a desire to investigate AI tools with specialized functionalities, highlighting the necessity of continuously adapting AI technologies to address the changing demands of education." (ElSayary et al., 2025 ) 2. "Integrating prompt engineering into teacher education curricula may therefore serve as a bridge between digital literacy and pedagogical creativity, preparing future teachers to use AI tools reflectively and effectively." (Tümen Akyıldız, 2026 ) 3. "The deliberate goal is to eventually lead students towards a responsible use of AI, but to do so, they need to understand how one can 'talk' to an AI so that it does what it is supposed to." (Walter, 2024 ) Basic Prompt Engineering (BPE). This dimension represents the technical capacity to structure clear, context-aware instructions using specific prompt components and to iteratively refine inputs. The theoretical grounding derives from the CLEAR framework (Lo, 2023 ), the PARTS framework (Park & Choo, 2024 ), and empirical work on prompt literacy (Hwang et al., 2025 ). Seven items capture competencies including prompt construction, component specification, role assignment, and systematic documentation. Pedagogical Prompt Engineering (PPE). This dimension represents the capacity to design prompts aligned with learning objectives, differentiate instruction, and critically evaluate AI outputs. It is grounded in the TPACK framework (Mishra & Koehler, 2006 ) and its AI-specific extension, Intelligent-TPACK (Celik, 2023 ). The items address learning objective alignment, differentiated content, critical evaluation, and engagement strategies. Ethical Prompting Practice (EPP). This dimension represents the capacity to protect student data privacy, mitigate algorithmic bias, and model ethical AI use. Walter ( 2024 ) positions ethical discussions as integral to AI-integrated learning, while Hsu ( 2025 ) documents data privacy concerns in GenAI interactions. Mzwri and Turcsányi-Szabó ( 2025 ) articulate that prompt engineering guides outputs toward fairness and inclusivity. Nine items address privacy protection, bias mitigation, source verification, and ethical modeling for students. Continuous Professional Development (CPD). This dimension represents the disposition to monitor AI advancements and seek ongoing skill improvement. Foundationally grounded in Schön's reflective practice framework and supported by ElSayary et al. ( 2025 ), this dimension recognizes that rapid AI evolution requires continuous updating (Lee & Palmer, 2025 ; Walter, 2024 ). Woo et al. ( 2025 b) demonstrated that targeted workshops significantly improve prompt engineering ability. The items address awareness, adaptation, feedback-driven refinement, and training participation. 3.Results As the initial step in the scale development process, the multivariate normality assumption of the dataset was tested using Mardia's coefficients. The statistically significant skewness (117) and kurtosis (888) values (p < .001) indicated that the data did not conform to a multivariate normal distribution. Accordingly, robust estimation methods were employed throughout the analyses. The number of dimensions was determined by triangulating three criteria: Horn's Parallel Analysis (see Fig. 1 ), the Scree Plot (see Fig. 2 ), and the Kaiser criterion (eigenvalue > 1). Convergent evaluation of these results supported a four-factor structure for the scale. The EFA findings for the four-factor prompt engineering scale are presented in Table 3 . Table 3 Exploratory Factor Analysis Results of the Prompt Engineering Scale Items Dimension Loadings Factor Statistic Factor 1 Factor 2 Factor 3 Factor 4 SS Loading % of Variance Cumulative % Ethical Prompting Practice 5.78 22.20 22.20 item_21 .98 item_19 .86 item_20 .85 item_23 .83 item_22 .81 item_24 .70 item_17 .69 item_18 .66 Continuous Professional Development 3.97 15.30 37.50 item_31 .82 item_27 .81 item_26 .79 item_32 .77 item_30 .77 item_28 .76 Basic Prompt Engineering 3.91 15.00 52.50 item_4 .86 item_6 .82 item_3 .80 item_5 .70 item_2 .65 item_8 .59 item_9 .50 Pedagogical Prompt Engineering 3.15 12.10 64.60 item_11 .97 item_10 .83 item_12 .77 item_15 .70 item_16 .46 Inter-Factor Correlations 1 2 3 1. Basic Prompt Engineering 2. Continuous Professional Development .59 3. Ethical Prompting Practice .63 .51 4. Pedagogical Prompt Engineering .56 .59 .65 Note. N = 300. KMO = .942; Bartlett's Test of Sphericity χ²(325) = 6173, p < .001. Model fit measures: RMSEA = .076 [.069, .084], TLI = .902, χ²(227) = 624, p < .001. Promax oblique rotation and principal axis factoring were used as the factor extraction and rotation methods, respectively. Factor loadings above .40 are bolded. The KMO value of .942 and Bartlett's Test of Sphericity, χ²(325) = 6173, p < .001, confirmed that the data were suitable for factor analysis. EFA conducted using principal axis factoring and Promax rotation yielded a four-factor structure in which all retained factors had eigenvalues exceeding 1. These four factors collectively accounted for 64.60% of the total variance. The individual contributions of each factor to the total variance were 22.20% (Basic Prompt Engineering), 15.30% (Continuous Professional Development), 15.00% (Ethical Prompting Practice), and 12.10% (Pedagogical Prompt Engineering), respectively. Item factor loadings ranged from .46 to .97, and inter-factor correlation coefficients ranged from .51 to .65. Examination of the model fit indices, RMSEA = .076 [.069, .084], TLI = .902, and χ²/df = 2.75, indicates that the model demonstrated an acceptable level of fit to the data (Kline, 2011 ; Tabachnick & Fidell, 2013 ). Table 4 Confirmatory Factor Analysis Results Basic Prompt Engineering B SE ( B ) z p β δ α AVE ω CR .865 .490 .871 .870 item_2 1.00 .590 0.652 item_3 1.34 0.149 8.99 .000 .775 0.400 item_4 1.49 0.160 9.27 .000 .832 0.309 item_5 1.42 0.159 8.92 .000 .756 0.428 item_6 1.30 0.150 8.68 .000 .756 0.429 item_8 1.05 0.143 7.33 .000 .575 0.669 item_9 1.14 0.150 7.61 .000 .585 0.658 Pedagogical Prompt Engineering .880 .605 .886 .882 item_10 1.00 .766 0.413 item_11 1.12 0.061 18.24 .000 .827 0.317 item_12 0.96 0.067 14.40 .000 .779 0.392 item_15 0.91 0.066 13.69 .000 .754 0.432 item_16 0.84 0.068 12.32 .000 .744 0.446 Ethical Prompting Practice .938 .659 .939 .939 item_17 1.000 0.674 0.545 item_18 1.179 0.075 15.63 .000 0.786 0.383 item_19 1.271 0.090 14.18 .000 0.856 0.267 item_20 1.319 0.093 14.24 .000 0.882 0.221 item_21 1.311 0.098 13.44 .000 0.856 0.268 item_22 1.182 0.106 11.19 .000 0.765 0.414 item_23 1.199 0.089 13.41 .000 0.847 0.282 item_24 1.155 0.091 12.72 .000 0.811 0.342 Continuous Professional Development .887 .566 .886 .888 item_26 1.000 0.733 0.463 item_27 1.048 0.058 18.14 .000 0.809 0.346 item_28 0.923 0.068 13.56 .000 0.771 0.406 item_30 1.003 0.071 14.23 .000 0.758 0.425 item_31 0.965 0.076 12.68 .000 0.676 0.544 item_32 1.033 0.073 14.16 .000 0.779 0.393 Note. B : Unstandardized factor loading; SE ( B ): Standard error of the unstandardized factor loading; β : Standardized factor loading; δ : Standardized error variance; AVE: Average Variance Extracted; α : Internal consistency coefficient; ω : McDonald's omega; CR: Composite Reliability. Whether the 26-item, four-factor structure identified through EFA (Basic Prompt Engineering, Pedagogical Prompt Engineering, Ethical Prompting Practice, and Continuous Professional Development) was confirmed was tested via CFA in the second study group comprising 305 participants. Examination of the robust fit indices derived from Robust Maximum Likelihood (RML) estimation yielded the following values: χ²/df = 2.03 (595/293), RMSEA = .064 [95% CI: .057, .071], SRMR = .057, CFI = .929, and TLI = .921, collectively indicating that the model demonstrated good fit to the data (Brown, 2015 ; Kline, 2011 ). Standardized factor loadings (β) ranged from .575 to .882, all statistically significant at p < .001. Examination of the reliability analysis results revealed that Cronbach's α, McDonald's ω, and CR values exceeded .80 across all subscales, reaching as high as .939 for the Ethical Prompting Practice subscale, indicating that the scale's internal consistency is very high. AVE values exceeding .50 demonstrate that the respective latent variable (factor) accounts for more than half of the variance in its associated items (Hair et al., 2019 ). Composite reliability (CR) coefficients ranged from .87 to .94, and CR values exceeding .70 across all factors confirm that the scale possesses a high level of internal consistency (Fornell & Larcker, 1981 ; Hair et al., 2019 ). Although the AVE value for the Basic Prompt Engineering subscale (.490) fell marginally below the .50 threshold, the high reliability coefficients for the same subscale (α = .865, ω = .871) and a CR value exceeding .70 indicate that the reliability of this subscale remains within acceptable limits (Fornell & Larcker, 1981 ). The research findings collectively support the conclusion that the developed instrument possesses a valid and reliable structure. A KMO value exceeding .90 indicates that the sample size was at an "excellent" level for factorization. Item factor loadings ranging from .46 to .97, combined with the total variance explained by the scale exceeding the 60% threshold at 64.60%, provide compelling evidence that the factor structure effectively represents the measured construct. Turning to the CFA results, CFI and TLI values exceeding .90, alongside RMSEA and SRMR values falling below .08, confirm that the four-factor model is statistically congruent with the theoretical structure. Inter-factor correlations ranging from .51 to .65 indicate that the subscales are interrelated yet measure conceptually distinct constructs. Examination of Cronbach's α, McDonald's ω, and CR values across all subscales further confirms that the scale is reliable. In conclusion, the 26-item, four-subscale 5-point Likert prompt engineering scale, comprising Basic Prompt Engineering, Pedagogical Prompt Engineering, Ethical Prompting Practice, and Continuous Professional Development, can be considered a valid and reliable instrument for measuring the relevant competencies of pre-service teachers. 4. Discussion This study addressed a measurement gap in AI-integrated teacher education by developing and validating the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS). The study was guided by three research questions addressing (1) the factor structure of the Ed-PESS based on EFA, (2) the extent to which CFA supports that structure, and (3) the reliability and convergent validity evidence for the scale and its subscales. Overall, the results provide initial validity evidence supporting the interpretation of Ed-PESS scores as a multidimensional measure of pre-service teachers' prompt engineering self-efficacy. To address non-normality, the CFA of the Ed-PESS was estimated using Maximum Likelihood with the Yuan and Bentler robust correction. In an independent validation sample (n = 305), model fit was acceptable (χ²/df = 2.03, RMSEA = .064 [95% CI: .057 to .071], SRMR = .057, CFI = .929, TLI = .921), providing internal structure evidence for the proposed four-factor solution. Converging EFA and CFA results supported a stable four-factor structure of the Ed-PESS: Basic Prompt Engineering, Pedagogical Prompt Engineering, Ethical Prompting Practice, and Continuous Professional Development. The EFA solution explained 64.60% of the total variance. In the CFA, standardized loadings ranged from .575 to .882 (all p < .001), indicating coherent relations between items and factors. Convergent validity was generally adequate (AVE: Pedagogical = .605; Ethical = .659; CPD = .566), while the Basic dimension was borderline (AVE = .490), suggesting broader or more heterogeneous operational content despite strong reliability. Internal consistency was high across dimensions (α and ω > .80; Ethical ≈ .94). Inter-factor correlations were moderate (r = .507-.645), consistent with related yet distinguishable facets within a broader self-efficacy construct. Content-focused validity evidence for the Ed-PESS was supported through a systematic review of the prompt engineering literature, expert review of item relevance and clarity (measurement, educational technology, and language experts), and pilot testing with target-population feedback followed by item refinement. Consistent with recent studies, the results support treating prompt related self-efficacy as a construct that can be strengthened through targeted instruction and operationalized for assessment in educational settings (e.g., Davila-Moran et al., 2025 ; Jang et al., 2025 ). At the same time, the four-factor pattern suggests that prompt engineering self-efficacy in teacher education may include related but distinct parts. This goes beyond the single factor structure reported in a competence scale developed with general users (Gibreel & Arpaci, 2025 ). 4.1. Interpreting the Hierarchy of Ed-PESS Basic Prompt Engineering. The first dimension, Basic Prompt Engineering, accounted for the largest proportion of variance in the EFA (22.2%), indicating that operational prompt-structuring is a core component of the Ed-PESS. This dimension showed strong internal consistency (α = .865; ω was approximately .868 to .871), while convergent validity was borderline (AVE = .490). This result suggests slightly limited convergent validity and may indicate that the factor covers a broader range of operational self-efficacy beliefs rather than a narrowly uniform domain. Its moderate correlation with Pedagogical Prompt Engineering (r = .591) suggests that the two dimensions are related but not the same. Confidence in structuring prompts is linked to confidence in instructional integration, but they represent different aspects of self-efficacy. Theoretically, this operational dimension fits Cognitive Load Theory (CLT). When the “syntax” of prompting, such as stating roles, constraints, context, and output formats, uses a lot of working memory, it can create extraneous load. This extra load can reduce the mental resources needed for pedagogical reasoning and for checking the quality of the output (Sweller, 1988 ). Therefore, self-efficacy in Basic Prompt Engineering can be seen as confidence in handling these basic parts smoothly and with little unnecessary mental effort, so attention can shift to instructional planning. This CLT-based view also matches Mendes et al. ( 2025 ), who highlight the importance of reducing extraneous load and increasing mental effort that supports learning in LLM-supported learning, especially at early stages. Finally, the basic elements captured here align with component-based prompt descriptions such as the Input, Instruction, Output, and Context (IIOC) framing reported by Jin et al. ( 2025 ). This supports the idea that Basic Prompt Engineering is a starting structure that can help later pedagogical use, but it does not guarantee it. Pedagogical Prompt Engineering. In the Ed-PESS, Basic and Pedagogical Prompt Engineering were moderately correlated (r = .591), and inter-factor correlations across the model ranged from r = .507 to r = .645, indicating a coherent yet multidimensional construct. Reliability was strong for both the Basic (α = .865; ω ≈ .868-.871) and Pedagogical (α = .880; ω ≈ .883-.886) dimensions, and convergent validity was adequate for Pedagogical Prompt Engineering (AVE = .605). Together, these results support interpreting pedagogical prompting as a distinct self-efficacy facet centred on instructional integration, not merely operational prompt construction. Theoretically, this distinction fits Shulman’s ( 1986 ) view of teaching as professional judgment. In this view, teachers combine procedures with content and clear reasons, while also responding to the demands of the situation. This includes paying attention to students’ ideas and misconceptions and choosing explanations that make the topic easier to learn. This is consistent with the TPACK/TPCK argument that effective technology integration depends on understanding the dynamic relationships among technology, pedagogy, and content, because knowing how to use technology is not equivalent to knowing how to teach with it (Mishra & Koehler, 2006 ). Celik’s ( 2023 ) AI-TPACK extension similarly emphasizes that technological knowledge alone is insufficient and underscores the centrality of pedagogical knowledge for ethically and effectively integrating AI-based tools. Additionally, from an instrument-mediated activity perspective, Rabardel’s ( 2002 ) theory of instrumental genesis provides a complementary lens: GenAI functions as an instructional instrument only when coupled with teachers’ utilization schemes adapted to the specificity of each situation. Accordingly, higher self-efficacy in Pedagogical Prompt Engineering reflects confidence in using GenAI in a pedagogically purposeful way. This includes interpreting instructional demands, turning outputs into usable learning resources, and monitoring and steering the process in context (Rabardel, 2002 ). Taken together, these findings suggest that teacher education should move beyond “chatting with AI” toward structured, case-based design experiences (e.g., prompt labs, microteaching with constraints, and iterative revision tasks) that cultivate instructional alignment and reflective judgment. Ethical Prompting Practice. In the Ed-PESS, Ethical Prompting Practice formed a highly coherent subdomain of prompt engineering self-efficacy. Internal consistency was excellent (α = .938; ω ≈ .939) and convergent validity was strong (AVE = .659), indicating substantial shared variance between the items and the latent ethical self-efficacy factor. Inter-factor correlations (r = .507-.645) suggest that ethical self-efficacy is meaningfully connected to the other dimensions while remaining empirically distinguishable. Bandura’s ( 2001 ) account of moral agency in Social Cognitive Theory provides a useful lens for interpreting Ethical Prompting Practice as ethically oriented self-efficacy: moral conduct is shaped by self-regulatory processes that must be actively engaged, particularly in contexts where responsibility can be diffused or displaced. In this view, ethical prompting self-efficacy refers to pre-service teachers’ confidence that they can act ethically when using GenAI. This means setting clear safeguards and limits in their GenAI interactions and keeping human responsibility for decisions. In the present scale, this is reflected in item content addressing privacy/data protection, bias and age-appropriateness guardrails, reliability-oriented constraints, and modelling/teaching ethical prompting norms to students. Based on this view of moral agency, UNESCO explains ethical AI use in education in AI Competency Framework for Teachers. The framework focuses on decision processes that are “human controlled and human accountable.” It also clearly states that “Teachers should remain accountable for pedagogical decisions” when they use AI (UNESCO, 2024 ). This policy view supports how we understand Ethical Prompting Practice in the Ed-PESS. It suggests that teachers should feel confident about protecting privacy and safety, reducing bias, and checking that outputs are reliable. They do this by writing prompts with clear rules and limits, instead of leaving responsibility to the tool. This idea is also supported by other studies. Agirdag ( 2025 ) highlights that ethical prompting involves improving prompts over time and noticing bias during the dialogue with AI. This matters because teachers need to shape prompts so the outputs are safe, fair, and suitable for learning. Selwyn et al. ( 2025 ) also show that teachers often check outputs carefully, fix problems, and sometimes rewrite the outputs. Overall, these studies support Ethical Prompting Practice as a practical responsibility in use, not only a technical prompting skill. Continuous Professional Development (CPD). In the Ed-PESS, CPD is a separate but connected part of prompt engineering self-efficacy. The CPD items showed strong internal consistency (α = .887; ω was approximately .884 to .886) and acceptable convergent validity (AVE = .566). The correlations between factors ranged from .507 to .645. This suggests that CPD is related to the other dimensions but still distinct. These results suggest that CPD is not just an attitude. Instead, it reflects self-efficacy for keeping prompt related practices up to date and improving them as GenAI tools and classroom expectations change. Artino ( 2012 ) argues that self-efficacy beliefs matter in technology-mediated learning because learners try to control their own learning. Zimmerman’s ( 2002 ) self-regulation framework also supports this view. It describes ongoing cycles of planning, doing, and reflecting, including setting goals, planning strategies, monitoring progress, evaluating results, and adapting. In line with this, CPD in the Ed-PESS refers to perceived capability to follow new developments in AI and prompting, apply updates in teaching, and keep improving through reflection (for example, revising prompts based on feedback and setting professional learning goals). In addition, this view is supported by recent studies. Tümen Akyıldız ( 2026 ) describes prompting as a new practice that can feel experimental and uncertain. For this reason, competent GenAI use requires repeated updating and adaptation. This supports the idea that CPD reflects confidence in improving over time by trying, checking results, and making changes through repeated cycles of writing prompts, evaluating outputs, and revising prompts. Consistent with this view, ElSayary et al. ( 2025 ) highlight that effective classroom use of GenAI requires AI literacy, continuous professional development, and repeated improvement of prompts and outputs. 4.2. Theoretical Implications: Prompting as "Teacher Agency" Taken together, the Ed-PESS supports viewing prompt engineering self-efficacy as teacher agency in GenAI-supported instruction. Teachers are not passive recipients of outputs. They actively regulate how GenAI is used and for what teaching goals. In an instrument mediated activity view, GenAI is an artifact. It becomes an instructional instrument only when it is linked to teachers’ utilization schemes. Prompt engineering self-efficacy is the confidence to build, adapt, and coordinate these schemes to support learning goals (Rabardel, 2002 ). In addition, moral agency makes this view clearer. Ethical Prompting Practice in GenAI use is not guaranteed by abstract reasoning alone. It depends on teachers’ confidence that they can activate self-regulatory safeguards and keep human accountability, especially in settings where responsibility can be spread across others or shifted to the tool (Bandura, 2006 ). Most importantly, teacher agency in GenAI-supported teaching works like a repeat cycle. Teachers check if the output fits the teaching goal and if it causes ethical problems. If the output is weak, they find the reason and then change the prompt. This process matches three parts of self-efficacy in the Ed-PESS: Pedagogical Prompt Engineering, Continuous Professional Development, and Ethical Prompting Practice. 4.3. Practical Implications for Teacher Education General lectures that introduce AI are not enough to build prompt engineering self-efficacy as teacher agency (Bandura, 2001 , 2006 ; Zimmerman, 2002 ). Teacher education curricula should include Prompt Labs. In these sessions, pre-service teachers practise writing prompts for clear teaching goals. They also add safety and ethics rules and improve prompts and outputs step by step in realistic teaching tasks. This gives mastery experiences that can strengthen self-efficacy (Bandura, 2006 ). These labs can follow guided cycles of planning, doing, and reflecting. In this way, teacher candidates learn why an output did not work and how to change the prompt, instead of using GenAI as a tool that gives an answer in one attempt (Zimmerman, 2002 ). To support evidence informed curriculum design, the Ed-PESS can be used as a pre-test and post-test in educational technology or methods courses. This can help instructors in two ways. First, it can show teacher candidates’ self-efficacy profiles across the different dimensions. Second, it can show whether a specific teaching activity improves the targeted parts of prompt engineering self-efficacy. Teacher education should also include clear modules on ethical use. These modules should train teacher candidates to write prompts with safety and fairness rules. For example, they can ask for gender neutral roles, require child appropriate content, and request fair examples. The courses should also teach routines for checking outputs and revising them. In this way, ethical prompting practice remains a teachable and assessable part of professional practice, not something that is expected to happen automatically when using the tool. 4.4 Limitations and Future Directions Although this study provides strong evidence on the internal structure and reliability of the Ed-PESS, several limitations affect interpretation and point to next steps. First, the Ed-PESS is a self-report scale. It shows what candidates think they can do when they design, adapt, and safeguard GenAI prompts. It does not show what they do in real tasks. Future studies should combine Ed-PESS scores with performance measures. For example, researchers can use rubric scored tasks where candidates write prompts, check outputs, and revise them under pedagogical and ethical rules. This can test criterion validity and show how beliefs match observed prompting performance. Second, we validated the scale in teacher education programs in one national context. Studies with more diverse cohorts and institutions are needed to improve generalizability. We did not test measurement invariance. Future research should test invariance across key groups, such as gender, year level, prior GenAI training, and AI tool familiarity. It should also test invariance across cultures before using the scale for group comparisons. Third, prompting is shaped by a fast-changing technology context. Models, interfaces, and modes change, and model differences can affect prompting and how it is evaluated. For this reason, the Ed-PESS should be seen as measuring basic self-efficacy for pedagogically purposeful, ethically responsible, and developmentally appropriate prompting, not model specific tricks. Regular monitoring and planned item review are needed to protect content validity as new classroom needs and risks appear. Overall, these steps position the Ed-PESS as a practical base for research and formative evaluation that tracks how teacher self-efficacy develops as GenAI tools change. 5.Conclusion The main purpose of this study was to develop and validate the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS). This responds to the growing need for reliable measurement in AI integrated teacher education. Psychometric results supported a clear four factor structure. The EFA explained 64.60 percent of the total variance. The CFA, with a robust correction, showed good model fit (CFI = .929, RMSEA = .064, SRMR = .057). Reliability evidence was also strong across the scale. These results suggest that the Ed-PESS can support teacher education in a practical way. It can help prepare future teachers to stay human in the loop and guide GenAI use toward pedagogically sound and ethically responsible goals. This study also makes a theoretical and methodological contribution. To our knowledge, it is the first validated instrument designed to assess prompt engineering self-efficacy beliefs among pre-service teachers. In practice, the Ed-PESS addresses a gap by enabling systematic assessment and targeted development of this emerging self-efficacy construct in teacher education curricula. Declarations Ethical Approval. This research received ethical approval from the Scientific Research and Publication Ethics Committee at [removed for blinded review] , as per decision number [removed for blinded review]. Declaration of generative AI and AI-assisted technologies in the writing process. During the preparation of this research, the authors used Paperpal to paraphrase their writing for more academic enhancement, and Claude Opus (AI tools) for language translation and text reduction. After using these AI tools, the author(s) reviewed and edited the content as necessary and take full responsibility for the content of the publication. Alongside using AI tools for language tasks, the author(s) thoroughly reviewed and accurately cited all references in this research. They verified each reference's authenticity, including DOI links. It's crucial to note that all data and findings come from properly cited sources, not AI-generated. The author(s) fully ensure the research's integrity and accuracy. Author Contributions Statement. Author 1 conceptualized the research, designed the study, and supervised the project. Author 2 performed the data analysis and validated the results. Author 3 contributed to the methodology and wrote the discussion section. Author 4 conducted the literature review and wrote the introduction. Author 5 collected the data and critically reviewed the manuscript. All authors reviewed and approved the final manuscript. Competing interests. The authors declare that they have no competing interests related to this research. Funding. The author reports no funding. Data Availability. The data and materials used in this study are available upon request from the corresponding author. References Agirdag, O. (2025). Beyond prompt engineering: Prompting (l)iteracy, linguistic capital, and educational inequality. Educational Theory . https://doi.org/10.1111/edth.70057 Andewi, W., Waziana, W., Wibisono, D., Putra, K. A., Hastomo, T., & Oktarin, I. B. (2025). From prompting to proficiency: A mixed-methods analysis of prompting with ChatGPT versus lecturer interaction in an EFL classroom. Journal of Studies in the English Language , 20 (2), 210–238. https://doi.org/10.64731/jsel.v20i2.282318 Arocha, J. G. (2025). Critical phenomenology of prompting in artificial intelligence. Sophia-Colección de Filosofía de la Educación , 39 , 141–165. https://doi.org/10.17163/soph.n39.2025.04 Artino, A. R. Jr. (2012). Academic self-efficacy: From educational theory to instructional practice. Perspectives on Medical Education , 1 (2), 76–85. https://doi.org/10.1007/s40037-012-0012-5 Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of Psychology , 52 (1), 1–26. https://doi.org/10.1146/annurev.psych.52.1.1 Bandura, A. (2006). Guide for constructing self-efficacy scales. In F. Pajares, & T. Urdan (Eds.), Self-efficacy beliefs of adolescents (pp. 307–337). Information Age Publishing. Biberman-Shalev, L. (2025). Prompting theory into practice: Utilizing ChatGPT-4 in a curriculum planning course. Education Sciences , 15 (2). Article 196. https://doi.org/10.3390/educsci15020196 Boateng, G. O., Neilands, T. B., Frongillo, E. A., Melgar-Quiñonez, H. R., & Young, S. L. (2018). Best practices for developing and validating scales for health, social, and behavioral research: A primer. Frontiers in Public Health , 6 , 149. https://doi.org/10.3389/fpubh.2018.00149 Boonstra, L. (2025). Prompt engineering [Whitepaper]. Google. https://www.kaggle.com/whitepaper-prompt-engineering Brown, T. A. (2015). Confirmatory factor analysis for applied research . The Guilford Press. Carl, K., Dignam, C., Kochan, M., Alston, C., & Green, D. (2024). Discovering prompt engineering: A qualitative study of nonexpert teachers' interactions with ChatGPT. Issues in Information Systems , 25 (4), 205–220. https://doi.org/10.48009/4_iis_2024_117 Celik, I. (2023). Towards Intelligent-TPACK: An empirical study on teachers’ professional knowledge to ethically integrate artificial intelligence (AI)-based tools into education. Computers in Human Behavior , 138 . https://doi.org/10.1016/j.chb.2022.107468 . Article 107468. Celik, I., Kontkanen, S., Laru, J., & Dalyanci, A. A. (2026). Co-constructing adaptive lesson plans with GenAI: Pre-service teachers' Intelligent-TPACK and prompt engineering strategies. Computers and Education , 241 , 105485. https://doi.org/10.1016/j.compedu.2025.105485 Chiu, T. K. F., Ahmad, Z., & Çoban, M. (2025). Development and validation of teacher artificial intelligence (AI) competence self-efficacy (TAICS) scale. Education and Information Technologies , 30 , 6667–6685. https://doi.org/10.1007/s10639-024-13094-z Correia, A. P., Hickey, S., & Xu, F. (2025). Realizing the possibilities of the large language models: Strategies for prompt engineering in educational inquiries. Theory Into Practice , 64 (4), 434–447. https://doi.org/10.1080/00405841.2025.2528545 Crosthwaite, P. R., Smala, S., & Spinelli, F. (2025). Prompting for pedagogy? Australian F-10 teachers' generative AI prompting use cases. Australian Educational Researcher , 52 (3), 1795–1825. https://doi.org/10.1007/s13384-024-00787-0 Davidson, T. R., & Karell, D. (2025). Integrating generative artificial intelligence into social science research: Measurement, prompting, and simulation. Sociological Methods and Research , 54 (3), 775–793. https://doi.org/10.1177/00491241251339184 Davila-Moran, R. C., Soto, S., Gomez, J. M. L., Silva Infantes, H. E., Arias Lizares, M., Huanca Rojas, A., L. M., & Cama Flores, S. J. (2025). Brief prompt-engineering clinic substantially improves AI literacy and reduces technology anxiety in first-year teacher-education students: A pre-post pilot study. Education Sciences , 15 (8). https://doi.org/10.3390/educsci15081010 . Article 1010. ElSayary, A., Kuhail, M. A., & Hojeij, Z. (2025). Examining the role of prompt engineering in utilizing generative AI tools for lesson planning: Insights from teachers’ experiences and perceptions. Human Behavior and Emerging Technologies . https://doi.org/10.1155/hbe2/9986139 Federiakin, D., Molerov, D., Zlatkin-Troitschanskaia, O., & Maur, A. (2024). Prompt engineering as a new 21st century skill. Frontiers in Education , 9 , 1366434. https://doi.org/10.3389/feduc.2024.1366434 Fornell, C., & Larcker, D. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research , 18 (1), 39–50. https://doi.org/10.2307/3151312 Gibreel, O., & Arpaci, I. (2025). Development and validation of the prompt engineering competence scale (PECS). Information Development . Advance online publication. https://doi.org/10.1177/02666669251336455 Gümüş, M., & Kukul, V. (2023). Developing a digital competence scale for teachers: Validity and reliability study. Education and Information Technologies , 28 , 2747–2765. https://doi.org/10.1007/s10639-022-11411-z Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2019). Multivariate data analysis . Cengage Learning EMEA. Hsu, H. P. (2025). From programming to prompting: Developing computational thinking through large language model-based generative artificial intelligence. TechTrends , 69 (3), 485–506. https://doi.org/10.1007/s11528-025-01052-6 Huang, K. L., Liu, Y. C., & Dong, M. Q. (2024). Incorporating AIGC into design ideation: A study on self-efficacy and learning experience acceptance under higher-order thinking. Thinking Skills and Creativity , 52 , 101508. https://doi.org/10.1016/j.tsc.2024.101508 Hwang, M., Jeens, R., & Lee, H. K. (2025). Exploring learner prompting behavior and its effect on ChatGPT-assisted English writing revision. The Asia-Pacific Education Researcher , 34 , 1157–1167. https://doi.org/10.1007/s40299-024-00930-6 Jacobsen, L. J., & Weber, K. E. (2025). The promises and pitfalls of large language models as feedback providers: A study of prompt engineering and the quality of AI-driven feedback. AI , 6 (2). https://doi.org/10.3390/ai6020035 Jang, A., Oh, M., & Song, M. O. (2025). Question-centered prompt engineering for nursing simulation debriefing: Model development and validation. Nurse Education in Practice , 88 ., Article 104541. https://doi.org/10.1016/j.nepr.2025.104541 Jin, Z., Meng, G., Wang, X., Wang, J., Liu, C., & Zhang, J. (2025). Understanding user prompting behavior in generative AI: A component analysis. Proceedings of the Association for Information Science and Technology, 62 (1), 941–945. https://doi.org/10.1002/pra2.1318 Kabeer, A., Bhat, R. A., Antony, S., & Tramboo, I. A. (2025). Enhancing creative writing skills in secondary school students through prompt engineering and artificial intelligence. Forum for Linguistic Studies , 7 (3), 800–815. https://doi.org/10.30564/fls.v7i3.8511 Kang, L., Shi, X., & Zhu, K. (2025). Uncovering the mediation of disciplinary literacy in the effect of GAI prompt engineering on pre-service teachers' instructional design. Education and Information Technologies , 30 (16), 22779–22802. https://doi.org/10.1007/s10639-025-13657-8 Kline, R. B. (2011). Principles and practice of structural equation modeling . The Guilford Press. Laupichler, M. C., Aster, A., Haverkamp, N., & Raupach, T. (2023). Development of the Scale for the assessment of non-experts' AI literacy: An exploratory factor analysis. Computers in Human Behavior Reports , 12 (4). Article 100338. https://doi.org/10.1016/j.chbr.2023.100338 Lawshe, C. H. (1975). A quantitative approach to content validity. Personnel Psychology , 28 (4), 563–575. https://doi.org/10.1111/j.1744-6570.1975.tb01393.x Lee, D., & Palmer, E. (2025). Prompt engineering in higher education: A systematic review to help inform curricula. International Journal of Educational Technology in Higher Education , 22 (1). https://doi.org/10.1186/s41239-025-00503-7 Lee, U., Jung, H., Jeon, Y., Sohn, Y., Hwang, W., Moon, J., & Kim, H. (2024). Few-shot is enough: Exploring ChatGPT prompt engineering method for automatic question generation in English education. Education and Information Technologies , 29 (9), 11483–11515. https://doi.org/10.1007/s10639-023-12249-8 Lo, L. S. (2023). The CLEAR path: A framework for enhancing information literacy through prompt engineering. Journal of Academic Librarianship , 49 (4), 102720. https://doi.org/10.1016/j.acalib.2023.102720 Long, D., & Magerko, B. (2020). What is AI literacy? Competencies and design considerations. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems (pp. 1–16). Association for Computing Machinery. https://doi.org/10.1145/3313831.3376727 Mendes, A., Parizotto, A., Garcia, R., Garcia, M., Guedes, G., Vilela, R., Valle, P., Balancieri, R., & Silva, W. (2025). CoderBot 2.0: Integrating LLM and prompt engineering in the evolution of an educational chatbot. In Simpósio Brasileiro de Informática na Educação (SBIE) (pp. 1497–1506). SBC. https://doi.org/10.5753/sbie.2025.12321 Mishra, P., & Koehler, M. J. (2006). Technological pedagogical content knowledge: A framework for teacher knowledge. Teachers College Record , 108 (6), 1017–1054. https://doi.org/10.1177/016146810610800610 Moorhouse, B. L., Ho, T. Y., Wu, C., & Wan, Y. (2025). Pre-service language teachers' task-specific large language model prompting practices. RELC Journal . https://doi.org/10.1177/00336882251313701 . Advance online publication. Mutanga, M. B., Msane, J., Mndaweni, T. N., Hlongwane, B. B., & Ngcobo, N. Z. (2025). Exploring the impact of LLM prompting on students' learning. Trends in Higher Education , 4 (3). Article 31. https://doi.org/10.3390/higheredu4030031 Mzwri, K., & Turcsányi-Szabó, M. (2025). The impact of prompt engineering and a generative AI-driven tool on autonomous learning: A case study. Education Sciences , 15 (2). Article 199. https://doi.org/10.3390/educsci15020199 Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021). Conceptualizing AI literacy: An exploratory review. Computers and Education: Artificial Intelligence , 2 , 100041. https://doi.org/10.1016/j.caeai.2021.100041 Oliveira, L., Tavares, C., Strzelecki, A., & Silva, M. (2025). Prompting minds: Evaluating how students perceive generative AI's critical thinking dispositions. Electronic Journal of e-Learning , 23 (2), 1–18. https://doi.org/10.34190/ejel.23.2.3986 Park, J., & Choo, S. (2024). Generative AI prompt engineering for educators: Practical strategies. Journal of Special Education Technology , 40 (3), 411–417. https://doi.org/10.1177/01626434241298954 Pratschke, B. M. (2024). Generative AI and education: Digital pedagogies, teaching innovation and learning design . Springer. https://doi.org/10.1007/978-3-031-67991-9 Qian, Y. (2025). Prompt engineering in education: A systematic review of approaches and educational applications. Journal of Educational Computing Research . https://doi.org/10.1177/07356331251365189 . Advance online publication. Rabardel, P. (2002). People and technology in the workplace: A cognitive ergonomic approach . HAL Open Science. Selwyn, N., Ljungqvist, M., & Sonesson, A. (2025). When the prompting stops: Exploring teachers’ work around the educational frailties of generative AI tools. Learning Media and Technology , 50 (3), 310–323. https://doi.org/10.1080/17439884.2025.2537959 Serra, P., & Oliveira, A. (2025). AI-powered prompt engineering for Education 4.0: Transforming digital resources into engaging learning experiences. Education Sciences , 15 (12). Article 1640. https://doi.org/10.3390/educsci15121640 Shulman, L. S. (1986). Those who understand: Knowledge growth in teaching. Educational Researcher , 15 (2), 4–14. https://doi.org/10.3102/0013189X015002004 Sigot, M., & Tassoti, S. (2025). An investigation of change in prompting strategies in a semester-long course on the use of GenAI. Journal of Chemical Education , 102 (6), 2507–2513. https://doi.org/10.1021/acs.jchemed.4c01287 Son, M., Won, Y. J., & Lee, S. (2025). Optimizing large language models: A deep dive into effective prompt engineering techniques. Applied Sciences , 15 (3), 1430. https://doi.org/10.3390/app15031430 Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science , 12 (2), 257–285. https://doi.org/10.1207/s15516709cog1202_4 Tabachnick, B. G., & Fidell, L. S. (2013). Using multivariate statistics (6th ed.). Pearson Education. Tümen Akyıldız, S. (2026). Prompt engineering and creative pedagogical thinking: Insights from a qualitative study with teacher candidates. Thinking Skills and Creativity , 60 ., Article 102105. https://doi.org/10.1016/j.tsc.2025.102105 Tzafilkou, K., Perifanou, M., & Economides, A. A. (2022). Development and validation of students' digital competence scale (SDiCoS). International Journal of Educational Technology in Higher Education , 19 (1). Article 30. https://doi.org/10.1186/s41239-022-00330-0 UNESCO. (2024). AI competency framework for teachers . UNESCO Publishing. Van Wyk, M. M. (2025). Student teachers' leveraging GenAI tools for academic writing, design, and prompting in an ODeL course. Open Praxis , 17 (1), 95–107. https://doi.org/10.55982/openpraxis.17.1.711 Van Zoost, S. (2025). Prompting teacher identities: A model for teacher subjectivities constituted through artificial intelligence. Journal of Teaching and Learning , 19 (4), 198–215. https://doi.org/10.22329/jtl.v19i4.10079 Velásquez-Henao, J. D., Franco-Cardona, C. J., & Cadavid-Higuita, L. (2023). Prompt engineering: A methodology for optimizing interactions with AI-language models in the field of engineering. DYNA , 90 (230), 9–17. https://doi.org/10.15446/dyna.v90n230.111700 Walter, Y. (2024). Embracing the future of artificial intelligence in the classroom: The relevance of AI literacy, prompt engineering, and critical thinking in modern education. International Journal of Educational Technology in Higher Education , 21 (1). https://doi.org/10.1186/s41239-024-00448-3 Wang, B., Rau, P. L. P., & Yuan, T. (2023). Measuring user competence in using artificial intelligence: Validity and reliability of artificial intelligence literacy scale. Behaviour & Information Technology , 42 (9), 1324–1337. https://doi.org/10.1080/0144929X.2022.2072768 Wang, Y. Y., & Chuang, Y. W. (2024). Artificial intelligence self-efficacy: Scale development and validation. Education and Information Technologies , 29 (4), 4785–4808. https://doi.org/10.1007/s10639-023-12015-w Woo, D. J., Wang, D., Yung, T., & Guo, K. (2025). Effects of a prompt engineering intervention on undergraduate students' AI self-efficacy, AI knowledge and prompt engineering ability: A mixed methods study. British Educational Research Journal . Advance online publication. https://doi.org/10.1002/berj.70087 Zimmerman, B. J. (2002). Becoming a self-regulated learner: An overview. Theory Into Practice , 41 (2), 64–70. https://doi.org/10.1207/s15430421tip4102_2 Additional Declarations No competing interests reported. Supplementary Files Appendix1.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9003817","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":598890698,"identity":"7f28ab27-0fe3-476d-ba5a-2246411a4660","order_by":0,"name":"Fatih KARATAS","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABI0lEQVRIie3RsWqDQBjA8U8EXS5xvRCIr3ASSIe25FVODuzi0NGhELM0Y9YTQp+hU6YOhoN0EVwtdGg4cA+FEreqTQslZ+jY4f5wcif8uA8F0On+YXYMIADo8Ri1T+vrnKoJSs1vYtYr+02MuIOYP8S4PxI4R+x8I24hcC8W+WZfPbyOSC62b7snGPULastIQRADwSH0Vhkzh711OSZFcENoCeNBQY15dkqmUBMEkcHrzdBYCz/haIJpCv5jQxSTIUe2ZModaVbVSsySZf7RkFknwe0toc8xs3AvFtSB0GoIJZ1EEoFIwDiWk0u0Fd4SB81g2Euy3TxRDubLdxSxa15vXg53wrUcUQ4O6ZXbf2bpXvWV28jpK1wv5W/R6XQ63R/6BKo9ad+Mf3YQAAAAAElFTkSuQmCC","orcid":"","institution":"Nevşehir Hacı Bektaş Veli University","correspondingAuthor":true,"prefix":"","firstName":"Fatih","middleName":"","lastName":"KARATAS","suffix":""},{"id":598890699,"identity":"e58537be-307d-4ae3-b013-8692a21633e2","order_by":1,"name":"Recep GÜR","email":"","orcid":"","institution":"Osman Gazi University","correspondingAuthor":false,"prefix":"","firstName":"Recep","middleName":"","lastName":"GÜR","suffix":""},{"id":598890700,"identity":"bb68911c-7c8f-45a7-99cb-56f8e80e3220","order_by":2,"name":"Barış ERİÇOK","email":"","orcid":"","institution":"Ordu University","correspondingAuthor":false,"prefix":"","firstName":"Barış","middleName":"","lastName":"ERİÇOK","suffix":""},{"id":598890701,"identity":"f3faa8c8-3c05-49ef-85f7-9fa01443f506","order_by":3,"name":"Fatma Başarır","email":"","orcid":"","institution":"Nevşehir Hacı Bektaş Veli University","correspondingAuthor":false,"prefix":"","firstName":"Fatma","middleName":"","lastName":"Başarır","suffix":""},{"id":598890702,"identity":"828ac37a-aaa3-46c9-bb28-d8c09ff692a9","order_by":4,"name":"Lokman TANRIKULU","email":"","orcid":"","institution":"Nevşehir Hacı Bektaş Veli University","correspondingAuthor":false,"prefix":"","firstName":"Lokman","middleName":"","lastName":"TANRIKULU","suffix":""}],"badges":[],"createdAt":"2026-03-01 20:53:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9003817/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9003817/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":103809835,"identity":"5ff050c1-9127-4a7f-bb70-4507a4e970d9","added_by":"auto","created_at":"2026-03-03 08:04:42","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":64026,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eParallel Analysis\u003c/em\u003e\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9003817/v1/f0c2391cbdad389987d7e277.png"},{"id":103809834,"identity":"aef6b5a4-07e9-42de-90ba-7e85001287f1","added_by":"auto","created_at":"2026-03-03 08:04:42","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":44634,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eScree Plot\u003c/em\u003e\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-9003817/v1/6de972b1b835490090a15c91.png"},{"id":103809836,"identity":"73a73b59-bc00-4e84-b0dd-394e205e35f2","added_by":"auto","created_at":"2026-03-03 08:04:42","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":429244,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003ePath Diagram\u003c/em\u003e\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-9003817/v1/ea405e58cfe86fb690b0519b.png"},{"id":104412795,"identity":"4ef4c27e-662a-4803-82ff-42ffb7498b92","added_by":"auto","created_at":"2026-03-11 13:01:12","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2252463,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9003817/v1/d5945da0-6c2f-49ab-8f24-cea98a69859b.pdf"},{"id":104401098,"identity":"081db7a5-c12b-43e5-b3fd-811a1b5d56c8","added_by":"auto","created_at":"2026-03-11 12:11:51","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":28328,"visible":true,"origin":"","legend":"","description":"","filename":"Appendix1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9003817/v1/003216ecddfea063a19375fa.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Educational Prompt Engineering Self-Efficacy Scale (Ed-PESS): Scale Development and Psychometric Validation","fulltext":[{"header":"1.Introduction","content":"\u003cp\u003eThe integration of artificial intelligence (AI) into educational settings is driving a profound transformation across diverse areas, from curriculum design to assessment practices, and from evolving teacher roles to issues of accessibility and inclusivity. With the widespread adoption of AI applications, fundamental aspects of teaching including lesson planning, content development, personalized feedback, and interactive learning experiences are being reshaped. This transformation extends beyond mere technological innovation. It represents a paradigm shift in how teaching and learning are conceptualized and enacted (Serra \u0026amp; Oliveira, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Within this context, AI integration is increasingly recognized as a strategic priority for educational institutions (Pratschke, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), gaining growing attention at the institutional level in higher education (Lee \u0026amp; Palmer, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDespite this growing integration, AI's pedagogical value depends not on the technology itself but on the quality of user AI interaction. Arocha (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) theoretically demonstrates that the educational outputs of AI systems are sensitive to the structure of user inputs, while Selwyn et al. (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) emphasize that these tools require deliberate guidance from users to generate meaningful outputs. Empirical findings also support this view. Users without systematic training tend to use AI as a search engine and accept outputs without critical evaluation (Sigot \u0026amp; Tassoti, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). In this context, the creation of appropriate prompts plays a decisive role in obtaining targeted educational outputs from AI tools (ElSayary et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Park \u0026amp; Choo, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In its most basic definition, a prompt is a text-based instruction or input provided by the user to an AI system to generate a response or complete a specific task (Arocha, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Correia et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Crosthwaite et al. (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) indicates that effective prompts are aligned with learning objectives, contextually rich, and well structured, whereas ineffective prompts consist of vague or overly simplistic commands. This suggests that prompt quality directly determines pedagogical outcomes (Davidson \u0026amp; Karell, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Consequently, the ability to craft effective prompts has emerged as a critical competency for AI use in education, and the conceptual framework for purposefully structuring this interaction is referred to as educational prompt engineering (Agirdag, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn the literature, prompt engineering is defined within a rigorous computational and semantic framework as the systematic structuring and refinement of user inputs to obtain desired outputs from AI systems (Lee \u0026amp; Palmer, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Pratschke, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Although prompt engineering is closely related to AI literacy, which encompasses general competencies such as understanding and evaluating AI technologies (Long \u0026amp; Magerko, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Ng et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), it represents a distinct, practice-oriented application of this broader competence, enabling teachers to take concrete actions toward their pedagogical goals and obtain purpose-aligned outputs from AI tools (Celik et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2026\u003c/span\u003e; Kang et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). While the concept was initially developed in technical contexts, the need for domain-specific strategies has led to the emergence of a new competency area termed educational prompt engineering, which intersects with pedagogical expertise in education. This emergence establishes the basis for investigating prompt engineering as a systematically teachable competency.\u003c/p\u003e \u003cp\u003ePrompt engineering is a teachable and developable skill; therefore, its pedagogy can be empirically investigated (Lee \u0026amp; Palmer, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Recent empirical research demonstrates that structured prompt engineering training significantly influences learners' prompting strategies and learning outcomes. Woo et al. (\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2025\u003c/span\u003ea) found that students used prompt engineering not merely as a technical operation but as a tool supporting cognitive processes such as overcoming writer's block and developing story plots. In the context of teacher education, Sigot and Tassoti (\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) observed that through a structured prompting framework, students transitioned from simple search-engine-like usage to strategic and iterative prompting practices. Similarly, Hwang et al. (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) reported that students without guidance focused on superficial linguistic features and struggled to construct goal-aligned prompts. Mutanga et al. (\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) also reported that most students exhibited usage patterns relying on superficial commands. Moreover, AI-supported instructional practices and prompt engineering interventions have been shown to strengthen task-specific self-efficacy beliefs (Andewi et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Huang et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Woo et al., \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), support self-regulated learning (Mzwri \u0026amp; Turcs\u0026aacute;nyi-Szab\u0026oacute;, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), foster creativity and critical thinking (Kabeer et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), and reduce technology anxiety (Davila-Moran et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Taken together, these findings suggest that prompt engineering skills do not develop spontaneously but require intentional, pedagogically grounded instruction. This highlights the importance of systematically integrating such training into teacher education curricula.\u003c/p\u003e \u003cp\u003eWhile systematically integrating prompt engineering training into teacher education curricula is essential, the effectiveness of such training also depends on how confident teachers feel in applying these skills in instructional contexts. This points to the need for a construct that captures teachers' confidence in designing and refining pedagogically grounded prompts. Given that these competencies and beliefs are predominantly formed during pre-service education, teacher education curricula should move beyond technical usage to also address the development of educational prompt engineering self-efficacy (Moorhouse et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Serra \u0026amp; Oliveira, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). In a systematic review, Qian (\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) demonstrates that prompt engineering in education is positioned as both a technical skill and a pedagogical strategy. Accordingly, scholars argue that pedagogically productive interaction with AI requires the holistic integration of pedagogical content knowledge, AI knowledge, and prompting skills (Correia et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Moorhouse et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Critical scholarship also indicates that educational prompt engineering should be understood not merely as a technical skill but within a broader framework encompassing pedagogical purposes and contextual sensitivity (Agirdag, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Arocha, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Selwyn et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eHowever, competency alone does not guarantee effective practice; teachers must also believe in their ability to apply these skills in pedagogical contexts. Identifying and fostering this self-efficacy is therefore essential for curriculum design, as it determines whether acquired competencies translate into meaningful classroom practice. Accordingly, pre-service teachers' educational prompt engineering self-efficacy, that is, their perceived capability to apply prompt design skills in instructional tasks such as lesson planning, material development, and assessment, emerges as a key construct warranting empirical attention (Crosthwaite et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Furthermore, drawing on the theoretical framework that conceptualizes self-efficacy as task- and context-specific (Bandura, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), prompt engineering self-efficacy can be defined as an individual's subjective belief in their ability to create pedagogically meaningful prompts when interacting with AI systems, to integrate these prompts into teaching and learning processes, and to sustain this interaction within an evaluation\u0026ndash;refinement cycle. Importantly, this belief extends beyond technical competence and is closely associated with how teachers position themselves in relation to AI. This imperative, however, first requires a clear conceptualization of prompt engineering self-efficacy as a distinct construct within teacher education.\u003c/p\u003e \u003cp\u003eVan Zoost (\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) suggests that prompt engineering serves as a technology of the self, moving teachers beyond mere users to a space where they negotiate their pedagogical decisions and professional identities. Empirical evidence further supports this perspective. Biberman-Shalev (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) found that pre-service teachers, through the process of prompt refinement, recognized the personalization limitations of AI and rediscovered the value of their own pedagogical roles. Beyond this identity dimension, these perspectives also imply that teachers with stronger self-efficacy beliefs are more likely to engage in purposeful and pedagogically aligned interactions with AI tools. Taken together, these findings suggest that educational prompt engineering self-efficacy reflects not merely a technical skill but a professional capacity that mediates between knowledge and practice, shaping how teachers position themselves as pedagogically informed agents in AI-integrated environments.\u003c/p\u003e \u003cp\u003eThese conceptual and empirical considerations carry significant implications for teacher education. For future teachers, prompt engineering self-efficacy is becoming a foundational area of professional belief, as fundamental as confidence in preparing lesson plans or developing assessment instruments. Within the professional roles that pre-service teachers will assume, traditional teaching competencies such as lesson planning, assessment, and instructional material development are expected to be carried out in integration with AI tools (ElSayary et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Park \u0026amp; Choo, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This integration, however, is not straightforward. As Pratschke (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) noted, experiences from Khan Academy have shown that while having AI generate a lesson plan is easy, having it generate a \"good\" lesson plan is a demanding process requiring mastery of pedagogical principles. This suggests that pre-service teachers need to develop not only technical familiarity with AI tools but also confidence in directing them toward pedagogical purposes. Regarding how such confidence can be cultivated, Hsu (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) proposes that within a constructivist prompting framework, it can be fostered through principles such as meaningful prompting and metacognitive analysis. From a broader institutional perspective, Correia et al. (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) directly define teachers' evolving roles as \"prompt engineers\" and relate this identity to UNESCO's AI competency framework. According to the authors, effective prompt design enables teachers to maintain their professional autonomy as ethical decision makers and pedagogical process designers, underscoring the importance of pre-service teachers acquiring these competencies during their initial training.\u003c/p\u003e \u003cp\u003eIn the specific context of teacher education, other studies examined the use of prompt engineering strategies within the Intelligent TPACK framework (Celik et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2026\u003c/span\u003e), the mediating role of disciplinary literacy (Kang et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), and the relationship with pedagogical creative thinking (T\u0026uuml;men Akyıldız, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2026\u003c/span\u003e). While these studies reveal the relationship between prompt engineering and pedagogical practices, they have predominantly been framed around general AI self-efficacy or perceptions toward technology, rather than directly focusing on a distinct, task- and context-specific self-efficacy construct unique to prompt engineering in pedagogical settings. As a learnable and developable capacity, this construct is also directly measurable, which in turn necessitates valid and reliable instruments for its systematic assessment.\u003c/p\u003e \u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003ch2\u003e1.1.Need for Educational Prompt Engineering Self-Efficacy Scale (Ed-PESS)\u003c/h2\u003e \u003cp\u003eThe growth of generative AI in educational contexts has led to the development of various measurement instruments targeting related constructs. The Artificial Intelligence Literacy Scale (AILS), comprising 12 items across four dimensions (awareness, usage, evaluation, and ethics) provides a validated framework for measuring general AI literacy (Wang et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Similarly, the AI Self-Efficacy Scale developed by Wang and Chuang (\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) assesses individuals' perceived confidence in using generative AI through a 22-item instrument with four dimensions: anthropomorphism, assistance, technological skill, and comfort with AI. In teacher education contexts specifically, Chiu et al. (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) introduced the Teacher Artificial Intelligence Competence Self-Efficacy Scale (TAICS), a 24-item instrument measuring six dimensions including AI assessments, AI knowledge, AI ethics, AI pedagogy, professional engagement, and human-centered education. The Intelligent-TPACK framework (Celik, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) extends the traditional TPACK model to address AI-specific pedagogical knowledge, encompassing Intelligent-TK, Intelligent-TCK, Intelligent-TPK, and ethical dimensions. Additionally, digital competence frameworks such as the Students' Digital Competence Scale (SDiCoS; Tzafilkou et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and the Digital Competency Scale for Teachers (G\u0026uuml;m\u0026uuml;ş \u0026amp; Kukul, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) have emerged to assess technology integration capabilities. Despite these contributions, significant limitations constrain their applicability to measuring prompt engineering or prompt engineering self-efficacy among pre-service teachers. First, most instruments were developed before the emergence of generative AI and prompt engineering as distinct constructs, focusing instead on general technology acceptance and use (Laupichler et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Second, existing measures emphasize general AI interaction capabilities rather than the specific belief structures underlying confidence in prompt design for pedagogical purposes. Third, none of these instruments capture the iterative, dialogic nature of prompt engineering confidence, which requires educators to engage in cycles of prompt formulation, output evaluation, and refinement to achieve instructionally appropriate responses (Celik et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2026\u003c/span\u003e; T\u0026uuml;men Akyıldız, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2026\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eBesides these studies, most recently and most closely related to the present study, Gibreel and Arpaci (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) developed the Prompt Engineering Competence Scale (PECS), a unidimensional nine-item instrument measuring users' general proficiency in prompt engineering. The PECS represents a significant contribution by establishing prompt engineering as a measurable construct. However, critical distinctions limit its applicability to education. First, the PECS measures competence rather than self-efficacy, yet self-efficacy beliefs are more predictive of behavioral engagement than objective skill levels (Bandura, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Second, the instrument was developed with general AI users, lacking pedagogical specificity essential for understanding how teachers design prompts to achieve instructional objectives. Consequently, a teacher or teacher candidate may demonstrate high PECS scores while lacking confidence in applying prompt engineering to educational purposes. Additionally, the PECS, while addressing prompt engineering competence, measures actual performance rather than self-efficacy beliefs and lacks pedagogical specificity relevant to teacher preparation contexts. These limitations reveal a clear methodological gap, specifically the absence of a psychometrically validated instrument designed to measure educational prompt engineering self-efficacy for pre-service teachers. The present study addresses this gap through the development and validation of the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS). While educational policy increasingly requires AI integration in teacher preparation (Kang et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; T\u0026uuml;men Akyıldız, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2026\u003c/span\u003e), no valid instrument exists to determine whether teacher candidates believe they can effectively engineer prompts to bring out pedagogically appropriate AI responses. The consequences of this problem are far-reaching, as teacher training curricula cannot systematically identify students with low prompt engineering self-efficacy, evaluate the impact of instructional interventions, or detect prompt-related anxiety that may slow down effective AI utilization.\u003c/p\u003e \u003cp\u003eSelf-efficacy theory (Bandura, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2006\u003c/span\u003e) provides the theoretical foundation for addressing this gap. Self-efficacy, defined by Bandura as belief in one's ability to achieve desired goals, encompasses competence and confidence in successfully performing tasks. Building on the research gap highlighted in the literature, the purpose of this study is to develop the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS) and to provide evidence of its validity and reliability. To ensure theoretical validity, the instrument\u0026rsquo;s development was grounded in a systematic literature review and thematic analysis, with every item derived from and supported by direct evidence from the literature. Accordingly, the study aimed to obtain evidence regarding the construct validity and reliability of the scale within the framework of current psychometric standards. To this end, the following research questions were addressed:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eWhat is the factor structure of the Ed-PESS based on exploratory factor analysis?\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eTo what extent does confirmatory factor analysis support the factor structure of the Ed-PESS?\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eWhat is the evidence for the reliability and convergent validity of the Ed-PESS and its subscales?\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Research Design\u003c/h2\u003e \u003cp\u003eThis study employed an instrument development and validation design to develop the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS). The scale development process was guided by Boateng et al.\u0026rsquo;s (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) three-phase, nine-step framework: item development (domain identification and item generation; content validity), scale development (question pre-testing, survey administration and sampling, item reduction, and latent factor extraction), and scale evaluation (tests of dimensionality, reliability, and validity). In Phase 1, the construct domain was defined and an initial item pool was created based on a systematic review of the prompt engineering and AI in teacher education literature. Expert review and pilot testing with the target group were then used to improve item relevance and clarity. In Phase 2, the draft scale was administered to Sample 1 of pre-service teachers, where item reduction procedures and exploratory factor analysis were used to identify the latent structure. In Phase 3, the retained items were tested in an independent Sample 2 (n\u0026thinsp;=\u0026thinsp;305) using confirmatory factor analysis. Because the data were not normally distributed, maximum likelihood estimation with a Yuan and Bentler robust correction was used. Reliability was examined using Cronbach\u0026rsquo;s alpha and McDonald\u0026rsquo;s omega, alongside internal structure validity indices (e.g., factor loadings and AVE).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Participants\u003c/h2\u003e \u003cp\u003eGiven the inherent requirements of the scale development process, this study was conducted with two independent participant groups. In the first phase, data were collected from 327 pre-service teachers to examine the construct validity of the 32-item draft scale through EFA. Following univariate (z-score, \u0026plusmn;\u0026thinsp;3) and multivariate (Mahalanobis distance) outlier screening, the final EFA sample comprised n₁ = 300 participants. For the CFA phase, a separate sample of 341 pre-service teachers was recruited independently; after removing five univariate and 31 multivariate outliers, the final CFA sample comprised n₂ = 305 participants. A participant-to-item ratio exceeding 10:1 in both groups provides evidence for sampling adequacy. Detailed demographic characteristics of both groups are presented in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cem\u003eDemographic Characteristics of Study Group 1 (EFA) and Study Group 2 (CFA)\u003c/em\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCategory\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eStudy Group 1 (EFA, n₁ = 300)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eStudy Group 2 (CFA, n₂ = 305)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eGender\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e219 (73.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e224 (73.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e81 (27.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e81 (26.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAge\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e17\u0026ndash;18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e66 (21.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e56 (18.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e74 (24.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e79 (26.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e22 and above\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRange\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e18\u0026ndash;40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e17\u0026ndash;38\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eYear of Study\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1st Year\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e101 (33.1%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2nd Year\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e83 (27.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e127 (41.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3rd Year\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e113 (37.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4th Year\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAcademic Program (Top Programs)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEnglish Language Teaching\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e24.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e9.2%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGerman Language Teaching\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e16.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSocial Studies Teaching\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e9.5%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePrimary AI Usage Purpose\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHomework \u0026amp; Presentation Prep\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e69 (23.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e31.1%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAI Usage Duration\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u0026ndash;2 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e36.4%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e30.5%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAI Training Status\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eReceived prior AI training\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e37 (12.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e20 (6.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo training; willing to receive\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e181 (60.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e140 (45.9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo training; not interested\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e79 (26.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e141 (46.2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e\u003cb\u003eNote.\u003c/b\u003e \u003cem\u003eEFA\u0026thinsp;=\u0026thinsp;Exploratory Factor Analysis; CFA\u0026thinsp;=\u0026thinsp;Confirmatory Factor Analysis; HBV\u0026thinsp;=\u0026thinsp;Hacı Bektaş Veli. Percentages may not sum to 100 due to rounding. Dashes (\u0026mdash;) indicate data not separately reported for that group.\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eAs seen in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, several patterns in the demographic profiles of both groups warrant methodological consideration. The pronounced gender imbalance, with female participants comprising approximately 73% of both samples, reflects the broader feminization of teacher education in Turkey rather than a sampling artifact. Both groups were predominantly composed of first- and second-year students with limited AI training experience; fewer than 13% in Group 1 and fewer than 7% in Group 2 had received any formal AI instruction. This near-absence of structured AI training, combined with the fact that nearly half of Group 2 reported no training and no intention to pursue it, suggests that the scale was validated under ecologically representative conditions for the current state of AI integration in Turkish pre-service teacher education. At the same time, it raises the question of whether self-efficacy scores may be systematically attenuated by limited prior exposure, a potential confound that longitudinal or intervention-based designs should address.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Instrument Development and Procedure\u003c/h2\u003e \u003cp\u003eData were collected via an online survey administered to pre-service teachers between May and November 2025; participation was voluntary, anonymous, and preceded by informed consent, with no incentives offered. Following a systematic literature review and researcher discussions, an initial pool of 38 items was drafted in accordance with 5 Likert-type scale development conventions, with ambiguous expressions deliberately avoided. Items were reviewed by three Measurement and Evaluation specialists, two Computer Education and Instructional Technology specialists, and two Turkish Language specialists for language, content, and psychometric quality; applying Lawshe's (1975) CVR criterion of .80 at p \u0026lt; .05, six items falling below this threshold were removed, and the remaining items were revised to yield a 32-item draft instrument comprising four subscales: Basic Prompt Engineering (9 items), Pedagogical Prompt Engineering (7 items), Ethical Prompting Practice (9 items), and Continuous Professional Development (7 items). The study was conducted with two independent samples: responses collected from 327 pre-service teachers were screened for univariate and multivariate outliers, yielding a final EFA sample of n₁ = 300 (retention rate\u0026thinsp;\u0026asymp;\u0026thinsp;91.7%); a separate sample of 341 participants was subsequently recruited for CFA, with analogous screening procedures producing a final sample of n₂ = 305 (retention rate\u0026thinsp;\u0026asymp;\u0026thinsp;89.4%).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.4. Data Analysis\u003c/h2\u003e \u003cp\u003eData analysis was carried out in three stages: data screening, construct validity analyses, and reliability analyses. Prior to conducting the analyses, the normality assumption of the dataset was evaluated using Mardia's coefficients. Additionally, z-scores (\u0026plusmn;\u0026thinsp;3 threshold) were employed for univariate outlier detection, and Mahalanobis distance values were used for multivariate outlier identification. To provide evidence of construct validity and to determine the factor structure of the scale, EFA was performed as the initial analytical step. The suitability of the dataset for factorization was assessed using the Kaiser-Meyer-Olkin (KMO) coefficient and Bartlett's Test of Sphericity. The KMO value was found to be .942, indicating excellent sampling adequacy, and Bartlett's test results were statistically significant, χ\u0026sup2;(325)\u0026thinsp;=\u0026thinsp;6173, p \u0026lt; .001. Principal Axis Factoring was selected as the extraction method, and given the presence of inter-factor correlations (r = .507 to .645), Promax rotation, an oblique rotation method, was employed. The number of factors was determined by examining eigenvalues greater than 1 and inspecting the scree plot. CFA was subsequently applied to confirm the structure identified through EFA. Because the data did not conform to a multivariate normal distribution, the Maximum Likelihood (ML) estimation method was used in conjunction with Yuan-Bentler (robust) correction. Model fit was evaluated using the following indices: Robust χ\u0026sup2;/df, RMSEA (Root Mean Square Error of Approximation), SRMR (Standardized Root Mean Square Residual), CFI (Comparative Fit Index), and TLI (Tucker-Lewis Index). To provide evidence of reliability, Cronbach's alpha (α), McDonald's omega (ω), and the composite reliability (CR) coefficient were computed to assess the internal consistency of both the total scale and its subscales. Average Variance Extracted (AVE) values were examined as evidence of convergent validity. All analyses were performed using the JAMOVI statistical software.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.5. Measures: Item Generation and Development\u003c/h2\u003e \u003cp\u003eThe item development process followed a theoretically rigorous approach supported by a systematic literature review to identify the construct's dimensional structure. Subsequently, thematic analysis of core papers established the framework for item generation, ensuring that each item was based on specific empirical or theoretical evidence. This literature-grounded methodology ensures initial content validity. Following Boateng et al.'s (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) scale development framework, the systematic search was conducted in \u0026ldquo;Web of Science\u0026rdquo; and \u0026ldquo;Scopus\u0026rdquo; databases to establish the theoretical foundation for item development. The review identified core papers addressing prompt engineering in educational contexts, from which principal themes were extracted through thematic analysis. Based on this synthesis, educational prompt engineering self-efficacy was operationally defined as a multidimensional construct comprising pre-service teachers' confidence in designing, implementing, and evaluating prompts for generative AI tools in pedagogically appropriate ways. This conceptualization integrates Bandura's (2006) self-efficacy theory with domain-specific competencies identified in the literature.\u003c/p\u003e \u003cp\u003eThe systematic literature review revealed four recurring themes that formed the theoretical architecture for item development: (1) technical prompt construction skills, (2) pedagogical integration of AI outputs, (3) ethical considerations in educational AI use, and (4) ongoing professional learning for AI integration. These themes were marked as candidate dimensions, with each theme's basic elements serving as the framework for item generation. Items were written as first-person self-efficacy statements (e.g., \"I can...\") using clear language appropriate for pre-service teachers, following established guidelines (Boateng et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Each item was supported by direct citations from the literature to ensure theoretical validity (see Appendix 1 for all items and literature evidence). An initial pool of 32 items was generated, distributed across the four thematically-derived dimensions: Basic Prompt Engineering (9 items), Pedagogical Prompt Engineering (7 items), Ethical Prompting Practice (9 items), and Continuous Professional Development (7 items). In addition to the Appendix, Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e provides the first two sample items and literature support for each dimension.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cem\u003eSample Items and Literature Support for Each Dimension\u003c/em\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDimension\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eScale Items\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDirect Quote (Evidence)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDimension 1: Basic Prompt Engineering (9 items),\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1. I can explain how AI tools generate content.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1. \"AI systems generate text (outputs) by making statistically informed predictions based on the patterns they have learned and responding to prompts (inputs) entered by users (Park \u0026amp; Choo, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\"\u003c/p\u003e \u003cp\u003e2. \"The session included demonstrations of interactions with ChatGPT-4, showcasing its text generation and text-to-image functionalities. Particular emphasis was placed on the importance of crafting effective prompts, as well as demonstrating how refining prompts could elicit more accurate and contextually appropriate responses from the AI (Biberman-Shalev, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\"\u003c/p\u003e \u003cp\u003e3. \"ChatGPT works by learning from a large amount of text data, which helps it understand grammar, vocabulary, and the meaning of words in different contexts. This learning process of ChatGPT is analogous to training ChatGPT's brain to understand language (Lee et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\"\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2. I can select the AI tool (e.g., ChatGPT, Claude, Gemini) that best fits my needs.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1. \"When prompt engineering, you will start by choosing a model. Prompts might need to be optimized for your specific model, regardless of whether you use Gemini language models in Vertex AI, GPT, Claude, or an open source model like Gemma or LLaMA (Boonstra, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\"\u003c/p\u003e \u003cp\u003e2. \"The most widely used large language models (LLMs) today include ChatGPT, Gemini, Claude, Mistral, and Llama, with ChatGPT remaining the most dominant overall (Mutanga et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\"\u003c/p\u003e \u003cp\u003e3. \"I preferred ChaGPT4.0. It is expensive, but it gives you value for money. I used my bursary and subscribed to this tool (Van Wyk, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\"\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDimension 2: Pedagogical Prompt Engineering (7 items),\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1. I can design prompts that generate content aligned with my learning objectives.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1. \"Goal-oriented and task-specific frameworks focus on designing prompts to fulfill clearly defined instructional or functional objectives. These frameworks align prompt structures with desired outcomes, such as eliciting student reasoning or performing a diagnostic task (Qian, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\"\u003c/p\u003e \u003cp\u003e2. \"For teachers to use GenAI tools effectively, they must craft prompts that encourage the AI to engage substantively with instructional goals and deliver responses aligned with intended learning outcomes (Celik et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2026\u003c/span\u003e).\"\u003c/p\u003e \u003cp\u003e3. \"Continuous refinement through iterative prompt writing ensures alignment with learning objectives and facilitates real-time lesson adaptation (Carl et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\"\u003c/p\u003e \u003cp\u003e4. \"Each instruction must be 'initiated with a clear and precise directive' and maintain alignment with the learning objectives and educational goals (Lee \u0026amp; Palmer, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\"\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2. I can develop prompts that generate differentiated content suitable for students' needs.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1. \"Adaptive learning is an instructional approach in which learning pathways are adjusted to align with learners' needs, prior knowledge, and interests... instruction is tailored to students' progress and characteristics (Celik et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2026\u003c/span\u003e).\"\u003c/p\u003e \u003cp\u003e2. \"Crucially, the task emphasized designing adaptive or personalized instructional strategies by encouraging pre-service teachers to request support from the AI that considered pupil diversity, engagement, and differentiated learning needs (Celik et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2026\u003c/span\u003e).\"\u003c/p\u003e \u003cp\u003e3. \"Real-time AI assistance in classrooms enhances lesson delivery and enables personalized learning for diverse student needs (Carl et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\"\u003c/p\u003e \u003cp\u003e4. \"Prompts emphasizing contextual clarity aid differentiation by adjusting outputs to learner needs (Qian, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\"\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDimension 3: Ethical Prompting Practice (9 items),\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1. I can develop classroom rules for prompt creation (e.g., 'do not use your name') to ensure privacy and data security.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1. \"Educate students about data privacy, including how their data is used by AI systems and ways to protect their digital footprint (Walter, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\" \u003c/p\u003e \u003cp\u003e2. \"One practical approach is to develop a personal checklist for appropriate and ethical uses of AI (Park \u0026amp; Choo, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\" \u003c/p\u003e \u003cp\u003e3. \"It should be clear how the expectations of the school look like so that students know exactly what they are allowed and what they are not allowed to do (Walter, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\"\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2. I can protect the confidentiality of student data while creating prompts.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1. \"Providers of LLM-based GenAI tools, such as OpenAI, use prompting data to train their AI models, which may raise concerns regarding data privacy and protection... To mitigate this, learners should avoid providing personal or identifiable information when interacting with LLM-based GenAI (Hsu, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\" \u003c/p\u003e \u003cp\u003e2. \"AI systems frequently handle private user data without following open procedures for its storage or use... many AI applications gather personal data without express consent (T\u0026uuml;men Akyıldız, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2026\u003c/span\u003e).\" \u003c/p\u003e \u003cp\u003e3. \"Ethical AI development is essential, focusing on transparency, unbiased content, and privacy-respecting practices (Walter, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\"\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDimension 4: Continuous Professional Development (7 items).\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1. I can follow current developments in AI tools and prompt engineering.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1. \"Regular training and workshops for educators will ensure they stay updated with the latest AI technology advancements.\" (Walter, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) \u003c/p\u003e \u003cp\u003e2. \"Provide workshops for career guidance that emphasize adaptability and the importance of continuous learning in an AI-evolving job landscape. Teach an agile mindset and provide sources to learn the newest developments.\" (Walter, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) \u003c/p\u003e \u003cp\u003e3. \"The rapid evolution of AI technologies can render existing frameworks quickly outdated. As new AI capabilities evolve, educators may find themselves needing to continuously adapt their teaching strategies, frameworks and assessment processes, or risk passively developing potential gaps in knowledge and skills.\" (Lee \u0026amp; Palmer, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2025\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2. I strive to use current developments in AI/prompt engineering in my teaching practices.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1. \"Several teachers indicated a desire to investigate AI tools with specialized functionalities, highlighting the necessity of continuously adapting AI technologies to address the changing demands of education.\" (ElSayary et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) \u003c/p\u003e \u003cp\u003e2. \"Integrating prompt engineering into teacher education curricula may therefore serve as a bridge between digital literacy and pedagogical creativity, preparing future teachers to use AI tools reflectively and effectively.\" (T\u0026uuml;men Akyıldız, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2026\u003c/span\u003e) \u003c/p\u003e \u003cp\u003e3. \"The deliberate goal is to eventually lead students towards a responsible use of AI, but to do so, they need to understand how one can 'talk' to an AI so that it does what it is supposed to.\" (Walter, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eBasic Prompt Engineering (BPE).\u003c/b\u003e This dimension represents the technical capacity to structure clear, context-aware instructions using specific prompt components and to iteratively refine inputs. The theoretical grounding derives from the CLEAR framework (Lo, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), the PARTS framework (Park \u0026amp; Choo, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and empirical work on prompt literacy (Hwang et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Seven items capture competencies including prompt construction, component specification, role assignment, and systematic documentation.\u003c/p\u003e \u003cp\u003e \u003cb\u003ePedagogical Prompt Engineering (PPE).\u003c/b\u003e This dimension represents the capacity to design prompts aligned with learning objectives, differentiate instruction, and critically evaluate AI outputs. It is grounded in the TPACK framework (Mishra \u0026amp; Koehler, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2006\u003c/span\u003e) and its AI-specific extension, Intelligent-TPACK (Celik, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The items address learning objective alignment, differentiated content, critical evaluation, and engagement strategies.\u003c/p\u003e \u003cp\u003e \u003cb\u003eEthical Prompting Practice (EPP).\u003c/b\u003e This dimension represents the capacity to protect student data privacy, mitigate algorithmic bias, and model ethical AI use. Walter (\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) positions ethical discussions as integral to AI-integrated learning, while Hsu (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) documents data privacy concerns in GenAI interactions. Mzwri and Turcs\u0026aacute;nyi-Szab\u0026oacute; (\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) articulate that prompt engineering guides outputs toward fairness and inclusivity. Nine items address privacy protection, bias mitigation, source verification, and ethical modeling for students.\u003c/p\u003e \u003cp\u003e \u003cb\u003eContinuous Professional Development (CPD).\u003c/b\u003e This dimension represents the disposition to monitor AI advancements and seek ongoing skill improvement. Foundationally grounded in Sch\u0026ouml;n's reflective practice framework and supported by ElSayary et al. (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), this dimension recognizes that rapid AI evolution requires continuous updating (Lee \u0026amp; Palmer, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Walter, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Woo et al. (\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2025\u003c/span\u003eb) demonstrated that targeted workshops significantly improve prompt engineering ability. The items address awareness, adaptation, feedback-driven refinement, and training participation.\u003c/p\u003e \u003c/div\u003e"},{"header":"3.Results","content":"\u003cp\u003eAs the initial step in the scale development process, the multivariate normality assumption of the dataset was tested using Mardia's coefficients. The statistically significant skewness (117) and kurtosis (888) values (p \u0026lt; .001) indicated that the data did not conform to a multivariate normal distribution. Accordingly, robust estimation methods were employed throughout the analyses.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe number of dimensions was determined by triangulating three criteria: Horn's Parallel Analysis (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), the Scree Plot (see Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), and the Kaiser criterion (eigenvalue\u0026thinsp;\u0026gt;\u0026thinsp;1). Convergent evaluation of these results supported a four-factor structure for the scale. The EFA findings for the four-factor prompt engineering scale are presented in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cem\u003eExploratory Factor Analysis Results of the Prompt Engineering Scale\u003c/em\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eItems\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e \u003cp\u003eDimension Loadings\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c8\" namest=\"c6\"\u003e \u003cp\u003eFactor Statistic\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFactor 1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eFactor 2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eFactor 3\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eFactor 4\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSS Loading\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e% of Variance\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eCumulative %\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eEthical Prompting Practice\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e5.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e22.20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e22.20\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e.70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eContinuous Professional Development\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e15.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e37.50\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.79\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eBasic Prompt Engineering\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e15.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e52.50\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePedagogical Prompt Engineering\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e12.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e64.60\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Taba\" border=\"1\"\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInter-Factor Correlations\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1. Basic Prompt Engineering\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2. Continuous Professional Development\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3. Ethical Prompting Practice\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4. Pedagogical Prompt Engineering\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.65\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e\u003cem\u003eNote. N\u0026thinsp;=\u0026thinsp;300. KMO = .942; Bartlett's Test of Sphericity χ\u0026sup2;(325)\u0026thinsp;=\u0026thinsp;6173, p \u0026lt; .001. Model fit measures: RMSEA = .076 [.069, .084], TLI = .902, χ\u0026sup2;(227)\u0026thinsp;=\u0026thinsp;624, p \u0026lt; .001. Promax oblique rotation and principal axis factoring were used as the factor extraction and rotation methods, respectively. Factor loadings above .40 are bolded.\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe KMO value of .942 and Bartlett's Test of Sphericity, χ\u0026sup2;(325)\u0026thinsp;=\u0026thinsp;6173, p \u0026lt; .001, confirmed that the data were suitable for factor analysis. EFA conducted using principal axis factoring and Promax rotation yielded a four-factor structure in which all retained factors had eigenvalues exceeding 1. These four factors collectively accounted for 64.60% of the total variance. The individual contributions of each factor to the total variance were 22.20% (Basic Prompt Engineering), 15.30% (Continuous Professional Development), 15.00% (Ethical Prompting Practice), and 12.10% (Pedagogical Prompt Engineering), respectively. Item factor loadings ranged from .46 to .97, and inter-factor correlation coefficients ranged from .51 to .65. Examination of the model fit indices, RMSEA = .076 [.069, .084], TLI = .902, and χ\u0026sup2;/df\u0026thinsp;=\u0026thinsp;2.75, indicates that the model demonstrated an acceptable level of fit to the data (Kline, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Tabachnick \u0026amp; Fidell, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cem\u003eConfirmatory Factor Analysis Results\u003c/em\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"11\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cb\u003eBasic Prompt Engineering\u003c/b\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003eB\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003eSE\u003c/em\u003e(\u003cem\u003eB\u003c/em\u003e)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003ez\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003ep\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cem\u003eβ\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003cem\u003eδ\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cem\u003eα\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eAVE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u003cem\u003eω\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c11\"\u003e \u003cp\u003eCR\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cb\u003e.865\u003c/b\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003e\u003cb\u003e.490\u003c/b\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u003cb\u003e.871\u003c/b\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c11\"\u003e \u003cp\u003e\u003cb\u003e.870\u003c/b\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e.590\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.652\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.149\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e8.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e.775\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.160\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9.27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e.832\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.309\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.159\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e8.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e.756\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.428\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.150\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e8.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e.756\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.429\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.143\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e.575\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.669\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.150\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e.585\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.658\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePedagogical Prompt Engineering\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cb\u003e.880\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e\u003cb\u003e.605\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u003cb\u003e.886\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e\u003cb\u003e.882\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e.766\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.413\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.061\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e18.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e.827\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.317\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.96\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.067\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e14.40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e.779\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.392\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.066\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e13.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e.754\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.432\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.068\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e12.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e.744\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.446\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eEthical Prompting Practice\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cb\u003e.938\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e\u003cb\u003e.659\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u003cb\u003e.939\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e\u003cb\u003e.939\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.674\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.545\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.179\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.075\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e15.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.786\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.383\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.271\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.090\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e14.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.856\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.267\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.319\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.093\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e14.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.882\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.221\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.311\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.098\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e13.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.856\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.268\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.182\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.106\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e11.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.765\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.414\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.199\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.089\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e13.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.847\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.282\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.155\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.091\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e12.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.811\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.342\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eContinuous Professional Development\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cb\u003e.887\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e\u003cb\u003e.566\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u003cb\u003e.886\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e\u003cb\u003e.888\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.733\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.463\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.048\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.058\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e18.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.809\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.346\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.923\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.068\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e13.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.771\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.406\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.071\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e14.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.758\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.425\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.965\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.076\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e12.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.676\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.544\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eitem_32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.033\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.073\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e14.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.779\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.393\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"11\"\u003e\u003cem\u003eNote. B\u003c/em\u003e: Unstandardized factor loading; \u003cem\u003eSE\u003c/em\u003e(\u003cem\u003eB\u003c/em\u003e): Standard error of the unstandardized factor loading; \u003cem\u003eβ\u003c/em\u003e: Standardized factor loading; \u003cem\u003eδ\u003c/em\u003e: Standardized error variance; AVE: Average Variance Extracted; \u003cem\u003eα\u003c/em\u003e: Internal consistency coefficient; \u003cem\u003eω\u003c/em\u003e: McDonald's omega; CR: Composite Reliability.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWhether the 26-item, four-factor structure identified through EFA (Basic Prompt Engineering, Pedagogical Prompt Engineering, Ethical Prompting Practice, and Continuous Professional Development) was confirmed was tested via CFA in the second study group comprising 305 participants. Examination of the robust fit indices derived from Robust Maximum Likelihood (RML) estimation yielded the following values: χ\u0026sup2;/df\u0026thinsp;=\u0026thinsp;2.03 (595/293), RMSEA = .064 [95% CI: .057, .071], SRMR = .057, CFI = .929, and TLI = .921, collectively indicating that the model demonstrated good fit to the data (Brown, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Kline, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Standardized factor loadings (β) ranged from .575 to .882, all statistically significant at p \u0026lt; .001.\u003c/p\u003e \u003cp\u003eExamination of the reliability analysis results revealed that Cronbach's α, McDonald's ω, and CR values exceeded .80 across all subscales, reaching as high as .939 for the Ethical Prompting Practice subscale, indicating that the scale's internal consistency is very high. AVE values exceeding .50 demonstrate that the respective latent variable (factor) accounts for more than half of the variance in its associated items (Hair et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Composite reliability (CR) coefficients ranged from .87 to .94, and CR values exceeding .70 across all factors confirm that the scale possesses a high level of internal consistency (Fornell \u0026amp; Larcker, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e1981\u003c/span\u003e; Hair et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Although the AVE value for the Basic Prompt Engineering subscale (.490) fell marginally below the .50 threshold, the high reliability coefficients for the same subscale (α\u0026thinsp;=\u0026thinsp;.865, ω\u0026thinsp;=\u0026thinsp;.871) and a CR value exceeding .70 indicate that the reliability of this subscale remains within acceptable limits (Fornell \u0026amp; Larcker, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e1981\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe research findings collectively support the conclusion that the developed instrument possesses a valid and reliable structure. A KMO value exceeding .90 indicates that the sample size was at an \"excellent\" level for factorization. Item factor loadings ranging from .46 to .97, combined with the total variance explained by the scale exceeding the 60% threshold at 64.60%, provide compelling evidence that the factor structure effectively represents the measured construct. Turning to the CFA results, CFI and TLI values exceeding .90, alongside RMSEA and SRMR values falling below .08, confirm that the four-factor model is statistically congruent with the theoretical structure. Inter-factor correlations ranging from .51 to .65 indicate that the subscales are interrelated yet measure conceptually distinct constructs. Examination of Cronbach's α, McDonald's ω, and CR values across all subscales further confirms that the scale is reliable. In conclusion, the 26-item, four-subscale 5-point Likert prompt engineering scale, comprising Basic Prompt Engineering, Pedagogical Prompt Engineering, Ethical Prompting Practice, and Continuous Professional Development, can be considered a valid and reliable instrument for measuring the relevant competencies of pre-service teachers.\u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThis study addressed a measurement gap in AI-integrated teacher education by developing and validating the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS). The study was guided by three research questions addressing (1) the factor structure of the Ed-PESS based on EFA, (2) the extent to which CFA supports that structure, and (3) the reliability and convergent validity evidence for the scale and its subscales. Overall, the results provide initial validity evidence supporting the interpretation of Ed-PESS scores as a multidimensional measure of pre-service teachers' prompt engineering self-efficacy.\u003c/p\u003e \u003cp\u003eTo address non-normality, the CFA of the Ed-PESS was estimated using Maximum Likelihood with the Yuan and Bentler robust correction. In an independent validation sample (n\u0026thinsp;=\u0026thinsp;305), model fit was acceptable (χ\u0026sup2;/df\u0026thinsp;=\u0026thinsp;2.03, RMSEA = .064 [95% CI: .057 to .071], SRMR = .057, CFI = .929, TLI = .921), providing internal structure evidence for the proposed four-factor solution.\u003c/p\u003e \u003cp\u003eConverging EFA and CFA results supported a stable four-factor structure of the Ed-PESS: Basic Prompt Engineering, Pedagogical Prompt Engineering, Ethical Prompting Practice, and Continuous Professional Development. The EFA solution explained 64.60% of the total variance. In the CFA, standardized loadings ranged from .575 to .882 (all p \u0026lt; .001), indicating coherent relations between items and factors. Convergent validity was generally adequate (AVE: Pedagogical = .605; Ethical = .659; CPD = .566), while the Basic dimension was borderline (AVE = .490), suggesting broader or more heterogeneous operational content despite strong reliability. Internal consistency was high across dimensions (α and ω\u0026thinsp;\u0026gt;\u0026thinsp;.80; Ethical \u0026asymp; .94). Inter-factor correlations were moderate (r = .507-.645), consistent with related yet distinguishable facets within a broader self-efficacy construct.\u003c/p\u003e \u003cp\u003eContent-focused validity evidence for the Ed-PESS was supported through a systematic review of the prompt engineering literature, expert review of item relevance and clarity (measurement, educational technology, and language experts), and pilot testing with target-population feedback followed by item refinement.\u003c/p\u003e \u003cp\u003eConsistent with recent studies, the results support treating prompt related self-efficacy as a construct that can be strengthened through targeted instruction and operationalized for assessment in educational settings (e.g., Davila-Moran et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Jang et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). At the same time, the four-factor pattern suggests that prompt engineering self-efficacy in teacher education may include related but distinct parts. This goes beyond the single factor structure reported in a competence scale developed with general users (Gibreel \u0026amp; Arpaci, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Interpreting the Hierarchy of Ed-PESS\u003c/h2\u003e \u003cp\u003e \u003cb\u003eBasic Prompt Engineering.\u003c/b\u003e The first dimension, Basic Prompt Engineering, accounted for the largest proportion of variance in the EFA (22.2%), indicating that operational prompt-structuring is a core component of the Ed-PESS. This dimension showed strong internal consistency (α\u0026thinsp;=\u0026thinsp;.865; ω was approximately .868 to .871), while convergent validity was borderline (AVE = .490). This result suggests slightly limited convergent validity and may indicate that the factor covers a broader range of operational self-efficacy beliefs rather than a narrowly uniform domain. Its moderate correlation with Pedagogical Prompt Engineering (r = .591) suggests that the two dimensions are related but not the same. Confidence in structuring prompts is linked to confidence in instructional integration, but they represent different aspects of self-efficacy.\u003c/p\u003e \u003cp\u003eTheoretically, this operational dimension fits Cognitive Load Theory (CLT). When the \u0026ldquo;syntax\u0026rdquo; of prompting, such as stating roles, constraints, context, and output formats, uses a lot of working memory, it can create extraneous load. This extra load can reduce the mental resources needed for pedagogical reasoning and for checking the quality of the output (Sweller, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e1988\u003c/span\u003e). Therefore, self-efficacy in Basic Prompt Engineering can be seen as confidence in handling these basic parts smoothly and with little unnecessary mental effort, so attention can shift to instructional planning. This CLT-based view also matches Mendes et al. (\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), who highlight the importance of reducing extraneous load and increasing mental effort that supports learning in LLM-supported learning, especially at early stages. Finally, the basic elements captured here align with component-based prompt descriptions such as the Input, Instruction, Output, and Context (IIOC) framing reported by Jin et al. (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). This supports the idea that Basic Prompt Engineering is a starting structure that can help later pedagogical use, but it does not guarantee it.\u003c/p\u003e \u003cp\u003e \u003cb\u003ePedagogical Prompt Engineering.\u003c/b\u003e In the Ed-PESS, Basic and Pedagogical Prompt Engineering were moderately correlated (r = .591), and inter-factor correlations across the model ranged from r = .507 to r = .645, indicating a coherent yet multidimensional construct. Reliability was strong for both the Basic (α\u0026thinsp;=\u0026thinsp;.865; ω\u0026thinsp;\u0026asymp;\u0026thinsp;.868-.871) and Pedagogical (α\u0026thinsp;=\u0026thinsp;.880; ω\u0026thinsp;\u0026asymp;\u0026thinsp;.883-.886) dimensions, and convergent validity was adequate for Pedagogical Prompt Engineering (AVE = .605). Together, these results support interpreting pedagogical prompting as a distinct self-efficacy facet centred on instructional integration, not merely operational prompt construction.\u003c/p\u003e \u003cp\u003eTheoretically, this distinction fits Shulman\u0026rsquo;s (\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e1986\u003c/span\u003e) view of teaching as professional judgment. In this view, teachers combine procedures with content and clear reasons, while also responding to the demands of the situation. This includes paying attention to students\u0026rsquo; ideas and misconceptions and choosing explanations that make the topic easier to learn. This is consistent with the TPACK/TPCK argument that effective technology integration depends on understanding the dynamic relationships among technology, pedagogy, and content, because knowing how to use technology is not equivalent to knowing how to teach with it (Mishra \u0026amp; Koehler, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Celik\u0026rsquo;s (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) AI-TPACK extension similarly emphasizes that technological knowledge alone is insufficient and underscores the centrality of pedagogical knowledge for ethically and effectively integrating AI-based tools. Additionally, from an instrument-mediated activity perspective, Rabardel\u0026rsquo;s (\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2002\u003c/span\u003e) theory of instrumental genesis provides a complementary lens: GenAI functions as an instructional instrument only when coupled with teachers\u0026rsquo; utilization schemes adapted to the specificity of each situation. Accordingly, higher self-efficacy in Pedagogical Prompt Engineering reflects confidence in using GenAI in a pedagogically purposeful way. This includes interpreting instructional demands, turning outputs into usable learning resources, and monitoring and steering the process in context (Rabardel, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Taken together, these findings suggest that teacher education should move beyond \u0026ldquo;chatting with AI\u0026rdquo; toward structured, case-based design experiences (e.g., prompt labs, microteaching with constraints, and iterative revision tasks) that cultivate instructional alignment and reflective judgment.\u003c/p\u003e \u003cp\u003e \u003cb\u003eEthical Prompting Practice.\u003c/b\u003e In the Ed-PESS, Ethical Prompting Practice formed a highly coherent subdomain of prompt engineering self-efficacy. Internal consistency was excellent (α\u0026thinsp;=\u0026thinsp;.938; ω\u0026thinsp;\u0026asymp;\u0026thinsp;.939) and convergent validity was strong (AVE = .659), indicating substantial shared variance between the items and the latent ethical self-efficacy factor. Inter-factor correlations (r = .507-.645) suggest that ethical self-efficacy is meaningfully connected to the other dimensions while remaining empirically distinguishable. Bandura\u0026rsquo;s (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2001\u003c/span\u003e) account of moral agency in Social Cognitive Theory provides a useful lens for interpreting Ethical Prompting Practice as ethically oriented self-efficacy: moral conduct is shaped by self-regulatory processes that must be actively engaged, particularly in contexts where responsibility can be diffused or displaced. In this view, ethical prompting self-efficacy refers to pre-service teachers\u0026rsquo; confidence that they can act ethically when using GenAI. This means setting clear safeguards and limits in their GenAI interactions and keeping human responsibility for decisions. In the present scale, this is reflected in item content addressing privacy/data protection, bias and age-appropriateness guardrails, reliability-oriented constraints, and modelling/teaching ethical prompting norms to students. Based on this view of moral agency, UNESCO explains ethical AI use in education in AI Competency Framework for Teachers. The framework focuses on decision processes that are \u0026ldquo;human controlled and human accountable.\u0026rdquo; It also clearly states that \u0026ldquo;Teachers should remain accountable for pedagogical decisions\u0026rdquo; when they use AI (UNESCO, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This policy view supports how we understand Ethical Prompting Practice in the Ed-PESS. It suggests that teachers should feel confident about protecting privacy and safety, reducing bias, and checking that outputs are reliable. They do this by writing prompts with clear rules and limits, instead of leaving responsibility to the tool. This idea is also supported by other studies. Agirdag (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) highlights that ethical prompting involves improving prompts over time and noticing bias during the dialogue with AI. This matters because teachers need to shape prompts so the outputs are safe, fair, and suitable for learning. Selwyn et al. (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) also show that teachers often check outputs carefully, fix problems, and sometimes rewrite the outputs. Overall, these studies support Ethical Prompting Practice as a practical responsibility in use, not only a technical prompting skill.\u003c/p\u003e \u003cp\u003e \u003cb\u003eContinuous Professional Development (CPD).\u003c/b\u003e In the Ed-PESS, CPD is a separate but connected part of prompt engineering self-efficacy. The CPD items showed strong internal consistency (α\u0026thinsp;=\u0026thinsp;.887; ω was approximately .884 to .886) and acceptable convergent validity (AVE = .566). The correlations between factors ranged from .507 to .645. This suggests that CPD is related to the other dimensions but still distinct. These results suggest that CPD is not just an attitude. Instead, it reflects self-efficacy for keeping prompt related practices up to date and improving them as GenAI tools and classroom expectations change. Artino (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2012\u003c/span\u003e) argues that self-efficacy beliefs matter in technology-mediated learning because learners try to control their own learning. Zimmerman\u0026rsquo;s (\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2002\u003c/span\u003e) self-regulation framework also supports this view. It describes ongoing cycles of planning, doing, and reflecting, including setting goals, planning strategies, monitoring progress, evaluating results, and adapting. In line with this, CPD in the Ed-PESS refers to perceived capability to follow new developments in AI and prompting, apply updates in teaching, and keep improving through reflection (for example, revising prompts based on feedback and setting professional learning goals). In addition, this view is supported by recent studies. T\u0026uuml;men Akyıldız (\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2026\u003c/span\u003e) describes prompting as a new practice that can feel experimental and uncertain. For this reason, competent GenAI use requires repeated updating and adaptation. This supports the idea that CPD reflects confidence in improving over time by trying, checking results, and making changes through repeated cycles of writing prompts, evaluating outputs, and revising prompts. Consistent with this view, ElSayary et al. (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) highlight that effective classroom use of GenAI requires AI literacy, continuous professional development, and repeated improvement of prompts and outputs.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Theoretical Implications: Prompting as \"Teacher Agency\"\u003c/h2\u003e \u003cp\u003eTaken together, the Ed-PESS supports viewing prompt engineering self-efficacy as teacher agency in GenAI-supported instruction. Teachers are not passive recipients of outputs. They actively regulate how GenAI is used and for what teaching goals. In an instrument mediated activity view, GenAI is an artifact. It becomes an instructional instrument only when it is linked to teachers\u0026rsquo; utilization schemes. Prompt engineering self-efficacy is the confidence to build, adapt, and coordinate these schemes to support learning goals (Rabardel, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). In addition, moral agency makes this view clearer. Ethical Prompting Practice in GenAI use is not guaranteed by abstract reasoning alone. It depends on teachers\u0026rsquo; confidence that they can activate self-regulatory safeguards and keep human accountability, especially in settings where responsibility can be spread across others or shifted to the tool (Bandura, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Most importantly, teacher agency in GenAI-supported teaching works like a repeat cycle. Teachers check if the output fits the teaching goal and if it causes ethical problems. If the output is weak, they find the reason and then change the prompt. This process matches three parts of self-efficacy in the Ed-PESS: Pedagogical Prompt Engineering, Continuous Professional Development, and Ethical Prompting Practice.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Practical Implications for Teacher Education\u003c/h2\u003e \u003cp\u003eGeneral lectures that introduce AI are not enough to build prompt engineering self-efficacy as teacher agency (Bandura, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2001\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Zimmerman, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Teacher education curricula should include Prompt Labs. In these sessions, pre-service teachers practise writing prompts for clear teaching goals. They also add safety and ethics rules and improve prompts and outputs step by step in realistic teaching tasks. This gives mastery experiences that can strengthen self-efficacy (Bandura, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). These labs can follow guided cycles of planning, doing, and reflecting. In this way, teacher candidates learn why an output did not work and how to change the prompt, instead of using GenAI as a tool that gives an answer in one attempt (Zimmerman, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2002\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo support evidence informed curriculum design, the Ed-PESS can be used as a pre-test and post-test in educational technology or methods courses. This can help instructors in two ways. First, it can show teacher candidates\u0026rsquo; self-efficacy profiles across the different dimensions. Second, it can show whether a specific teaching activity improves the targeted parts of prompt engineering self-efficacy. Teacher education should also include clear modules on ethical use. These modules should train teacher candidates to write prompts with safety and fairness rules. For example, they can ask for gender neutral roles, require child appropriate content, and request fair examples. The courses should also teach routines for checking outputs and revising them. In this way, ethical prompting practice remains a teachable and assessable part of professional practice, not something that is expected to happen automatically when using the tool.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Limitations and Future Directions\u003c/h2\u003e \u003cp\u003eAlthough this study provides strong evidence on the internal structure and reliability of the Ed-PESS, several limitations affect interpretation and point to next steps. First, the Ed-PESS is a self-report scale. It shows what candidates think they can do when they design, adapt, and safeguard GenAI prompts. It does not show what they do in real tasks. Future studies should combine Ed-PESS scores with performance measures. For example, researchers can use rubric scored tasks where candidates write prompts, check outputs, and revise them under pedagogical and ethical rules. This can test criterion validity and show how beliefs match observed prompting performance. Second, we validated the scale in teacher education programs in one national context. Studies with more diverse cohorts and institutions are needed to improve generalizability. We did not test measurement invariance. Future research should test invariance across key groups, such as gender, year level, prior GenAI training, and AI tool familiarity. It should also test invariance across cultures before using the scale for group comparisons. Third, prompting is shaped by a fast-changing technology context. Models, interfaces, and modes change, and model differences can affect prompting and how it is evaluated. For this reason, the Ed-PESS should be seen as measuring basic self-efficacy for pedagogically purposeful, ethically responsible, and developmentally appropriate prompting, not model specific tricks. Regular monitoring and planned item review are needed to protect content validity as new classroom needs and risks appear. Overall, these steps position the Ed-PESS as a practical base for research and formative evaluation that tracks how teacher self-efficacy develops as GenAI tools change.\u003c/p\u003e \u003c/div\u003e"},{"header":"5.Conclusion","content":"\u003cp\u003eThe main purpose of this study was to develop and validate the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS). This responds to the growing need for reliable measurement in AI integrated teacher education. Psychometric results supported a clear four factor structure. The EFA explained 64.60 percent of the total variance. The CFA, with a robust correction, showed good model fit (CFI = .929, RMSEA = .064, SRMR = .057). Reliability evidence was also strong across the scale. These results suggest that the Ed-PESS can support teacher education in a practical way. It can help prepare future teachers to stay human in the loop and guide GenAI use toward pedagogically sound and ethically responsible goals. This study also makes a theoretical and methodological contribution. To our knowledge, it is the first validated instrument designed to assess prompt engineering self-efficacy beliefs among pre-service teachers. In practice, the Ed-PESS addresses a gap by enabling systematic assessment and targeted development of this emerging self-efficacy construct in teacher education curricula.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthical Approval.\u0026nbsp;\u003c/strong\u003eThis research received ethical approval from the Scientific Research and Publication Ethics Committee at [removed for blinded review] , as per decision number [removed for blinded review].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeclaration of generative AI and AI-assisted technologies in the writing process.\u003c/strong\u003e During the preparation of this research, the authors used \u003cem\u003ePaperpal\u003c/em\u003e to paraphrase their writing for more academic enhancement, and \u003cem\u003eClaude\u003c/em\u003e \u003cem\u003eOpus\u003c/em\u003e (AI tools) for language translation and text reduction. After using these AI tools, the author(s) reviewed and edited the content as necessary and take full responsibility for the content of the publication. Alongside using AI tools for language tasks, the author(s) thoroughly reviewed and accurately cited all references in this research. They verified each reference\u0026apos;s authenticity, including DOI links. It\u0026apos;s crucial to note that all data and findings come from properly cited sources, not AI-generated. The author(s) fully ensure the research\u0026apos;s integrity and accuracy.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions Statement.\u0026nbsp;\u003c/strong\u003eAuthor 1 conceptualized the research, designed the study, and supervised the project. Author 2 performed the data analysis and validated the results. Author 3 contributed to the methodology and wrote the discussion section. Author 4 conducted the literature review and wrote the introduction. Author 5 collected the data and critically reviewed the manuscript. All authors reviewed and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests.\u0026nbsp;\u003c/strong\u003eThe authors declare that they have no competing interests related to this research.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding.\u0026nbsp;\u003c/strong\u003eThe author reports no funding.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability.\u0026nbsp;\u003c/strong\u003eThe data and materials used in this study are available upon request from the corresponding author.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAgirdag, O. (2025). Beyond prompt engineering: Prompting (l)iteracy, linguistic capital, and educational inequality. \u003cem\u003eEducational Theory\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/edth.70057\u003c/span\u003e\u003cspan address=\"10.1111/edth.70057\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndewi, W., Waziana, W., Wibisono, D., Putra, K. A., Hastomo, T., \u0026amp; Oktarin, I. B. (2025). From prompting to proficiency: A mixed-methods analysis of prompting with ChatGPT versus lecturer interaction in an EFL classroom. \u003cem\u003eJournal of Studies in the English Language\u003c/em\u003e, \u003cem\u003e20\u003c/em\u003e(2), 210\u0026ndash;238. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.64731/jsel.v20i2.282318\u003c/span\u003e\u003cspan address=\"10.64731/jsel.v20i2.282318\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArocha, J. G. (2025). Critical phenomenology of prompting in artificial intelligence. \u003cem\u003eSophia-Colecci\u0026oacute;n de Filosof\u0026iacute;a de la Educaci\u0026oacute;n\u003c/em\u003e, \u003cem\u003e39\u003c/em\u003e, 141\u0026ndash;165. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.17163/soph.n39.2025.04\u003c/span\u003e\u003cspan address=\"10.17163/soph.n39.2025.04\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArtino, A. R. Jr. (2012). Academic self-efficacy: From educational theory to instructional practice. \u003cem\u003ePerspectives on Medical Education\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e(2), 76\u0026ndash;85. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s40037-012-0012-5\u003c/span\u003e\u003cspan address=\"10.1007/s40037-012-0012-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBandura, A. (2001). Social cognitive theory: An agentic perspective. \u003cem\u003eAnnual Review of Psychology\u003c/em\u003e, \u003cem\u003e52\u003c/em\u003e(1), 1\u0026ndash;26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1146/annurev.psych.52.1.1\u003c/span\u003e\u003cspan address=\"10.1146/annurev.psych.52.1.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBandura, A. (2006). Guide for constructing self-efficacy scales. In F. Pajares, \u0026amp; T. Urdan (Eds.), \u003cem\u003eSelf-efficacy beliefs of adolescents\u003c/em\u003e (pp. 307\u0026ndash;337). Information Age Publishing.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBiberman-Shalev, L. (2025). Prompting theory into practice: Utilizing ChatGPT-4 in a curriculum planning course. \u003cem\u003eEducation Sciences\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(2). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eArticle 196. https://doi.org/10.3390/educsci15020196\u003c/span\u003e\u003cspan address=\"Article 196. 10.3390/educsci15020196\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoateng, G. O., Neilands, T. B., Frongillo, E. A., Melgar-Qui\u0026ntilde;onez, H. R., \u0026amp; Young, S. L. (2018). Best practices for developing and validating scales for health, social, and behavioral research: A primer. \u003cem\u003eFrontiers in Public Health\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e, 149. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fpubh.2018.00149\u003c/span\u003e\u003cspan address=\"10.3389/fpubh.2018.00149\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoonstra, L. (2025). \u003cem\u003ePrompt engineering\u003c/em\u003e [Whitepaper]. Google. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.kaggle.com/whitepaper-prompt-engineering\u003c/span\u003e\u003cspan address=\"https://www.kaggle.com/whitepaper-prompt-engineering\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown, T. A. (2015). \u003cem\u003eConfirmatory factor analysis for applied research\u003c/em\u003e. The Guilford Press.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarl, K., Dignam, C., Kochan, M., Alston, C., \u0026amp; Green, D. (2024). Discovering prompt engineering: A qualitative study of nonexpert teachers' interactions with ChatGPT. \u003cem\u003eIssues in Information Systems\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(4), 205\u0026ndash;220. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.48009/4_iis_2024_117\u003c/span\u003e\u003cspan address=\"10.48009/4_iis_2024_117\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCelik, I. (2023). Towards Intelligent-TPACK: An empirical study on teachers\u0026rsquo; professional knowledge to ethically integrate artificial intelligence (AI)-based tools into education. \u003cem\u003eComputers in Human Behavior\u003c/em\u003e, \u003cem\u003e138\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.chb.2022.107468\u003c/span\u003e\u003cspan address=\"10.1016/j.chb.2022.107468\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Article 107468.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCelik, I., Kontkanen, S., Laru, J., \u0026amp; Dalyanci, A. A. (2026). Co-constructing adaptive lesson plans with GenAI: Pre-service teachers' Intelligent-TPACK and prompt engineering strategies. \u003cem\u003eComputers and Education\u003c/em\u003e, \u003cem\u003e241\u003c/em\u003e, 105485. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.compedu.2025.105485\u003c/span\u003e\u003cspan address=\"10.1016/j.compedu.2025.105485\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChiu, T. K. F., Ahmad, Z., \u0026amp; \u0026Ccedil;oban, M. (2025). Development and validation of teacher artificial intelligence (AI) competence self-efficacy (TAICS) scale. \u003cem\u003eEducation and Information Technologies\u003c/em\u003e, \u003cem\u003e30\u003c/em\u003e, 6667\u0026ndash;6685. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10639-024-13094-z\u003c/span\u003e\u003cspan address=\"10.1007/s10639-024-13094-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCorreia, A. P., Hickey, S., \u0026amp; Xu, F. (2025). Realizing the possibilities of the large language models: Strategies for prompt engineering in educational inquiries. \u003cem\u003eTheory Into Practice\u003c/em\u003e, \u003cem\u003e64\u003c/em\u003e(4), 434\u0026ndash;447. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/00405841.2025.2528545\u003c/span\u003e\u003cspan address=\"10.1080/00405841.2025.2528545\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCrosthwaite, P. R., Smala, S., \u0026amp; Spinelli, F. (2025). Prompting for pedagogy? Australian F-10 teachers' generative AI prompting use cases. \u003cem\u003eAustralian Educational Researcher\u003c/em\u003e, \u003cem\u003e52\u003c/em\u003e(3), 1795\u0026ndash;1825. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s13384-024-00787-0\u003c/span\u003e\u003cspan address=\"10.1007/s13384-024-00787-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavidson, T. R., \u0026amp; Karell, D. (2025). Integrating generative artificial intelligence into social science research: Measurement, prompting, and simulation. \u003cem\u003eSociological Methods and Research\u003c/em\u003e, \u003cem\u003e54\u003c/em\u003e(3), 775\u0026ndash;793. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/00491241251339184\u003c/span\u003e\u003cspan address=\"10.1177/00491241251339184\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavila-Moran, R. C., Soto, S., Gomez, J. M. L., Silva Infantes, H. E., Arias Lizares, M., Huanca Rojas, A., L. M., \u0026amp; Cama Flores, S. J. (2025). Brief prompt-engineering clinic substantially improves AI literacy and reduces technology anxiety in first-year teacher-education students: A pre-post pilot study. \u003cem\u003eEducation Sciences\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(8). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/educsci15081010\u003c/span\u003e\u003cspan address=\"10.3390/educsci15081010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Article 1010.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eElSayary, A., Kuhail, M. A., \u0026amp; Hojeij, Z. (2025). Examining the role of prompt engineering in utilizing generative AI tools for lesson planning: Insights from teachers\u0026rsquo; experiences and perceptions. \u003cem\u003eHuman Behavior and Emerging Technologies\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1155/hbe2/9986139\u003c/span\u003e\u003cspan address=\"10.1155/hbe2/9986139\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFederiakin, D., Molerov, D., Zlatkin-Troitschanskaia, O., \u0026amp; Maur, A. (2024). Prompt engineering as a new 21st century skill. \u003cem\u003eFrontiers in Education\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e, 1366434. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/feduc.2024.1366434\u003c/span\u003e\u003cspan address=\"10.3389/feduc.2024.1366434\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFornell, C., \u0026amp; Larcker, D. (1981). Evaluating structural equation models with unobservable variables and measurement error. \u003cem\u003eJournal of Marketing Research\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e(1), 39\u0026ndash;50. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2307/3151312\u003c/span\u003e\u003cspan address=\"10.2307/3151312\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGibreel, O., \u0026amp; Arpaci, I. (2025). Development and validation of the prompt engineering competence scale (PECS). \u003cem\u003eInformation Development\u003c/em\u003e. Advance online publication. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/02666669251336455\u003c/span\u003e\u003cspan address=\"10.1177/02666669251336455\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eG\u0026uuml;m\u0026uuml;ş, M., \u0026amp; Kukul, V. (2023). Developing a digital competence scale for teachers: Validity and reliability study. \u003cem\u003eEducation and Information Technologies\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e, 2747\u0026ndash;2765. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10639-022-11411-z\u003c/span\u003e\u003cspan address=\"10.1007/s10639-022-11411-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHair, J. F., Black, W. C., Babin, B. J., \u0026amp; Anderson, R. E. (2019). \u003cem\u003eMultivariate data analysis\u003c/em\u003e. Cengage Learning EMEA.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHsu, H. P. (2025). From programming to prompting: Developing computational thinking through large language model-based generative artificial intelligence. \u003cem\u003eTechTrends\u003c/em\u003e, \u003cem\u003e69\u003c/em\u003e(3), 485\u0026ndash;506. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11528-025-01052-6\u003c/span\u003e\u003cspan address=\"10.1007/s11528-025-01052-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, K. L., Liu, Y. C., \u0026amp; Dong, M. Q. (2024). Incorporating AIGC into design ideation: A study on self-efficacy and learning experience acceptance under higher-order thinking. \u003cem\u003eThinking Skills and Creativity\u003c/em\u003e, \u003cem\u003e52\u003c/em\u003e, 101508. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.tsc.2024.101508\u003c/span\u003e\u003cspan address=\"10.1016/j.tsc.2024.101508\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHwang, M., Jeens, R., \u0026amp; Lee, H. K. (2025). Exploring learner prompting behavior and its effect on ChatGPT-assisted English writing revision. \u003cem\u003eThe Asia-Pacific Education Researcher\u003c/em\u003e, \u003cem\u003e34\u003c/em\u003e, 1157\u0026ndash;1167. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s40299-024-00930-6\u003c/span\u003e\u003cspan address=\"10.1007/s40299-024-00930-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJacobsen, L. J., \u0026amp; Weber, K. E. (2025). The promises and pitfalls of large language models as feedback providers: A study of prompt engineering and the quality of AI-driven feedback. \u003cem\u003eAI\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(2). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ai6020035\u003c/span\u003e\u003cspan address=\"10.3390/ai6020035\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJang, A., Oh, M., \u0026amp; Song, M. O. (2025). Question-centered prompt engineering for nursing simulation debriefing: Model development and validation. \u003cem\u003eNurse Education in Practice\u003c/em\u003e, \u003cem\u003e88\u003c/em\u003e., Article 104541. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.nepr.2025.104541\u003c/span\u003e\u003cspan address=\"10.1016/j.nepr.2025.104541\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJin, Z., Meng, G., Wang, X., Wang, J., Liu, C., \u0026amp; Zhang, J. (2025). Understanding user prompting behavior in generative AI: A component analysis. \u003cem\u003eProceedings of the Association for Information Science and Technology, 62\u003c/em\u003e(1), 941\u0026ndash;945. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/pra2.1318\u003c/span\u003e\u003cspan address=\"10.1002/pra2.1318\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKabeer, A., Bhat, R. A., Antony, S., \u0026amp; Tramboo, I. A. (2025). Enhancing creative writing skills in secondary school students through prompt engineering and artificial intelligence. \u003cem\u003eForum for Linguistic Studies\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(3), 800\u0026ndash;815. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.30564/fls.v7i3.8511\u003c/span\u003e\u003cspan address=\"10.30564/fls.v7i3.8511\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKang, L., Shi, X., \u0026amp; Zhu, K. (2025). Uncovering the mediation of disciplinary literacy in the effect of GAI prompt engineering on pre-service teachers' instructional design. \u003cem\u003eEducation and Information Technologies\u003c/em\u003e, \u003cem\u003e30\u003c/em\u003e(16), 22779\u0026ndash;22802. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10639-025-13657-8\u003c/span\u003e\u003cspan address=\"10.1007/s10639-025-13657-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKline, R. B. (2011). \u003cem\u003ePrinciples and practice of structural equation modeling\u003c/em\u003e. The Guilford Press.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLaupichler, M. C., Aster, A., Haverkamp, N., \u0026amp; Raupach, T. (2023). Development of the Scale for the assessment of non-experts' AI literacy: An exploratory factor analysis. \u003cem\u003eComputers in Human Behavior Reports\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(4). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eArticle 100338. https://doi.org/10.1016/j.chbr.2023.100338\u003c/span\u003e\u003cspan address=\"Article 100338. 10.1016/j.chbr.2023.100338\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLawshe, C. H. (1975). A quantitative approach to content validity. \u003cem\u003ePersonnel Psychology\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e(4), 563\u0026ndash;575. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1744-6570.1975.tb01393.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1744-6570.1975.tb01393.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee, D., \u0026amp; Palmer, E. (2025). Prompt engineering in higher education: A systematic review to help inform curricula. \u003cem\u003eInternational Journal of Educational Technology in Higher Education\u003c/em\u003e, \u003cem\u003e22\u003c/em\u003e(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s41239-025-00503-7\u003c/span\u003e\u003cspan address=\"10.1186/s41239-025-00503-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee, U., Jung, H., Jeon, Y., Sohn, Y., Hwang, W., Moon, J., \u0026amp; Kim, H. (2024). Few-shot is enough: Exploring ChatGPT prompt engineering method for automatic question generation in English education. \u003cem\u003eEducation and Information Technologies\u003c/em\u003e, \u003cem\u003e29\u003c/em\u003e(9), 11483\u0026ndash;11515. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10639-023-12249-8\u003c/span\u003e\u003cspan address=\"10.1007/s10639-023-12249-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLo, L. S. (2023). The CLEAR path: A framework for enhancing information literacy through prompt engineering. \u003cem\u003eJournal of Academic Librarianship\u003c/em\u003e, \u003cem\u003e49\u003c/em\u003e(4), 102720. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.acalib.2023.102720\u003c/span\u003e\u003cspan address=\"10.1016/j.acalib.2023.102720\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLong, D., \u0026amp; Magerko, B. (2020). What is AI literacy? Competencies and design considerations. In \u003cem\u003eProceedings of the 2020 CHI Conference on Human Factors in Computing Systems\u003c/em\u003e (pp. 1\u0026ndash;16). Association for Computing Machinery. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1145/3313831.3376727\u003c/span\u003e\u003cspan address=\"10.1145/3313831.3376727\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMendes, A., Parizotto, A., Garcia, R., Garcia, M., Guedes, G., Vilela, R., Valle, P., Balancieri, R., \u0026amp; Silva, W. (2025). CoderBot 2.0: Integrating LLM and prompt engineering in the evolution of an educational chatbot. In \u003cem\u003eSimp\u0026oacute;sio Brasileiro de Inform\u0026aacute;tica na Educa\u0026ccedil;\u0026atilde;o (SBIE)\u003c/em\u003e (pp. 1497\u0026ndash;1506). SBC. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5753/sbie.2025.12321\u003c/span\u003e\u003cspan address=\"10.5753/sbie.2025.12321\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMishra, P., \u0026amp; Koehler, M. J. (2006). Technological pedagogical content knowledge: A framework for teacher knowledge. \u003cem\u003eTeachers College Record\u003c/em\u003e, \u003cem\u003e108\u003c/em\u003e(6), 1017\u0026ndash;1054. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/016146810610800610\u003c/span\u003e\u003cspan address=\"10.1177/016146810610800610\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoorhouse, B. L., Ho, T. Y., Wu, C., \u0026amp; Wan, Y. (2025). Pre-service language teachers' task-specific large language model prompting practices. \u003cem\u003eRELC Journal\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/00336882251313701\u003c/span\u003e\u003cspan address=\"10.1177/00336882251313701\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Advance online publication.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMutanga, M. B., Msane, J., Mndaweni, T. N., Hlongwane, B. B., \u0026amp; Ngcobo, N. Z. (2025). Exploring the impact of LLM prompting on students' learning. \u003cem\u003eTrends in Higher Education\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e(3). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eArticle 31. https://doi.org/10.3390/higheredu4030031\u003c/span\u003e\u003cspan address=\"Article 31. 10.3390/higheredu4030031\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMzwri, K., \u0026amp; Turcs\u0026aacute;nyi-Szab\u0026oacute;, M. (2025). The impact of prompt engineering and a generative AI-driven tool on autonomous learning: A case study. \u003cem\u003eEducation Sciences\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(2). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eArticle 199. https://doi.org/10.3390/educsci15020199\u003c/span\u003e\u003cspan address=\"Article 199. 10.3390/educsci15020199\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNg, D. T. K., Leung, J. K. L., Chu, S. K. W., \u0026amp; Qiao, M. S. (2021). Conceptualizing AI literacy: An exploratory review. \u003cem\u003eComputers and Education: Artificial Intelligence\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e, 100041. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.caeai.2021.100041\u003c/span\u003e\u003cspan address=\"10.1016/j.caeai.2021.100041\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOliveira, L., Tavares, C., Strzelecki, A., \u0026amp; Silva, M. (2025). Prompting minds: Evaluating how students perceive generative AI's critical thinking dispositions. \u003cem\u003eElectronic Journal of e-Learning\u003c/em\u003e, \u003cem\u003e23\u003c/em\u003e(2), 1\u0026ndash;18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.34190/ejel.23.2.3986\u003c/span\u003e\u003cspan address=\"10.34190/ejel.23.2.3986\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePark, J., \u0026amp; Choo, S. (2024). Generative AI prompt engineering for educators: Practical strategies. \u003cem\u003eJournal of Special Education Technology\u003c/em\u003e, \u003cem\u003e40\u003c/em\u003e(3), 411\u0026ndash;417. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/01626434241298954\u003c/span\u003e\u003cspan address=\"10.1177/01626434241298954\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePratschke, B. M. (2024). \u003cem\u003eGenerative AI and education: Digital pedagogies, teaching innovation and learning design\u003c/em\u003e. Springer. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-3-031-67991-9\u003c/span\u003e\u003cspan address=\"10.1007/978-3-031-67991-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQian, Y. (2025). Prompt engineering in education: A systematic review of approaches and educational applications. \u003cem\u003eJournal of Educational Computing Research\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/07356331251365189\u003c/span\u003e\u003cspan address=\"10.1177/07356331251365189\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Advance online publication.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRabardel, P. (2002). \u003cem\u003ePeople and technology in the workplace: A cognitive ergonomic approach\u003c/em\u003e. HAL Open Science.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSelwyn, N., Ljungqvist, M., \u0026amp; Sonesson, A. (2025). When the prompting stops: Exploring teachers\u0026rsquo; work around the educational frailties of generative AI tools. \u003cem\u003eLearning Media and Technology\u003c/em\u003e, \u003cem\u003e50\u003c/em\u003e(3), 310\u0026ndash;323. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/17439884.2025.2537959\u003c/span\u003e\u003cspan address=\"10.1080/17439884.2025.2537959\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSerra, P., \u0026amp; Oliveira, A. (2025). AI-powered prompt engineering for Education 4.0: Transforming digital resources into engaging learning experiences. \u003cem\u003eEducation Sciences\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(12). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eArticle 1640. https://doi.org/10.3390/educsci15121640\u003c/span\u003e\u003cspan address=\"Article 1640. 10.3390/educsci15121640\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShulman, L. S. (1986). Those who understand: Knowledge growth in teaching. \u003cem\u003eEducational Researcher\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(2), 4\u0026ndash;14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3102/0013189X015002004\u003c/span\u003e\u003cspan address=\"10.3102/0013189X015002004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSigot, M., \u0026amp; Tassoti, S. (2025). An investigation of change in prompting strategies in a semester-long course on the use of GenAI. \u003cem\u003eJournal of Chemical Education\u003c/em\u003e, \u003cem\u003e102\u003c/em\u003e(6), 2507\u0026ndash;2513. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1021/acs.jchemed.4c01287\u003c/span\u003e\u003cspan address=\"10.1021/acs.jchemed.4c01287\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSon, M., Won, Y. J., \u0026amp; Lee, S. (2025). Optimizing large language models: A deep dive into effective prompt engineering techniques. \u003cem\u003eApplied Sciences\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(3), 1430. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/app15031430\u003c/span\u003e\u003cspan address=\"10.3390/app15031430\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSweller, J. (1988). Cognitive load during problem solving: Effects on learning. \u003cem\u003eCognitive Science\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(2), 257\u0026ndash;285. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1207/s15516709cog1202_4\u003c/span\u003e\u003cspan address=\"10.1207/s15516709cog1202_4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTabachnick, B. G., \u0026amp; Fidell, L. S. (2013). \u003cem\u003eUsing multivariate statistics\u003c/em\u003e (6th ed.). Pearson Education.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eT\u0026uuml;men Akyıldız, S. (2026). Prompt engineering and creative pedagogical thinking: Insights from a qualitative study with teacher candidates. \u003cem\u003eThinking Skills and Creativity\u003c/em\u003e, \u003cem\u003e60\u003c/em\u003e., Article 102105. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.tsc.2025.102105\u003c/span\u003e\u003cspan address=\"10.1016/j.tsc.2025.102105\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTzafilkou, K., Perifanou, M., \u0026amp; Economides, A. A. (2022). Development and validation of students' digital competence scale (SDiCoS). \u003cem\u003eInternational Journal of Educational Technology in Higher Education\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eArticle 30. https://doi.org/10.1186/s41239-022-00330-0\u003c/span\u003e\u003cspan address=\"Article 30. 10.1186/s41239-022-00330-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUNESCO. (2024). \u003cem\u003eAI competency framework for teachers\u003c/em\u003e. UNESCO Publishing.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Wyk, M. M. (2025). Student teachers' leveraging GenAI tools for academic writing, design, and prompting in an ODeL course. \u003cem\u003eOpen Praxis\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(1), 95\u0026ndash;107. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.55982/openpraxis.17.1.711\u003c/span\u003e\u003cspan address=\"10.55982/openpraxis.17.1.711\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Zoost, S. (2025). Prompting teacher identities: A model for teacher subjectivities constituted through artificial intelligence. \u003cem\u003eJournal of Teaching and Learning\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(4), 198\u0026ndash;215. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.22329/jtl.v19i4.10079\u003c/span\u003e\u003cspan address=\"10.22329/jtl.v19i4.10079\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVel\u0026aacute;squez-Henao, J. D., Franco-Cardona, C. J., \u0026amp; Cadavid-Higuita, L. (2023). Prompt engineering: A methodology for optimizing interactions with AI-language models in the field of engineering. \u003cem\u003eDYNA\u003c/em\u003e, \u003cem\u003e90\u003c/em\u003e(230), 9\u0026ndash;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.15446/dyna.v90n230.111700\u003c/span\u003e\u003cspan address=\"10.15446/dyna.v90n230.111700\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWalter, Y. (2024). Embracing the future of artificial intelligence in the classroom: The relevance of AI literacy, prompt engineering, and critical thinking in modern education. \u003cem\u003eInternational Journal of Educational Technology in Higher Education\u003c/em\u003e, \u003cem\u003e21\u003c/em\u003e(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s41239-024-00448-3\u003c/span\u003e\u003cspan address=\"10.1186/s41239-024-00448-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, B., Rau, P. L. P., \u0026amp; Yuan, T. (2023). Measuring user competence in using artificial intelligence: Validity and reliability of artificial intelligence literacy scale. \u003cem\u003eBehaviour \u0026amp; Information Technology\u003c/em\u003e, \u003cem\u003e42\u003c/em\u003e(9), 1324\u0026ndash;1337. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/0144929X.2022.2072768\u003c/span\u003e\u003cspan address=\"10.1080/0144929X.2022.2072768\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, Y. Y., \u0026amp; Chuang, Y. W. (2024). Artificial intelligence self-efficacy: Scale development and validation. \u003cem\u003eEducation and Information Technologies\u003c/em\u003e, \u003cem\u003e29\u003c/em\u003e(4), 4785\u0026ndash;4808. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10639-023-12015-w\u003c/span\u003e\u003cspan address=\"10.1007/s10639-023-12015-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWoo, D. J., Wang, D., Yung, T., \u0026amp; Guo, K. (2025). Effects of a prompt engineering intervention on undergraduate students' AI self-efficacy, AI knowledge and prompt engineering ability: A mixed methods study. \u003cem\u003eBritish Educational Research Journal\u003c/em\u003e. Advance online publication. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/berj.70087\u003c/span\u003e\u003cspan address=\"10.1002/berj.70087\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZimmerman, B. J. (2002). Becoming a self-regulated learner: An overview. \u003cem\u003eTheory Into Practice\u003c/em\u003e, \u003cem\u003e41\u003c/em\u003e(2), 64\u0026ndash;70. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1207/s15430421tip4102_2\u003c/span\u003e\u003cspan address=\"10.1207/s15430421tip4102_2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"prompt engineering, educational prompt engineering, prompt engineering self-efficacy, AI, scale development","lastPublishedDoi":"10.21203/rs.3.rs-9003817/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9003817/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePre-service teachers' confidence in designing pedagogically purposeful prompts for generative AI tools remains an underexamined construct in teacher education. This study developed and validated the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS) to address the absence of a psychometrically sound instrument measuring this domain. Scale development followed Boateng et al.'s three-phase framework, grounded in a systematic literature review and expert review procedures. Two independent samples of pre-service teachers were recruited: an exploratory factor analysis (EFA) sample (n\u0026thinsp;=\u0026thinsp;300) and a confirmatory factor analysis (CFA) sample (n\u0026thinsp;=\u0026thinsp;305). EFA using principal axis factoring with Promax rotation produced a four-factor structure explaining 64.60% of total variance. CFA with Yuan-Bentler robust correction confirmed acceptable model fit. The final 26-item scale comprises four dimensions: Basic Prompt Engineering, Pedagogical Prompt Engineering, Ethical Prompting Practice, and Continuous Professional Development. Internal consistency was high across all subscales. The Ed-PESS provides a valid, reliable instrument for assessing prompt engineering self-efficacy in pre-service teacher education. It supports formative curriculum evaluation and targeted intervention design.\u003c/p\u003e","manuscriptTitle":"Educational Prompt Engineering Self-Efficacy Scale (Ed-PESS): Scale Development and Psychometric Validation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-03 08:04:37","doi":"10.21203/rs.3.rs-9003817/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"642777b7-d274-4527-9d8f-4811fed8711b","owner":[],"postedDate":"March 3rd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-03-05T22:23:25+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-03 08:04:37","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9003817","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9003817","identity":"rs-9003817","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}