Profiling Preservice Teachers’ Computational Thinking: The Role of Metacognition and Coding Experience

Profiling Preservice Teachers’ Computational Thinking: The Role of Metacognition and Coding Experience | 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 Profiling Preservice Teachers’ Computational Thinking: The Role of Metacognition and Coding Experience Tanya Chichekian, Marian Cutumisu, Annie Savard This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6283533/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 In our technology-driven world, developing computational thinking skills among preservice teachers is essential for fostering digital innovation and preparing them to effectively integrate technology into their future classrooms. This study investigates how metacognitive strategies and prior coding experience impact preservice teachers’ computational thinking skills and online problem-solving behaviors during an interactive task. A sample of preservice teachers (n=129) completed a survey assessing their computational thinking skills and metacognitive strategies and then engaged in completing an online task. Latent profile analysis using self-reported computational thinking skills identified three distinct profiles: Developing, Novice, and Proficient. Key results revealed that higher levels of self-reported computational thinking skills correlated positively with better performance on the online task, as well as lower perceptions of task difficulty. Both metacognitive strategies and prior coding experience were significant predictors of computational thinking skills profile membership. This study highlights the importance of integrating targeted coding exercises and metacognitive skill-building activities into teacher education programs to better prepare future educators to implement digital technologies effectively and confidently in their teaching practices. The findings provide actionable insights for designing teacher training that fosters computational thinking competencies essential for the modern digital era. Educational Psychology computational thinking preservice teachers teacher education coding metacognition Figures Figure 1 Introduction In our increasingly technology-oriented society, computational thinking (CT) has become a critical skill for engaging with digital innovation (Kotsopoulos et al., 2017 ). This cognitive ability is essential for solving complex and novel problems (Barr & Stephenson, 2011 ). According to Wing ( 2006 ), CT is a way to comprehend human behavior through computing principles. It can also be seen as a way of thinking algorithmically, drawing from computer science principles as a guiding framework, at times metaphorically (Shodiev, 2015 ). Hoyles and Noss (2015) define CT as including skills such as abstraction (viewing problems at varying levels of detail), algorithmic thinking (expressing problems as a series of step-by-step instructions), decomposition (tackling a problem by solving smaller constituent issues), and pattern recognition (identifying connections between a new problem and previous ones). While traditionally associated with computer science, CT is a versatile skillset that transcends disciplines, enhancing problem-solving capabilities and analytical skills over time (Barr & Stephenson, 2011 ). Therefore, understanding how pedagogical experiences contribute to developing students’ CT skills is imperative in higher education. This knowledge provides valuable insights into shaping students’ perceptions and attitudes towards CT, with the potential to facilitate the integration of these skills into various subjects across different educational levels (Author 1). In an era where technology is omnipresent, developing CT skills among future educators has increasingly become a focal point. The significance of these skills for preservice teachers is a rapidly expanding field of interest. Despite this, research on the factors influencing these skills (Author 2) among preservice teachers remains limited. While prior research has explored CT development in K-12 students or in-service teachers, this study addresses a critical gap by focusing on preservice teachers, a population uniquely positioned to shape future generations’ computational literacy. By investigating the interplay between metacognitive strategies and coding experience, this study advances understanding of how these factors synergistically influence CT skill acquisition. Furthermore, it employs latent profile analysis (LPA) to categorize preservice teachers into distinct CT proficiency profiles, offering a novel framework for designing differentiated pedagogical interventions in teacher education programs. Specifically, this study seeks to answer the following research questions: (1) How can preservice teachers be differentiated based on their CT skills? (2) To what extent do preservice teachers’ metacognitive strategies and prior coding experience predict CT skills? (3) In what ways do online behaviors on a CT interactive task differ based on preservice teachers’ CT skills? Literature Review Over the past decade, various researchers have explored different models for computational thinking (CT), with an emphasis on aspects such as concepts, practices, and perspectives (Barr & Stephenson, 2011 ; Weintrop et al., 2016 ). From an educational standpoint, an effective CT model should transcend mere coding and programming language knowledge, prioritizing pedagogical approaches to teaching CT through specific classroom activities. For instance, Shute et al. ( 2017 ) have proposed a CT model that represents a spectrum of increasing levels of foundational understanding and CT skills. This conceptual foundation is particularly relevant for educational experiences that aim to design problem-solving scenarios with solutions that can be applied in different contexts. According to Shute et al. ( 2017 ), CT involves the ability to solve problems effectively and efficiently, whether algorithmically with or without computer assistance, and to create solutions that can be reused in various situations. This definition underscores that CT is fundamentally a mindset and a way of acting, which can be demonstrated through specific skills that can serve as the basis for evaluating CT abilities through performance-based assessments. The development of CT skills in educational settings has been approached through both unplugged (non-digital) and plugged (digital) methods. Unplugged methods, which do not require the use of computers, focus on the conceptual underpinnings of computing and have been particularly effective in making abstract concepts accessible to beginners (Kotsopoulos et al., 2017 ). These methods often involve hands-on activities, such as puzzles, games, and physical manipulatives that help learners grasp foundational concepts like algorithms, decomposition, and pattern recognition without the need for digital tools. Conversely, plugged methods involve direct interaction with technology, offering a more hands-on approach to developing computational skills. These methods are crucial for understanding how to apply computational concepts in real-world contexts, particularly within programming environments (Shute et al., 2017 ) where students can experiment with coding languages and computational tools. For example, tools like Scratch and robotics kits have been widely used to teach CT through interactive, project-based learning (Jaipal-Jamani & Angeli, 2017 ; Mouza et al., 2017 ). Research comparing these approaches (Román-González et al., 2019 ) suggests that a blend of both may be most effective, with unplugged activities building foundational understanding and plugged activities solidifying practical application. This shift from unplugged to plugged is particularly pertinent in teacher education, where developing CT capabilities is seen not only as a skill but as a fundamental teaching competency. Challenges in Enhancing Preservice Teachers’ CT Skills Despite the growing recognition of CT's importance, there remains a significant disparity in the preparedness of preservice teachers to teach CT skills. Research indicates that many preservice teachers enter their programs with minimal exposure to computational concepts, leading to a steep learning curve and uneven skill development across cohorts (Bower & Falkner, 2015 ; Sadik, 2017). This lack of preparedness is compounded by the fact that teacher education programs often lack structured CT-focused coursework, leaving many preservice teachers ill-equipped to integrate CT into their future classrooms (Grover & Pea, 2013 ; Tang et al., 2020 ). Moreover, the design and execution of CT activities can significantly impact preservice teachers' comprehension and skill development. For instance, some preservice teachers may develop misconceptions about CT, particularly if the activities are not aligned with their prior knowledge or if the instructional strategies fail to address their individual learning needs (Sadik, 2017). Common misconceptions among preservice teachers include viewing CT as solely programming-related or believing that it is only relevant to STEM subjects. Some may struggle with the abstract nature of CT concepts or lack confidence in their ability to teach these skills (Bers, 2018 ). Addressing these misconceptions may require explicit instruction on how to effectively design and implement CT-focused interventions in teacher education programs. For example, Jaipal-Jamani and Angeli ( 2017 ) found that a robotics course significantly improved preservice teachers' self-efficacy, science learning, and CT skills. Similarly, Mouza et al. ( 2017 ) reported significant improvements in preservice teachers' ability to think computationally after integrating CT into a 15-week course using tools like Scratch. While past studies advocate how various types of interventions have proven effective in engaging preservice teachers and fostering CT skills (Brennan & Resnick, 2012 ) through Block-based programming tools like Scratch and educational robotics kits, the key is to provide structured activities that progressively build CT concepts and encourage creative problem-solving. Constructionist learning environments, which encourage learners to construct their own understanding through active exploration, have also been shown to be effective in enriching preservice teachers' CT skills (Butler & Leahy, 2021 ). These environments provide opportunities for preservice teachers to engage in hands-on, project-based learning, which can enhance their ability to apply computational concepts in real-world contexts. However, the design of these interventions must be carefully tailored to the needs of preservice teachers, taking into account their prior knowledge, especially in coding, and learning strategies. The Role of Metacognitive Strategies and Prior Coding Experience on CT Skills Metacognitive strategies, which involve self-regulation and reflective thinking, have been shown to play a crucial role in the development of CT skills, particularly in teacher preparation programs where the dual goal is to develop both CT skills and the ability to teach these skills effectively. According to Pintrich et al. ( 1993 ), metacognitive strategies help learners adapt and apply computational methods in problem-solving scenarios, leading to deeper and more transferable learning outcomes. In the context of CT, metacognitive strategies can enhance preservice teachers' ability to plan, monitor, and evaluate their problem-solving processes, thereby improving their overall computational proficiency (Cheng et al., 2023 ). Recent studies have also highlighted the importance of integrating metacognitive training into CT-focused interventions. For example, Argelagós et al. ( 2022 ) found that preservice teachers who engaged in metacognitive reflection during CT tasks demonstrated significant improvements in their problem-solving skills and ability to transfer these skills to new contexts. Prior coding experience has also been identified as a significant predictor of CT skills among preservice teachers. Research by Chan et al. ( 2021 ) found that preservice teachers with prior coding experience demonstrated higher levels of digital literacy, problem-solving skills, and comfort with coding compared to their peers without such experience. This finding aligns with the broader literature on CT, which emphasizes the importance of practical coding knowledge in enhancing computational proficiency (Grover & Pea, 2013 ; Shute et al., 2017 ). However, the relationship between prior coding experience and CT skills is not always straightforward. While coding experience can provide a foundation for understanding computational concepts, it does not guarantee the development of higher-order CT skills, such as abstraction and algorithmic thinking (Bower & Falkner, 2015 ). This suggests that teacher education programs must go beyond teaching coding languages and focus on fostering a deeper understanding of computational principles and their application in diverse contexts. Recent systematic reviews (e.g., Dong et al., 2024 ; Liu et al., 2024 ) highlight a paucity of research on how preservice teachers develop CT skills, particularly through the lens of metacognitive reflection and prior coding exposure. While studies have examined CT components in isolation (e.g., coding drills or problem-solving frameworks), few have explored how these elements interact within preservice teacher populations. Liao et al. ( 2022 ) found that prior coding experience positively correlates with CT skills, suggesting that familiarity with programming languages can facilitate the development of computational thinking. However, the interplay between coding experience and metacognitive strategies remains unclear. It is possible that preservice teachers with prior coding experience may rely more on metacognitive strategies to navigate complex problems, but this hypothesis has yet to be tested empirically. This gap is significant, as preservice teachers must not only master CT themselves but also learn to teach it, a dual challenge requiring both technical proficiency and reflective pedagogical strategies. Our study bridges this gap by situating metacognition and coding experience as interdependent predictors of CT, aligning with calls for more nuanced investigations into teacher-specific CT development pathways (Tang et al., 2020 ). The present research This study extends the application of Shute et al.’s ( 2017 ) model by focusing on preservice teachers, a group for whom the integration of CT skills is crucial yet understudied. Preservice teachers not only need to develop their own CT skills but also must be prepared to teach these skills, making their mastery doubly significant. This dual need aligns with Hoyles and Noss's (2015) perspective that CT should be embedded in pedagogy, not just as a subject to be taught but as a method of teaching and learning. Furthermore, the notion that CT can enhance problem-solving capabilities beyond computer science (Barr & Stephenson, 2011 ) provides a theoretical basis for investigating CT across educational disciplines. This is particularly important in teacher education, where these skills can significantly impact how future educators integrate technology into their teaching practices. This study also examines the roles of metacognitive strategies and previous coding experiences in shaping CT skills among preservice teachers. Metacognitive strategies are hypothesized to enhance the acquisition of CT skills as they help learners adapt and apply computational methods in problem-solving scenarios, potentially leading to deeper and more transferable learning outcomes (Pintrich et al., 1993 ). Moreover, prior coding experience not only fosters familiarity with computational tools and languages but is considered as a practical component of CT skills development. As a response to calls for more empirical research on how specific educational backgrounds influence CT (Liao et al., 2022 ), this study will also examine how preservice teachers’ prior coding knowledge interacts with their metacognitive strategies to predict a set of CT skills. The interaction between metacognition and prior coding experience is particularly relevant because individuals with coding backgrounds may be better equipped to leverage metacognitive strategies when problem-solving in computational contexts. Research in cognitive psychology suggests that metacognitive skills, such as planning, monitoring, and evaluating one's own thinking processes, are crucial for effective learning and problem-solving (Flavell, 1979 ). By investigating how these factors interplay, this study contributes to the field by providing empirical evidence on how these factors interact to influence CT proficiency and online problem-solving behaviors, thereby potentially informing the design of more effective teacher education programs. Methods Participants This quantitative study was conducted with 128 preservice teachers (17 males, 101 females, 8 other, and 2 missing) from a Faculty of Education. There were n = 19 participants in Year 1, n = 9 in Year 2, n = 8 in Year 3, n = 73 in Year 4, and n = 19 in Year 4+. Participants were recruited from various programs: n = 73 from Elementary Education, n = 37 from Secondary Education, and n = 18 from Physical Education, TESL, or Music. Finally, n = 62 had some familiarity with a coding language, such as Scratch, Python, C++, JavaScript, and HTML. Data Collection A research assistant visited classes whose professors had agreed upon to explain the project and asked participants’ consent before filling out an online survey on Qualtrics. Participants responded to Likert-type scales and were asked to complete a CT interactive task in which they were randomly placed in by level of difficulty. CT Interactive Task Design The CT task named, Programming Lamps was derived from one of the Bebras tasks found in the Bebras international contest (Dagiene & Futschek, 2008) which consists of a set of activities requiring students to transfer and project their CT skills to solve “real-life” problems. Given that these tasks do not require any prior knowledge of any software or hardware, they could be administered to individuals without any prior programming experience. Three computer science interns worked on optimizing a series of CT tasks (including Programming Lamps ) by adding difficulty levels (i.e., easy, medium, hard) with incremental complexity, an animation of the participants’ responses as a form of feedback, as well as an animation of the correct solution. A first pilot of the optimized tasks was tested with close friends and colleagues to ask for feedback and refine the prototype including the three difficulty levels of Programming Lamps . Feedback questions focused on the user experience, namely ease of understanding regarding the description and goal of the task, text and response field formatting, and picture quality. Suggestions on how to improve user experience were also part of the feedback that was sought. A second pilot test was conducted with a student population from an online crowdsourcing platform to validate the optimized interactive tasks with other forms of combined assessment (e.g., surveys on CT skills). These optimized CT tasks also easily catered to the implementation of learning analytics. For example, programming for click counts from the beginning to the end of task completion was embedded as a form of data-mining assessment and a playground was added before beginning the actual CT task as a form of formative-iterative assessment. Automated feedback was also incorporated by providing students with two occasions to correctly complete the actual task. After submitting their response, participants received a replay of their solution as feedback if their submitted response was incorrect before continuing with a second attempt. Measures Demographic information was collected from the participants, such as gender, program, year of study, and prior coding experience. Computational thinking (CT) skills. The CT skills were adapted from the Callysto Computational Thinking test – student version (Author 3). We specifically asked preservice teachers to rate their agreement on a scale of 1–7 their perceptions regarding their CT skills, namely in relation to digital technology (DL, 3 items, α = .77), problem-solving (PS, 3 items, α = .77), and comfort with coding (CD, 3 items, α = .73) for this study. CT task score and online learning behaviors. The optimized CT task Programming Lamps rendered either a score ( L_Score ) of “1” when the participants did not get the right answer on both attempts, a “2” when the participants answered correctly on the second attempt after viewing automated feedback (i.e., a replay of their wrong solution), or a “3” when the participants reached the correct answer on the first attempt. The time spent on playground ( PG_Time ), the number of clicks (L_clicks) , and time spent on completing the task ( L_Time ) were also recorded. Perceptions of difficulty level (L_diff). At the end of the task, preservice teachers were asked to rate from 1 (very easy) to 10 (very hard), the difficulty level they associated with the task. Metacognitive learning strategies (MetaCog). Preservice teachers reported the use of different metacognitive strategies (Pintrich et al., 1993 ) when preparing a lesson plan for teaching a complex concept (9 items, α = .80) Data Analysis Latent profile analysis (LPA) was employed to explore the underlying profiles within the data based on preservice teachers’ responses regarding their CT skills (DL-Digital Literacy, PS-Problem Solving, and CD-Comfort in Coding). We began by standardizing these variables, followed by an estimation of a series of models with one to five profiles to determine the optimal number of profiles, using fit indices such as the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), entropy, and the Bootstrap Likelihood Ratio Test (BLRT). To determine the extent to which metacognitive strategies and prior coding experience predicted group membership, a linear discriminant analysis (LDA) was conducted. LDA is particularly useful for smaller datasets as it is designed to model differences in groups by assuming Gaussian distributions and can provide insights into which behaviors are most discriminative between profiles. To examine how outcomes related to online behaviors (i.e., PG_Time, L_Time, L_Clicks, and L_Diff ) varied based on preservice teachers’ CT skills, a one-way ANOVA was conducted across the three profiles, followed by post hoc LSD comparisons. Based on the first research question regarding the differentiation of preservice teachers based on CT skills, it was hypothesized that preservice teachers will exhibit distinct profiles characterized by clusters of high, moderate, and low proficiency in CT components (H1a) and that they will differ significantly from one another (H1b). More specifically, it was expected that preservice teachers with prior coding experience (e.g., familiarity with Scratch, Python, C++, JavaScript, HTML) will report higher levels of CT skills (Digital Literacy (DL), Problem Solving (PS), and Comfort with Coding (CD)) than those without such experience. This hypothesis aligns with Shute et al.’s ( 2017 ) CT model, which posits that learners progress through increasing levels of proficiency. Given that CT is a multidimensional construct (Hoyles & Noss, 2015) this hypothesis also parallels Román-González et al.’s ( 2019 ) suggestion concerning the variability in CT skill acquisition, particularly among novices. Regarding the second research question about the predictors of CT skills, it was hypothesized that prior coding experience and metacognitive strategies would independently predict preservice teachers’ CT skills. Specifically, preservice teachers with prior coding experience (e.g., familiarity with Scratch, Python, or HTML) were expected to exhibit stronger CT proficiency, as coding provides foundational exposure to algorithmic thinking and problem-solving (Chan et al., 2021 ). Similarly, those using more sophisticated metacognitive strategies were anticipated to demonstrate higher CT skills, as metacognition enables adaptive learning and reflection (Pintrich et al., 1993 ; Argelagós et al., 2022 ) (H2a). It was also hypothesized that the combination of prior coding experience and metacognitive strategies would most effectively differentiate preservice teachers’ CT proficiency profiles (H2b). Drawing on cognitive psychology frameworks (Flavell, 1979 ), we posited that preservice teachers with both coding experience and strong metacognitive strategies would be classified into profiles indicative of the highest CT proficiency, as these factors synergistically enhance problem-solving efficiency and strategic application of computational principles. Finally, the third research question about online behaviors and CT skills, it was hypothesized that preservice teachers with higher CT skills would demonstrate more efficient and adaptive online behaviors during CT tasks. This includes spending less time on the playground (PGTime) and on completing the task (L_Time), making fewer clicks (L_Clicks), and perceiving the task as less difficult (L_Diff) compared to those in profiles with moderate or lower CT skills (H3). The latter is based on problem-solving efficiency and transferability acting as key CT markers (Shute et al., 2017 ). Moreover, studies on metacognition (Cheng et al., 2023 ) suggest that skilled learners engage in deliberate planning and reflection, which manifest in observable behaviors like reduced errors and strategic reuse of patterns. Overall, these hypotheses are based on the assumption that higher CT skill levels lead to quicker comprehension and more efficient problem-solving. Enhanced navigation and problem-solving strategies, stemming from greater familiarity and comfort with computational concepts, are considered critical factors. These insights are expected to inform the design of educational interventions aimed at boosting specific CT skills and improving the overall learning experience in teacher-education programs. Results Table 1 shows the means and the correlations of the study variables and provides several key observations based on the correlations and descriptive statistics. INSERT TABLE 1 HERE Experience in coding ( CodExp ) was positively correlated with digital literacy ( CT_DL , r = .22, p < .05) and comfort with coding ( CT_CD , r = .29, p < .01), whereas it was negatively correlated with the perceived difficulty of computational tasks ( L_Diff , r = − .22, p < .05). This indicates that coding experience can not only enhance CT skills but also alleviate any cognitive stress associated with computational tasks. Moreover, the level of perceived difficulty ( L_Diff ) in CT tasks showed strong negative correlations with both problem-solving skills (CT_PS, r = − .31, p < .01) and comfort with coding ( CT_CD , r = − .32, p < .001). These relationships suggest that those with increased levels of CT skills may perceive these tasks as less challenging. Metacognitive strategies ( MetaCog ) also demonstrated a positive correlation with problem-solving skills ( CT_PS , r = .45, p < .001), emphasizing the role of reflective and strategic thinking in effectively tackling complex problems. Furthermore, behavioral metrics during computational tasks such as the number of clicks ( L_Clicks ) and total time spent on tasks ( L_Time ) revealed that a higher number of clicks negatively correlated with performance scores ( L_Score , r = − .51, p < .001). This suggests that more efficient navigation through tasks (i.e., requiring fewer clicks) is associated with higher performance. Lastly, performance on the CT task ( L_Score ) was positively associated with the time spent on the playground ( PG_Time , r = .295, p < .01), highlighting the benefits of training in enhancing performance outcomes. How can preservice teachers be differentiated based on their CT skills? The LPA indicated that AIC and BIC values decreased as the number of profiles increased, indicating improved fit with additional profiles. However, the selection of the optimal model was not based solely on the lowest AIC and BIC values but also considered entropy, profile probability bounds, and the BLRT. Specifically, the three-profile model exhibited a substantial decrease in AIC (995.93) and BIC (1035.85) compared to the two-profile model, coupled with high entropy (.84) indicating clear delineation between profiles. Probabilities of most likely profile membership ranged from .86 to .96, suggesting high classification accuracy. Importantly, the BLRT for the three-profile model ( p = .01) confirmed its statistical superiority over the two-profile model. Models with more than three profiles did not show meaningful improvements in fit or entropy to justify the added complexity. Both the four-profile (entropy = .77) and five-profile (entropy = .76) models exhibited lower entropy values than the three-profile model, indicating diminishing returns in profile distinctiveness and classification accuracy. Moreover, the marginal reductions in AIC and BIC were not accompanied by significant BLRT improvements, suggesting that these models might overfit the data. Figure 1 displays the profile plot with the standardized scores of digital literacy (DL), problem solving (PS), and coding (CD) across the three profiles as proposed by the LPA. INSERT FIGURE 1 HERE Profile 1: Developing (n = 58) Most preservice teachers (7M, 45F, 5 other, and one missing) in Profile 1 were from Elementary (n = 34), followed by Secondary (n = 16), and other programs (n = 8). Before completing the CT task, this group spent 30.14 seconds on the playground, which rendered to 44.8% completing the task on the first attempt, 13.8% on the second attempt, and 25.9% incorrectly completing the task. Profile 1displayed moderate levels of digital literacy ( M = 5.53, SD = 0.74), problem-solving skills ( M = 5.45, SD = 0.46), and coding ( M = 2.49, SD = 0.87), suggesting that participants in this profile were still developing foundational CT skills. Profile 2: Novice (n = 43) Preservice teachers (3M, 39F, and one missing) in Profile 2 were also mostly from Elementary (n = 31), followed by Secondary (n = 8), and other programs (n = 4). This group spent on average 39.93 seconds training on the playground. After completing the CT task, 48.8% correctly completed the task on the first attempt, 9.3% on the second attempt, and 18.6% did not correctly complete the task. Given that Profile 2 yielded below-average levels of digital literacy ( M = 3.88, SD = 0.83), problem-solving skills ( M = 4.40, SD = 0.76), and coding ( M = 1.76, SD = 0.69), participants in this profile were still considered novices and in their initial stages of acquiring CT skills. Profile 3: Proficient (n = 27) Preservice teachers (7M, 17F, and 2 missing) in Profile 3 composed of students in Secondary (n = 13), Elementary (n = 8), and other programs (n = 6) comprised a distinct subgroup characterized by its proficiency in CT skills. This group yielded above-average scores in digital literacy (M = 6.22, SD = 0.76), problem-solving skills (M = 6.14, SD = 0.61), and coding (M = 4.70, SD = 0.95). Profile 3 participants spent on average 70.80 seconds on the playground, with a majority (70.4%) successfully completing the task on their first attempt, 11.1% on the second attempt, whereas 7.4% incorrectly completed the task. The ANOVA results for CT_DL [ F (2,125) = 90.969, p < .001], CT_PS [ F (2,125) = 75.508, p < .001], and CT_CD [ F (2,125) = 107.373, p < .001] provide strong evidence of significant differences in these measures across the three groups. To what extent do preservice teachers’ metacognitive strategies and prior coding experience predict CT skills? A Linear Discriminant Analysis (LDA) was performed to explore how well two predictors, Coding Experience ( CodExp ) and Metacognitive Strategies ( MetaCog ), could discriminate among the three profiles. Descriptive statistics revealed that Group 3 (N = 27) was predominantly more experienced in coding (74%), compared to Group 2 (34.9%) and Group 1 (46.6%). Moreover, Group 1 (M = 5.99, SD = 0.70) and Group 3 (M = 5.99, SD = 0.66) displayed similar means in MetaCog . Finally, Group 2 had a significant lower mean (M = 5.53, SD = 0.77) compared to Group 1 ( p < .004) and Group 3 ( p < .024). Wilks’ lambda indicated significant differences among the groups for both MetaCog [λ = .912, F (2, 125) = 6.068, p = .003] and CodExp [λ = .919, F (2, 125) = 5.499, p = .005] across the groups. Box’s M test confirmed the assumption of equal covariance matrices was not violated, M = 4.747, F(6, 88111.212) = .771, p = .593. The analysis yielded two discriminant functions, with the first function explaining 82.5% of the variance between groups (canonical R² = .358), and the second function explaining an additional 17.5% (canonical R² = .174). Collectively, the discriminant functions significantly differentiated the groups, Wilks’ Lambda = .845, χ²(4) = 20.913, p < .001, with the second function alone approaching significance, χ²(1) = 3.831, p = .050. The standardized canonical discriminant function coefficients for the first function were .716 for MetaCog and .658 for CodExp , whereas for the second function they were − .701 for MetaCog and .755 for CodExp . Group centroids indicated that Group 1 was characterized by scores of .130 on the first function and − .182 on the second function, Group 2 by scores of − .494 on the first and .092 on the second, and Group 3 by scores of .506 on the first and .244 on the second. The analysis also provided classification functions for each group, with the coefficients for CodExp ranging from 4.821 to 6.415 and for MetaCog from 10.672 to 11.526. These functions effectively classified the participants into their respective groups based on the predictors. In what ways do online behaviors on a CT task differ based on preservice teachers’ CT skills? Participants were randomly assigned to different difficulty levels, with 24–30% allocated to each level. A one-way ANOVA was conducted to examine the differences across the three profiles in time spent on playground ( PG_Time ), time spent completing the CT task ( L_Time ), and the number of clicks while completing the CT task ( L_Clicks ), and perceived task difficulty ( L_Diff ). A Kruskal-Wallis H test was also conducted to determine if there were differences in perceived task difficulty (L_Diff) , across the three different profiles. The mean scores and standard deviations for each variable are presented in Table 2. INSERT TABLE 2 HERE The one-way ANOVA revealed that there was a statistically significant difference in L_Time [ F (2, 99) = 4.072, p = .020] across the profiles. No significant differences were found for PG_Time [ F (2, 121) = .891, p = .413] or L_Clicks [ F (2, 99) = 1.692, p = .189]. The Kruskal-Wallis H test also revealed a significant difference in the medians of L_Diff across the profiles, [H(2) = 17.271, p < .001], suggesting that not all groups perceived task difficulty similarly. Post hoc comparisons using the LSD test indicated that the mean score for L_Time for Profile 1 ( M = 131.42, SD = 72.92) was significantly different ( p = .007) than that for Profile 3 ( M = 87.91, SD = 55.44). No significant differences were found between Profiles 1 and 2 ( M = 106.65, SD = 50.79), p = .095, and between Profiles 2 and 3, p = .281. Post-hoc analyses using pairwise comparisons with Bonferroni correction also indicated a significant difference in L_Diff scores (H = 28.804, SE = 7.491, p < .001) between Profiles 1 and 3 and between Profiles 2 and 3 (H = 29.380, SE = 8.118, p < .001). No significant differences were found between Profiles 1 and 2 (H = − .576, SE = 6.833, p = .933). These findings suggest that learner profiles significantly differ in terms of how long they engage with learning tasks and their perceived difficulty, with Profile 3 participants typically finding tasks less challenging and spending less time on them compared to participants belonging to the other profiles. Discussion This study examined the variations in preservice teachers’ computational thinking (CT) skills, the predictive role of prior coding experience and metacognitive strategies on these skills, and how these skills influence their behavior during an online CT task. Our findings extend Shute et al.’s ( 2017 ) CT model by demonstrating that metacognitive strategies and coding experience play a significant role in CT proficiency, a dynamic previously underexplored in preservice teacher education. The identification of three distinct profiles (Developing, Novice, Proficient) provides empirical support for Román-González et al.’s ( 2019 ) assertion that CT skill acquisition is non-linear and context-dependent. This profiling approach is novel in the CT literature and offers a replicable framework for assessing preservice teachers’ readiness to integrate computational principles into pedagogy. The findings revealed significant differentiation among preservice teachers’ CT skills based on a three-profile model identified through Latent Profile Analysis (LPA). The most proficient group in CT, Profile 3, displayed above-average levels of digital literacy, problem-solving, and comfort with coding, which translated into higher performance scores and a lesser perceived difficulty of the CT task, thus corroborating with the hypothesis that higher CT skills enhance problem-solving efficiency (Cheng et al., 2023 ). Specifically, a recent systematic review revealed a close link between CT and problem solving in the 37 studies reviewed (Wu, Asmara, Huang, & Permata Hapsari, 2024 ). This finding supports the notion that well-developed CT skills can significantly ease the cognitive load of problem-solving learning situations, a result consistent with earlier studies highlighting the importance of strong CT skills in reducing perceived task difficulty and enhancing problem-solving efficiency (Barr & Stephenson, 2011 ). Our results also highlighted the critical role of prior coding experience and the use of metacognitive strategies in predicting varying levels of CT skills. The positive correlation between coding experience and CT skills underscores the value of practical coding knowledge in enhancing computational proficiency. This aligns with previous research by Liao et al. ( 2022 ), highlighting the significance of background knowledge in CT development. As hypothesized, preservice teachers with prior coding experience also found the CT task less daunting and displayed higher problem-solving skills, emphasizing the need for integrating practical coding exercises in teacher-education programs. The study’s results also confirm the hypothesis that metacognitive strategies are crucial in navigating CT tasks effectively. The use of these strategies was strongly correlated with better problem-solving skills, echoing the findings of Pintrich et al. ( 1993 ), which emphasize the importance of strategic and reflective thinking in learning environments. This supports the conceptual framework that positions metacognitive engagement as integral to mastering CT tasks. These observations suggest that both technical skills and reflective practices are essential for fostering effective computational thinking, aligning with Shute et al.’s ( 2017 ) emphasis on a pedagogical approach that integrates metacognitive skills into CT education. Moreover, the relationship between time spent on training activities (i.e., the playground, PG_Time) and subsequent performance on a CT task underscores the importance of preparatory engagement in enhancing the application of CT skills. This is consistent with findings in line with educational theories that advocate for experiential learning, in which a positive correlation was demonstrated between prior engagement with relevant learning activities and improved skill application in actual problem-solving contexts (Argelagós et al., 2022 ). This study contributes to the theoretical understanding of computational thinking by validating and extending Shute et al.’s ( 2017 ) model, particularly in the context of preservice teacher education. Our findings highlight the importance of metacognitive strategies as a mediating factor in CT skill development, a dimension that has been underexplored in existing frameworks. By demonstrating the interplay between coding experience and metacognitive engagement, this study bridges the gap between technical proficiency and reflective practice, offering a more holistic view of CT development. Furthermore, the differentiation of preservice teachers into distinct profiles (Developing, Novice, and Proficient) aligns with Román-González et al.’s ( 2019 ) assertion that CT skill acquisition is not uniform but varies based on individual experiences and cognitive strategies. This nuanced understanding of CT development can inform future theoretical models by emphasizing the role of metacognition and prior experience in shaping computational proficiency. While this study offers significant insights, it also reveals important limitations that call for further research. The cross-sectional nature of our data limits our ability to discern the developmental trajectories of CT skills among preservice teachers, which would indicate how these competencies evolve over time. To address these limitations and build on the findings of this study, further research could explore longitudinal studies to track the progression of CT skills over time and the long-term impact of metacognitive strategies and coding experience on these skills. Such studies would provide deeper insights into the developmental aspects of CT skills and help in designing interventions that are more effective over the long term. Additionally, experimental and comparative studies could be conducted to directly measure the effects of specific teaching interventions on CT skills development. These studies would help clarify the causality between teaching methods and CT skills acquisition, which this study suggests but cannot definitively establish due to its methodological constraints. Additionally, while prior studies often rely on self-reports or isolated performance metrics, our mixed-methods design, combining self-assessments, behavioral analytics (e.g., clicks, time-on-task), and task performance provides a holistic view of CT development. This methodological innovation addresses critiques of oversimplified CT assessments (Grover & Pea, 2013 ) and sets a precedent for future research in teacher education. Implications This research makes a theoretical contribution by providing valuable insights into the factors influencing CT skills among preservice teachers. It also makes a practical contribution by offering a basis for enhancing educational practices and curriculum design to better prepare future educators for the increasingly digital world. Lastly, a methodological contribution was presented through the use of detailed log data from online learning platforms, which provided a rich resource for understanding learner interactions in problem-solving situations that require the application of computational thinking skills. This approach aids not only in the immediate improvement of individual learning strategies but also informs broader educational practices and curriculum development. This study leads to multiple implications. First, the integration of CT tasks into preservice teacher education can serve as a valuable diagnostic and developmental tool in educational practices and policy development. The clear delineation in CT skill levels among preservice teachers suggests a need for differentiated instructional strategies that cater to specific needs of each group, thereby enhancing overall CT proficiency before these individuals enter the professional field. These findings also underscore the importance of personalized learning interventions that can be tailored to focus on enhancing coding experience and fostering metacognitive abilities as these proved to play a pivotal role in predicting CT skills. For instance, learners in the novice group might benefit more from targeted support and incremental task levels, which could help in gradually building their confidence and skills. Furthermore, given the demonstrated importance of coding experience and metacognitive strategies, the findings advocate for the integration of structured coding tasks and metacognition training in teacher-education programs. This approach could improve CT skills uniformly across different learner profiles, equipping future educators with the necessary skills not only to tackle digital tasks but also to implement technology-driven pedagogies effectively in their teaching practices. From a practical standpoint, this study underscores the need for teacher education programs to adopt a dual-focused approach that combines technical coding skills with metacognitive training. For instance, programs could incorporate project-based learning activities that require preservice teachers to reflect on their problem-solving processes, such as journaling or peer feedback sessions. Tools like Scratch or robotics kits could be used to provide hands-on coding experience, while metacognitive strategies could be reinforced through guided reflection prompts or self-assessment checklists. Additionally, the findings suggest the value of adaptive learning technologies that tailor tasks to individual skill levels, ensuring that novice learners receive the scaffolding they need while proficient learners are challenged with more complex problems. Policymakers could support these efforts by allocating resources for professional development workshops focused on CT integration and by incentivizing the adoption of evidence-based practices in teacher education curricula and ensuring preservice teachers graduate equipped to model CT across disciplines. In the context of 21st-century education, where computational thinking is increasingly recognized as a core competency, this study highlights the critical role of teacher preparation in meeting the demands of a technology-driven world. As K-12 curricula worldwide begin to integrate CT into subjects beyond computer science, preservice teachers must be equipped not only with technical skills but also with the ability to foster these skills in their students. Our findings suggest that teacher education programs should prioritize the development of both coding proficiency and metacognitive strategies, as these are essential for enabling future educators to model and teach CT effectively. By doing so, preservice teachers will be better prepared to create learning environments that encourage algorithmic thinking, abstraction, and problem-solving across disciplines, ultimately contributing to a more computationally literate society. While this study provides valuable insights, it is not without limitations. First, the reliance on self-reported data for CT skills and metacognitive strategies may introduce bias, as participants’ perceptions may not always align with their actual abilities. Future studies could complement self-reports with performance-based assessments to provide a more objective measure of CT proficiency. Second, the study was conducted within a single institution, which may limit the generalizability of the findings. Replicating this study in diverse educational contexts, including different countries or regions, would help determine the extent to which these findings apply across settings. Finally, the cross-sectional design of the study precludes causal inferences about the relationship between coding experience, metacognitive strategies, and CT skills. Longitudinal studies tracking preservice teachers’ CT development over time could provide deeper insights into how these factors interact and evolve. Additionally, experimental studies could test the effectiveness of specific interventions, such as metacognitive training modules or coding bootcamps, in enhancing CT skills. Exploring the role of cultural and contextual factors in CT development could also yield valuable insights, particularly in non-Western educational systems where CT integration may be at an earlier stage. This study advances the field by demonstrating that CT proficiency in preservice teachers is not merely a function of technical skill but a cognitive-metacognitive interplay. By linking coding experience to reduced task difficulty and metacognition to enhanced problem-solving, we provide a roadmap for reimagining teacher education. These insights are timely, as global K-12 reforms increasingly demand educators who can seamlessly integrate CT into diverse subjects, a competency our findings show is achievable through targeted, evidence-based training In conclusion, this study highlights the critical role of prior coding experience and metacognitive strategies in shaping preservice teachers’ computational thinking skills. By identifying distinct profiles of CT proficiency and demonstrating the importance of preparatory engagement and reflective practice, this research provides a foundation for designing more effective teacher education programs. As the demand for computational thinking continues to grow in K-12 education, equipping preservice teachers with the skills and strategies to teach CT effectively will be essential for preparing students to thrive in a technology-driven world. Ultimately, this study not only advances our understanding of CT development but also underscores the importance of fostering both technical and cognitive skills in future educators, paving the way for a more computationally literate and innovative society. Declarations Acknowledgments We extend our gratitude to [name blinded], a graduate student, for her invaluable assistance with data collection during our classroom visits. We also thank the three college students, [names blinded] whose dedication and creativity in designing the computational thinking tasks were instrumental to the success of this project. References Argelagós E, Garcia C, Privado J, Wopereis I (2022) Fostering information problem solving skills through online task-centred instruction in higher education. Comput Educ 180:104433. https://doi.org/10.1016/j.compedu.2022.104433 Author 1 Blinded for review Author 2 Blinded for review Author 3. Blinded for review Barr V, Stephenson C (2011) Bringing computational thinking to K-12: What is involved and what is the role of the computer science education community? ACM Inroads 2(1):48–54. https://doi.org/10.1145/1929887.1929905 Bers MU (2018) Coding as playground: Programming and computational thinking in the early childhood classroom. Routledge, London, UK Bower M, Falkner K (2015), January Computational thinking, the notional machine, pre-service teachers, and research opportunities. In Proceedings of the 17th Australasian computing education conference (ACE 2015) (Vol. 27, p. 30). Australian Computer Society Brennan K, Resnick M (2012), April New frameworks for studying and assessing the development of computational thinking. In Proceedings of the 2012 annual meeting of the American educational research association, Vancouver, Canada (Vol. 1, p. 25) Butler D, Leahy M (2021) Developing preservice teachers’ understanding of computational thinking: A constructionist approach. Br J Edu Technol 52(3):1060–1077. https://doi.org/10.1111/bjet.13090 Chan SW, Looi CK, Ho WK, Huang W, Seow P, Wu L (2021) Learning number patterns through computational thinking activities: A Rasch model analysis. Heliyon 7(9):e07922. https://doi.org/10.1016/j.heliyon.2021.e07922 Cheng Y-P, Lai C-F, Chen Y-T, Wang W-S, Huang Y-M, Wu T-T (2023) Enhancing student's computational thinking skills with student-generated questions strategy in a game-based learning platform. Comput Educ 200:1–20. https://doi.org/10.1016/j.compedu.2023.104794 Dagienė V, Futschek G (2008) Bebras international contest on informatics and computer literacy: Criteria for good tasks. In: Mittermeir RT, Sysło MM (eds) Informatics Education - Supporting Computational Thinking . ISSEP 2008. Lecture Notes in Computer Science, vol 5090. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-69924-8_2 Dong W, Li Y, Sun L, Liu Y (2024) Developing pre-service teachers’ computational thinking: A systematic literature review. Int J Technol Des Educ 34:191–127. https://doi.org/10.1007/s10798-023-09811-3 Flavell JH (1979) Metacognition and cognitive monitoring: A new area of cognitive–developmental inquiry. Am Psychol 34(10):906–911. https://doi.org/10.1037/0003-066X.34.10.906 Grover S, Pea R (2013) Computational thinking in K-12: A review of the state of the field. Educational Researcher 42(1):38–43. https://doi.org/10.3102/0013189X12463051 Hoyles C, Noss R A computational lens on design research. ZDM, 47 (6), 1039–1045. https://doi.org/10.1007/s11858-015-0731-2 Jaipal-Jamani K, Angeli C (2017) Effect of robotics on elementary preservice teachers’ self-efficacy, science learning, and computational thinking. J Sci Edu Technol 26(2):175–192. https://doi.org/10.1007/s10956-016-9663-z Kotsopoulos D, Floyd L, Khan S, Namukasa IK, Somanath S, Weber J, Yiu C (2017) A pedagogical framework for computational thinking. Digit Experiences Math Educ 3(2):154–171. https://doi.org/10.1007/s40751-017-0031-2 Liao CH, Chiang CT, Chen IC, Parker KR (2022) Exploring the relationship between computational thinking and learning satisfaction for non-STEM college students. Int J Educational Technol High Educ 19(1):43. https://doi.org/10.1186/s41239-022-00347-5 Liu Z, Gearty Z, Richard E et al (2024) Bringing computational thinking into classrooms: a systematic review on supporting teachers in integrating computational thinking into K-12 classrooms. Int J STEM Educ 11:51, 1–18. https://doi.org/10.1186/s40594-024-00510-6 Mishra P, Koehler MJ (2006) Technological pedagogical content knowledge: A framework for teacher knowledge. Teachers Coll Record 108(6):1017–1054. https://doi.org/10.1111/j.1467-9620.2006.00684.x Mouza C, Yang H, Pan Y-C, Ozden SY, Pollock L (2017) Resetting educational technology coursework for pre-service teachers: A computational thinking approach to the development of technological pedagogical content knowledge (TPACK). Australasian J Educational Technol 33(3). https://doi.org/10.14742/ajet.3521 Pintrich PR, Smith DAF, Garcia T, McKeachie WJ (1993) Reliability and predictive validity of the Motivated Strategies for Learning Questionnaire (Mslq). Educ Psychol Meas 53(3):801–813. https://doi.org/10.1177/0013164493053003024 Román-González M, Moreno-León J, Robles G (2019) Combining assessment tools for a comprehensive evaluation of computational thinking interventions. In: Kong SC, Abelson H (eds) Computational Thinking Education. Springer, Singapore. https://doi.org/10.1007/978-981-13-6528-7_6 Sadik O, Ottenbreit-Leftwich AT, Nadiruzzaman H (2017) Computational thinking conceptions and misconceptions: Progression of preservice teacher thinking during computer science lesson planning. In: Rich P, Hodges C (eds) Emerging Research, Practice, and Policy on Computational Thinking. Educational Communications and Technology: Issues and Innovations. Springer, Cham. https://doi.org/10.1007/978-3-319-52691-1_14 Shodiev H (2015) Computational thinking and simulation in teaching science and mathematics. In M. G. Cojocaru, I. S. Kotsireas, R. N. Makarov, R. V. N. Melnik, & H. Shodiev (Eds.), Interdisciplinary Topics in Applied Mathematics, Modeling and Computational Science (pp. 405–410). Springer. https://doi.org/10.1007/978-3-319-12307-3_58 Shute VJ, Sun C, Asbell-Clarke J (2017) Demystifying computational thinking. Educational Res Rev 22:142–158. https://doi.org/10.1016/j.edurev.2017.09.003 Tang X, Yin Y, Lin Q, Hadad R, Zhai X (2020) Assessing computational thinking: A systematic review of empirical studies. Comput Educ 148:103798. https://doi.org/10.1016/j.compedu. 2019.103798 Weintrop D, Beheshti E, Horn M, Orton K, Jona K, Trouille L, Wilensky U (2016) Defining computational thinking for mathematics and science classrooms. J Sci Edu Technol 25(1):127–147. https://doi.org/10.1007/s10956-015-9581-5 Wing JM (2006) Computational thinking. Commun ACM 49(3):33–35. https://doi.org/10.1145/1118178.1118215 Wu TT, Asmara A, Huang YM, Permata Hapsari I (2024) Identification of problem-solving techniques in computational thinking studies: Systematic literature review. SAGE Open 14(2). https://doi.org/10.1177/21582440241249897 Tables Table 1 Means and Standard Deviations for Online Behaviors Variable Profile Mean SD 95% CI for Mean PG_Time (in seconds) 1 62.28 30.41 [54.13, 70.42] 2 60.15 38.75 [47.92, 72.38] 3 70.80 30.11 [58.89, 82.71] L_Time (in seconds) 1 131.42 72.92 [110.24, 152.59] 2 106.65 50.79 [87.69, 125.62] 3 87.91 55.44 [64.50, 111.31] L_Clicks 1 49.81 27.13 [41.93, 57.69] 2 51.70 28.07 [41.22, 62.18] 3 38.75 28.80 [26.59, 50.91] L_Diff (on a scale of 10) 1 5.42 3.21 [4.49, 6.35] 2 5.90 3.60 [4.55, 7.25] 3 2.75 2.82 [1.56, 3.94] PG_Time: Time spent on playground L_Time: Time spent completing the task (Programming Lamps) L_Clicks: Number of clicks while completing the task (Programming Lamps) L_Diff: Perceived level of task difficulty Table 2 Means and Standard Deviations for Online Behaviors Variable Profile Mean SD 95% CI for Mean PG_Time (in seconds) 1 62.28 30.41 [54.13, 70.42] 2 60.15 38.75 [47.92, 72.38] 3 70.80 30.11 [58.89, 82.71] L_Time (in seconds) 1 131.42 72.92 [110.24, 152.59] 2 106.65 50.79 [87.69, 125.62] 3 87.91 55.44 [64.50, 111.31] L_Clicks 1 49.81 27.13 [41.93, 57.69] 2 51.70 28.07 [41.22, 62.18] 3 38.75 28.80 [26.59, 50.91] L_Diff (on a scale of 10) 1 5.42 3.21 [4.49, 6.35] 2 5.90 3.60 [4.55, 7.25] 3 2.75 2.82 [1.56, 3.94] PG_Time: Time spent on playground L_Time: Time spent completing the task (Programming Lamps) L_Clicks: Number of clicks while completing the task (Programming Lamps) L_Diff: Perceived level of task difficulty Additional Declarations The authors declare no competing interests. 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-6283533","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":432457795,"identity":"6d0a6957-e965-4110-848f-eed68c833207","order_by":0,"name":"Tanya Chichekian","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0002-6156-0779","institution":"Université de Sherbrooke","correspondingAuthor":true,"prefix":"","firstName":"Tanya","middleName":"","lastName":"Chichekian","suffix":""},{"id":432457796,"identity":"ed6c7802-be23-46fa-90d3-e7221877321e","order_by":1,"name":"Marian Cutumisu","email":"","orcid":"https://orcid.org/0000-0003-2475-9647","institution":"McGill University","correspondingAuthor":false,"prefix":"","firstName":"Marian","middleName":"","lastName":"Cutumisu","suffix":""},{"id":432457797,"identity":"2f574460-8209-42f7-97ba-89d112c6fca2","order_by":2,"name":"Annie Savard","email":"","orcid":"https://orcid.org/0000-0002-0285-0952","institution":"McGill University","correspondingAuthor":false,"prefix":"","firstName":"Annie","middleName":"","lastName":"Savard","suffix":""}],"badges":[],"createdAt":"2025-03-22 12:10:28","currentVersionCode":1,"declarations":{"humanSubjects":true,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":true,"humanSubjectConsent":true,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-6283533/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6283533/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":79160311,"identity":"ae81fe7d-ef0c-4dfc-a464-161df8267d23","added_by":"auto","created_at":"2025-03-25 07:18:01","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":23282,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6283533/v1/1840c9cbac6896f9c74092b0.jpg"},{"id":79161658,"identity":"a39a8e45-da07-4b74-acfa-ed25b2de07ca","added_by":"auto","created_at":"2025-03-25 07:34:06","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":900195,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6283533/v1/75bc35ee-037c-439d-9082-0ca99b22905d.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eProfiling Preservice Teachers’ Computational Thinking: The Role of Metacognition and Coding Experience\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eIn our increasingly technology-oriented society, computational thinking (CT) has become a critical skill for engaging with digital innovation (Kotsopoulos et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). This cognitive ability is essential for solving complex and novel problems (Barr \u0026amp; Stephenson, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). According to Wing (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), CT is a way to comprehend human behavior through computing principles. It can also be seen as a way of thinking algorithmically, drawing from computer science principles as a guiding framework, at times metaphorically (Shodiev, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Hoyles and Noss (2015) define CT as including skills such as abstraction (viewing problems at varying levels of detail), algorithmic thinking (expressing problems as a series of step-by-step instructions), decomposition (tackling a problem by solving smaller constituent issues), and pattern recognition (identifying connections between a new problem and previous ones). While traditionally associated with computer science, CT is a versatile skillset that transcends disciplines, enhancing problem-solving capabilities and analytical skills over time (Barr \u0026amp; Stephenson, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Therefore, understanding how pedagogical experiences contribute to developing students\u0026rsquo; CT skills is imperative in higher education. This knowledge provides valuable insights into shaping students\u0026rsquo; perceptions and attitudes towards CT, with the potential to facilitate the integration of these skills into various subjects across different educational levels (Author 1).\u003c/p\u003e \u003cp\u003eIn an era where technology is omnipresent, developing CT skills among future educators has increasingly become a focal point. The significance of these skills for preservice teachers is a rapidly expanding field of interest. Despite this, research on the factors influencing these skills (Author 2) among preservice teachers remains limited. While prior research has explored CT development in K-12 students or in-service teachers, this study addresses a critical gap by focusing on preservice teachers, a population uniquely positioned to shape future generations\u0026rsquo; computational literacy. By investigating the interplay between metacognitive strategies and coding experience, this study advances understanding of how these factors synergistically influence CT skill acquisition. Furthermore, it employs latent profile analysis (LPA) to categorize preservice teachers into distinct CT proficiency profiles, offering a novel framework for designing differentiated pedagogical interventions in teacher education programs. Specifically, this study seeks to answer the following research questions: (1) How can preservice teachers be differentiated based on their CT skills? (2) To what extent do preservice teachers\u0026rsquo; metacognitive strategies and prior coding experience predict CT skills? (3) In what ways do online behaviors on a CT interactive task differ based on preservice teachers\u0026rsquo; CT skills?\u003c/p\u003e"},{"header":"Literature Review","content":"\u003cp\u003eOver the past decade, various researchers have explored different models for computational thinking (CT), with an emphasis on aspects such as concepts, practices, and perspectives (Barr \u0026amp; Stephenson, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Weintrop et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). From an educational standpoint, an effective CT model should transcend mere coding and programming language knowledge, prioritizing pedagogical approaches to teaching CT through specific classroom activities. For instance, Shute et al. (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) have proposed a CT model that represents a spectrum of increasing levels of foundational understanding and CT skills. This conceptual foundation is particularly relevant for educational experiences that aim to design problem-solving scenarios with solutions that can be applied in different contexts. According to Shute et al. (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), CT involves the ability to solve problems effectively and efficiently, whether algorithmically with or without computer assistance, and to create solutions that can be reused in various situations. This definition underscores that CT is fundamentally a mindset and a way of acting, which can be demonstrated through specific skills that can serve as the basis for evaluating CT abilities through performance-based assessments.\u003c/p\u003e \u003cp\u003eThe development of CT skills in educational settings has been approached through both unplugged (non-digital) and plugged (digital) methods. Unplugged methods, which do not require the use of computers, focus on the conceptual underpinnings of computing and have been particularly effective in making abstract concepts accessible to beginners (Kotsopoulos et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). These methods often involve hands-on activities, such as puzzles, games, and physical manipulatives that help learners grasp foundational concepts like algorithms, decomposition, and pattern recognition without the need for digital tools.\u003c/p\u003e \u003cp\u003eConversely, plugged methods involve direct interaction with technology, offering a more hands-on approach to developing computational skills. These methods are crucial for understanding how to apply computational concepts in real-world contexts, particularly within programming environments (Shute et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) where students can experiment with coding languages and computational tools. For example, tools like Scratch and robotics kits have been widely used to teach CT through interactive, project-based learning (Jaipal-Jamani \u0026amp; Angeli, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Mouza et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Research comparing these approaches (Rom\u0026aacute;n-Gonz\u0026aacute;lez et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) suggests that a blend of both may be most effective, with unplugged activities building foundational understanding and plugged activities solidifying practical application. This shift from unplugged to plugged is particularly pertinent in teacher education, where developing CT capabilities is seen not only as a skill but as a fundamental teaching competency.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eChallenges in Enhancing Preservice Teachers\u0026rsquo; CT Skills\u003c/h2\u003e \u003cp\u003eDespite the growing recognition of CT's importance, there remains a significant disparity in the preparedness of preservice teachers to teach CT skills. Research indicates that many preservice teachers enter their programs with minimal exposure to computational concepts, leading to a steep learning curve and uneven skill development across cohorts (Bower \u0026amp; Falkner, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Sadik, 2017). This lack of preparedness is compounded by the fact that teacher education programs often lack structured CT-focused coursework, leaving many preservice teachers ill-equipped to integrate CT into their future classrooms (Grover \u0026amp; Pea, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Tang et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Moreover, the design and execution of CT activities can significantly impact preservice teachers' comprehension and skill development. For instance, some preservice teachers may develop misconceptions about CT, particularly if the activities are not aligned with their prior knowledge or if the instructional strategies fail to address their individual learning needs (Sadik, 2017). Common misconceptions among preservice teachers include viewing CT as solely programming-related or believing that it is only relevant to STEM subjects. Some may struggle with the abstract nature of CT concepts or lack confidence in their ability to teach these skills (Bers, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAddressing these misconceptions may require explicit instruction on how to effectively design and implement CT-focused interventions in teacher education programs. For example, Jaipal-Jamani and Angeli (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) found that a robotics course significantly improved preservice teachers' self-efficacy, science learning, and CT skills. Similarly, Mouza et al. (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) reported significant improvements in preservice teachers' ability to think computationally after integrating CT into a 15-week course using tools like Scratch. While past studies advocate how various types of interventions have proven effective in engaging preservice teachers and fostering CT skills (Brennan \u0026amp; Resnick, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2012\u003c/span\u003e) through Block-based programming tools like Scratch and educational robotics kits, the key is to provide structured activities that progressively build CT concepts and encourage creative problem-solving. Constructionist learning environments, which encourage learners to construct their own understanding through active exploration, have also been shown to be effective in enriching preservice teachers' CT skills (Butler \u0026amp; Leahy, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). These environments provide opportunities for preservice teachers to engage in hands-on, project-based learning, which can enhance their ability to apply computational concepts in real-world contexts. However, the design of these interventions must be carefully tailored to the needs of preservice teachers, taking into account their prior knowledge, especially in coding, and learning strategies.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eThe Role of Metacognitive Strategies and Prior Coding Experience on CT Skills\u003c/h3\u003e\n\u003cp\u003eMetacognitive strategies, which involve self-regulation and reflective thinking, have been shown to play a crucial role in the development of CT skills, particularly in teacher preparation programs where the dual goal is to develop both CT skills and the ability to teach these skills effectively. According to Pintrich et al. (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e1993\u003c/span\u003e), metacognitive strategies help learners adapt and apply computational methods in problem-solving scenarios, leading to deeper and more transferable learning outcomes. In the context of CT, metacognitive strategies can enhance preservice teachers' ability to plan, monitor, and evaluate their problem-solving processes, thereby improving their overall computational proficiency (Cheng et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Recent studies have also highlighted the importance of integrating metacognitive training into CT-focused interventions. For example, Argelag\u0026oacute;s et al. (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) found that preservice teachers who engaged in metacognitive reflection during CT tasks demonstrated significant improvements in their problem-solving skills and ability to transfer these skills to new contexts.\u003c/p\u003e \u003cp\u003ePrior coding experience has also been identified as a significant predictor of CT skills among preservice teachers. Research by Chan et al. (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) found that preservice teachers with prior coding experience demonstrated higher levels of digital literacy, problem-solving skills, and comfort with coding compared to their peers without such experience. This finding aligns with the broader literature on CT, which emphasizes the importance of practical coding knowledge in enhancing computational proficiency (Grover \u0026amp; Pea, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Shute et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). However, the relationship between prior coding experience and CT skills is not always straightforward. While coding experience can provide a foundation for understanding computational concepts, it does not guarantee the development of higher-order CT skills, such as abstraction and algorithmic thinking (Bower \u0026amp; Falkner, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). This suggests that teacher education programs must go beyond teaching coding languages and focus on fostering a deeper understanding of computational principles and their application in diverse contexts.\u003c/p\u003e \u003cp\u003eRecent systematic reviews (e.g., Dong et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Liu et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) highlight a paucity of research on how preservice teachers develop CT skills, particularly through the lens of metacognitive reflection and prior coding exposure. While studies have examined CT components in isolation (e.g., coding drills or problem-solving frameworks), few have explored how these elements interact within preservice teacher populations. Liao et al. (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) found that prior coding experience positively correlates with CT skills, suggesting that familiarity with programming languages can facilitate the development of computational thinking. However, the interplay between coding experience and metacognitive strategies remains unclear. It is possible that preservice teachers with prior coding experience may rely more on metacognitive strategies to navigate complex problems, but this hypothesis has yet to be tested empirically. This gap is significant, as preservice teachers must not only master CT themselves but also learn to teach it, a dual challenge requiring both technical proficiency and reflective pedagogical strategies. Our study bridges this gap by situating metacognition and coding experience as interdependent predictors of CT, aligning with calls for more nuanced investigations into teacher-specific CT development pathways (Tang et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\n\u003ch3\u003eThe present research\u003c/h3\u003e\n\u003cp\u003eThis study extends the application of Shute et al.\u0026rsquo;s (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) model by focusing on preservice teachers, a group for whom the integration of CT skills is crucial yet understudied. Preservice teachers not only need to develop their own CT skills but also must be prepared to teach these skills, making their mastery doubly significant. This dual need aligns with Hoyles and Noss's (2015) perspective that CT should be embedded in pedagogy, not just as a subject to be taught but as a method of teaching and learning. Furthermore, the notion that CT can enhance problem-solving capabilities beyond computer science (Barr \u0026amp; Stephenson, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) provides a theoretical basis for investigating CT across educational disciplines. This is particularly important in teacher education, where these skills can significantly impact how future educators integrate technology into their teaching practices.\u003c/p\u003e \u003cp\u003eThis study also examines the roles of metacognitive strategies and previous coding experiences in shaping CT skills among preservice teachers. Metacognitive strategies are hypothesized to enhance the acquisition of CT skills as they help learners adapt and apply computational methods in problem-solving scenarios, potentially leading to deeper and more transferable learning outcomes (Pintrich et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e1993\u003c/span\u003e). Moreover, prior coding experience not only fosters familiarity with computational tools and languages but is considered as a practical component of CT skills development. As a response to calls for more empirical research on how specific educational backgrounds influence CT (Liao et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), this study will also examine how preservice teachers\u0026rsquo; prior coding knowledge interacts with their metacognitive strategies to predict a set of CT skills. The interaction between metacognition and prior coding experience is particularly relevant because individuals with coding backgrounds may be better equipped to leverage metacognitive strategies when problem-solving in computational contexts. Research in cognitive psychology suggests that metacognitive skills, such as planning, monitoring, and evaluating one's own thinking processes, are crucial for effective learning and problem-solving (Flavell, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e1979\u003c/span\u003e). By investigating how these factors interplay, this study contributes to the field by providing empirical evidence on how these factors interact to influence CT proficiency and online problem-solving behaviors, thereby potentially informing the design of more effective teacher education programs.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eParticipants\u003c/h2\u003e \u003cp\u003eThis quantitative study was conducted with 128 preservice teachers (17 males, 101 females, 8 other, and 2 missing) from a Faculty of Education. There were \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;19 participants in Year 1, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;9 in Year 2, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;8 in Year 3, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;73 in Year 4, and \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;19 in Year 4+. Participants were recruited from various programs: \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;73 from Elementary Education, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;37 from Secondary Education, and \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;18 from Physical Education, TESL, or Music. Finally, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;62 had some familiarity with a coding language, such as Scratch, Python, C++, JavaScript, and HTML.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eData Collection\u003c/h2\u003e \u003cp\u003e A research assistant visited classes whose professors had agreed upon to explain the project and asked participants\u0026rsquo; consent before filling out an online survey on Qualtrics. Participants responded to Likert-type scales and were asked to complete a CT interactive task in which they were randomly placed in by level of difficulty.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eCT Interactive Task Design\u003c/h3\u003e\n\u003cp\u003eThe CT task named, \u003cem\u003eProgramming Lamps\u003c/em\u003e was derived from one of the Bebras tasks found in the Bebras international contest (Dagiene \u0026amp; Futschek, 2008) which consists of a set of activities requiring students to transfer and project their CT skills to solve \u0026ldquo;real-life\u0026rdquo; problems. Given that these tasks do not require any prior knowledge of any software or hardware, they could be administered to individuals without any prior programming experience.\u003c/p\u003e \u003cp\u003eThree computer science interns worked on optimizing a series of CT tasks (including \u003cem\u003eProgramming Lamps\u003c/em\u003e) by adding difficulty levels (i.e., easy, medium, hard) with incremental complexity, an animation of the participants\u0026rsquo; responses as a form of feedback, as well as an animation of the correct solution. A first pilot of the optimized tasks was tested with close friends and colleagues to ask for feedback and refine the prototype including the three difficulty levels of \u003cem\u003eProgramming Lamps\u003c/em\u003e. Feedback questions focused on the user experience, namely ease of understanding regarding the description and goal of the task, text and response field formatting, and picture quality. Suggestions on how to improve user experience were also part of the feedback that was sought. A second pilot test was conducted with a student population from an online crowdsourcing platform to validate the optimized interactive tasks with other forms of combined assessment (e.g., surveys on CT skills).\u003c/p\u003e \u003cp\u003eThese optimized CT tasks also easily catered to the implementation of learning analytics. For example, programming for click counts from the beginning to the end of task completion was embedded as a form of data-mining assessment and a playground was added before beginning the actual CT task as a form of formative-iterative assessment. Automated feedback was also incorporated by providing students with two occasions to correctly complete the actual task. After submitting their response, participants received a replay of their solution as feedback if their submitted response was incorrect before continuing with a second attempt.\u003c/p\u003e\n\u003ch3\u003eMeasures\u003c/h3\u003e\n\u003cp\u003eDemographic information was collected from the participants, such as gender, program, year of study, and prior coding experience.\u003c/p\u003e \u003cp\u003e \u003cb\u003eComputational thinking (CT) skills.\u003c/b\u003e The CT skills were adapted from the Callysto Computational Thinking test \u0026ndash; student version (Author 3). We specifically asked preservice teachers to rate their agreement on a scale of 1\u0026ndash;7 their perceptions regarding their CT skills, namely in relation to digital technology (DL, 3 items, α\u0026thinsp;=\u0026thinsp;.77), problem-solving (PS, 3 items, α\u0026thinsp;=\u0026thinsp;.77), and comfort with coding (CD, 3 items, α\u0026thinsp;=\u0026thinsp;.73) for this study.\u003c/p\u003e \u003cp\u003e \u003cb\u003eCT task score and online learning behaviors.\u003c/b\u003e The optimized CT task \u003cem\u003eProgramming Lamps\u003c/em\u003e rendered either a score (\u003cem\u003eL_Score\u003c/em\u003e) of \u0026ldquo;1\u0026rdquo; when the participants did not get the right answer on both attempts, a \u0026ldquo;2\u0026rdquo; when the participants answered correctly on the second attempt after viewing automated feedback (i.e., a replay of their wrong solution), or a \u0026ldquo;3\u0026rdquo; when the participants reached the correct answer on the first attempt. The time spent on playground (\u003cem\u003ePG_Time\u003c/em\u003e), the number of clicks \u003cem\u003e(L_clicks)\u003c/em\u003e, and time spent on completing the task (\u003cem\u003eL_Time\u003c/em\u003e) were also recorded.\u003c/p\u003e \u003cp\u003e \u003cb\u003ePerceptions of difficulty level\u003c/b\u003e \u003cb\u003e(L_diff).\u003c/b\u003e At the end of the task, preservice teachers were asked to rate from 1 (very easy) to 10 (very hard), the difficulty level they associated with the task.\u003c/p\u003e \u003cp\u003e \u003cb\u003eMetacognitive learning strategies\u003c/b\u003e \u003cb\u003e(MetaCog).\u003c/b\u003e Preservice teachers reported the use of different metacognitive strategies (Pintrich et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e1993\u003c/span\u003e) when preparing a lesson plan for teaching a complex concept (9 items, α\u0026thinsp;=\u0026thinsp;.80)\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eData Analysis\u003c/h2\u003e \u003cp\u003eLatent profile analysis (LPA) was employed to explore the underlying profiles within the data based on preservice teachers\u0026rsquo; responses regarding their CT skills (DL-Digital Literacy, PS-Problem Solving, and CD-Comfort in Coding). We began by standardizing these variables, followed by an estimation of a series of models with one to five profiles to determine the optimal number of profiles, using fit indices such as the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), entropy, and the Bootstrap Likelihood Ratio Test (BLRT).\u003c/p\u003e \u003cp\u003eTo determine the extent to which metacognitive strategies and prior coding experience predicted group membership, a linear discriminant analysis (LDA) was conducted. LDA is particularly useful for smaller datasets as it is designed to model differences in groups by assuming Gaussian distributions and can provide insights into which behaviors are most discriminative between profiles.\u003c/p\u003e \u003cp\u003eTo examine how outcomes related to online behaviors (i.e., \u003cem\u003ePG_Time, L_Time, L_Clicks, and L_Diff\u003c/em\u003e) varied based on preservice teachers\u0026rsquo; CT skills, a one-way ANOVA was conducted across the three profiles, followed by post hoc LSD comparisons.\u003c/p\u003e \u003cp\u003eBased on the first research question regarding the differentiation of preservice teachers based on CT skills, it was hypothesized that preservice teachers will exhibit distinct profiles characterized by clusters of high, moderate, and low proficiency in CT components (H1a) and that they will differ significantly from one another (H1b). More specifically, it was expected that preservice teachers with prior coding experience (e.g., familiarity with Scratch, Python, C++, JavaScript, HTML) will report higher levels of CT skills (Digital Literacy (DL), Problem Solving (PS), and Comfort with Coding (CD)) than those without such experience. This hypothesis aligns with Shute et al.\u0026rsquo;s (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) CT model, which posits that learners progress through increasing levels of proficiency. Given that CT is a multidimensional construct (Hoyles \u0026amp; Noss, 2015) this hypothesis also parallels Rom\u0026aacute;n-Gonz\u0026aacute;lez et al.\u0026rsquo;s (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) suggestion concerning the variability in CT skill acquisition, particularly among novices.\u003c/p\u003e \u003cp\u003eRegarding the second research question about the predictors of CT skills, it was hypothesized that prior coding experience and metacognitive strategies would independently predict preservice teachers\u0026rsquo; CT skills. Specifically, preservice teachers with prior coding experience (e.g., familiarity with Scratch, Python, or HTML) were expected to exhibit stronger CT proficiency, as coding provides foundational exposure to algorithmic thinking and problem-solving (Chan et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Similarly, those using more sophisticated metacognitive strategies were anticipated to demonstrate higher CT skills, as metacognition enables adaptive learning and reflection (Pintrich et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e1993\u003c/span\u003e; Argelag\u0026oacute;s et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) (H2a). It was also hypothesized that the combination of prior coding experience and metacognitive strategies would most effectively differentiate preservice teachers\u0026rsquo; CT proficiency profiles (H2b). Drawing on cognitive psychology frameworks (Flavell, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e1979\u003c/span\u003e), we posited that preservice teachers with both coding experience and strong metacognitive strategies would be classified into profiles indicative of the highest CT proficiency, as these factors synergistically enhance problem-solving efficiency and strategic application of computational principles.\u003c/p\u003e \u003cp\u003eFinally, the third research question about online behaviors and CT skills, it was hypothesized that preservice teachers with higher CT skills would demonstrate more efficient and adaptive online behaviors during CT tasks. This includes spending less time on the playground (PGTime) and on completing the task (L_Time), making fewer clicks (L_Clicks), and perceiving the task as less difficult (L_Diff) compared to those in profiles with moderate or lower CT skills (H3). The latter is based on problem-solving efficiency and transferability acting as key CT markers (Shute et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Moreover, studies on metacognition (Cheng et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) suggest that skilled learners engage in deliberate planning and reflection, which manifest in observable behaviors like reduced errors and strategic reuse of patterns.\u003c/p\u003e \u003cp\u003eOverall, these hypotheses are based on the assumption that higher CT skill levels lead to quicker comprehension and more efficient problem-solving. Enhanced navigation and problem-solving strategies, stemming from greater familiarity and comfort with computational concepts, are considered critical factors. These insights are expected to inform the design of educational interventions aimed at boosting specific CT skills and improving the overall learning experience in teacher-education programs.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003eTable\u0026nbsp;1 shows the means and the correlations of the study variables and provides several key observations based on the correlations and descriptive statistics.\u003c/p\u003e \u003cp\u003eINSERT TABLE 1 HERE\u003c/p\u003e \u003cp\u003eExperience in coding (\u003cem\u003eCodExp\u003c/em\u003e) was positively correlated with digital literacy (\u003cem\u003eCT_DL\u003c/em\u003e, r\u0026thinsp;=\u0026thinsp;.22, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.05) and comfort with coding (\u003cem\u003eCT_CD\u003c/em\u003e, r\u0026thinsp;=\u0026thinsp;.29, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.01), whereas it was negatively correlated with the perceived difficulty of computational tasks (\u003cem\u003eL_Diff\u003c/em\u003e, r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;.22, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.05). This indicates that coding experience can not only enhance CT skills but also alleviate any cognitive stress associated with computational tasks. Moreover, the level of perceived difficulty (\u003cem\u003eL_Diff\u003c/em\u003e) in CT tasks showed strong negative correlations with both problem-solving skills (CT_PS, r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;.31, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.01) and comfort with coding (\u003cem\u003eCT_CD\u003c/em\u003e, r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;.32, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001). These relationships suggest that those with increased levels of CT skills may perceive these tasks as less challenging. Metacognitive strategies (\u003cem\u003eMetaCog\u003c/em\u003e) also demonstrated a positive correlation with problem-solving skills (\u003cem\u003eCT_PS\u003c/em\u003e, r\u0026thinsp;=\u0026thinsp;.45, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001), emphasizing the role of reflective and strategic thinking in effectively tackling complex problems.\u003c/p\u003e \u003cp\u003eFurthermore, behavioral metrics during computational tasks such as the number of clicks (\u003cem\u003eL_Clicks\u003c/em\u003e) and total time spent on tasks (\u003cem\u003eL_Time\u003c/em\u003e) revealed that a higher number of clicks negatively correlated with performance scores (\u003cem\u003eL_Score\u003c/em\u003e, r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;.51, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001). This suggests that more efficient navigation through tasks (i.e., requiring fewer clicks) is associated with higher performance. Lastly, performance on the CT task (\u003cem\u003eL_Score\u003c/em\u003e) was positively associated with the time spent on the playground (\u003cem\u003ePG_Time\u003c/em\u003e, r\u0026thinsp;=\u0026thinsp;.295, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.01), highlighting the benefits of training in enhancing performance outcomes.\u003c/p\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eHow can preservice teachers be differentiated based on their CT skills?\u003c/h2\u003e \u003cp\u003eThe LPA indicated that AIC and BIC values decreased as the number of profiles increased, indicating improved fit with additional profiles. However, the selection of the optimal model was not based solely on the lowest AIC and BIC values but also considered entropy, profile probability bounds, and the BLRT. Specifically, the three-profile model exhibited a substantial decrease in AIC (995.93) and BIC (1035.85) compared to the two-profile model, coupled with high entropy (.84) indicating clear delineation between profiles. Probabilities of most likely profile membership ranged from .86 to .96, suggesting high classification accuracy. Importantly, the BLRT for the three-profile model (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.01) confirmed its statistical superiority over the two-profile model. Models with more than three profiles did not show meaningful improvements in fit or entropy to justify the added complexity. Both the four-profile (entropy\u0026thinsp;=\u0026thinsp;.77) and five-profile (entropy\u0026thinsp;=\u0026thinsp;.76) models exhibited lower entropy values than the three-profile model, indicating diminishing returns in profile distinctiveness and classification accuracy. Moreover, the marginal reductions in AIC and BIC were not accompanied by significant BLRT improvements, suggesting that these models might overfit the data. Figure\u0026nbsp;1 displays the profile plot with the standardized scores of digital literacy (DL), problem solving (PS), and coding (CD) across the three profiles as proposed by the LPA.\u003c/p\u003e \u003cp\u003eINSERT FIGURE 1 HERE\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eProfile 1: Developing (n\u0026thinsp;=\u0026thinsp;58)\u003c/h2\u003e \u003cp\u003eMost preservice teachers (7M, 45F, 5 other, and one missing) in Profile 1 were from Elementary (n\u0026thinsp;=\u0026thinsp;34), followed by Secondary (n\u0026thinsp;=\u0026thinsp;16), and other programs (n\u0026thinsp;=\u0026thinsp;8). Before completing the CT task, this group spent 30.14 seconds on the playground, which rendered to 44.8% completing the task on the first attempt, 13.8% on the second attempt, and 25.9% incorrectly completing the task. Profile 1displayed moderate levels of digital literacy (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5.53, \u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.74), problem-solving skills (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5.45, \u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.46), and coding (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.49, \u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.87), suggesting that participants in this profile were still developing foundational CT skills.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eProfile 2: Novice (n\u0026thinsp;=\u0026thinsp;43)\u003c/h2\u003e \u003cp\u003ePreservice teachers (3M, 39F, and one missing) in Profile 2 were also mostly from Elementary (n\u0026thinsp;=\u0026thinsp;31), followed by Secondary (n\u0026thinsp;=\u0026thinsp;8), and other programs (n\u0026thinsp;=\u0026thinsp;4). This group spent on average 39.93 seconds training on the playground. After completing the CT task, 48.8% correctly completed the task on the first attempt, 9.3% on the second attempt, and 18.6% did not correctly complete the task. Given that Profile 2 yielded below-average levels of digital literacy (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;3.88, \u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.83), problem-solving skills (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;4.40, \u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.76), and coding (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.76, \u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.69), participants in this profile were still considered novices and in their initial stages of acquiring CT skills.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eProfile 3: Proficient (n\u0026thinsp;=\u0026thinsp;27)\u003c/h2\u003e \u003cp\u003ePreservice teachers (7M, 17F, and 2 missing) in Profile 3 composed of students in Secondary (n\u0026thinsp;=\u0026thinsp;13), Elementary (n\u0026thinsp;=\u0026thinsp;8), and other programs (n\u0026thinsp;=\u0026thinsp;6) comprised a distinct subgroup characterized by its proficiency in CT skills. This group yielded above-average scores in digital literacy (M\u0026thinsp;=\u0026thinsp;6.22, SD\u0026thinsp;=\u0026thinsp;0.76), problem-solving skills (M\u0026thinsp;=\u0026thinsp;6.14, SD\u0026thinsp;=\u0026thinsp;0.61), and coding (M\u0026thinsp;=\u0026thinsp;4.70, SD\u0026thinsp;=\u0026thinsp;0.95). Profile 3 participants spent on average 70.80 seconds on the playground, with a majority (70.4%) successfully completing the task on their first attempt, 11.1% on the second attempt, whereas 7.4% incorrectly completed the task.\u003c/p\u003e \u003cp\u003eThe ANOVA results for CT_DL [\u003cem\u003eF\u003c/em\u003e(2,125)\u0026thinsp;=\u0026thinsp;90.969, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001], CT_PS [\u003cem\u003eF\u003c/em\u003e(2,125)\u0026thinsp;=\u0026thinsp;75.508, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001], and CT_CD [\u003cem\u003eF\u003c/em\u003e(2,125)\u0026thinsp;=\u0026thinsp;107.373, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001] provide strong evidence of significant differences in these measures across the three groups.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eTo what extent do preservice teachers\u0026rsquo; metacognitive strategies and prior coding experience predict CT skills?\u003c/h2\u003e \u003cp\u003eA Linear Discriminant Analysis (LDA) was performed to explore how well two predictors, Coding Experience (\u003cem\u003eCodExp\u003c/em\u003e) and Metacognitive Strategies (\u003cem\u003eMetaCog\u003c/em\u003e), could discriminate among the three profiles. Descriptive statistics revealed that Group 3 (N\u0026thinsp;=\u0026thinsp;27) was predominantly more experienced in coding (74%), compared to Group 2 (34.9%) and Group 1 (46.6%). Moreover, Group 1 (M\u0026thinsp;=\u0026thinsp;5.99, SD\u0026thinsp;=\u0026thinsp;0.70) and Group 3 (M\u0026thinsp;=\u0026thinsp;5.99, SD\u0026thinsp;=\u0026thinsp;0.66) displayed similar means in \u003cem\u003eMetaCog\u003c/em\u003e. Finally, Group 2 had a significant lower mean (M\u0026thinsp;=\u0026thinsp;5.53, SD\u0026thinsp;=\u0026thinsp;0.77) compared to Group 1 (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.004) and Group 3 (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.024).\u003c/p\u003e \u003cp\u003eWilks\u0026rsquo; lambda indicated significant differences among the groups for both \u003cem\u003eMetaCog\u003c/em\u003e [λ\u0026thinsp;=\u0026thinsp;.912, \u003cem\u003eF\u003c/em\u003e(2, 125)\u0026thinsp;=\u0026thinsp;6.068, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.003] and \u003cem\u003eCodExp\u003c/em\u003e [λ\u0026thinsp;=\u0026thinsp;.919, \u003cem\u003eF\u003c/em\u003e(2, 125)\u0026thinsp;=\u0026thinsp;5.499, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.005] across the groups. Box\u0026rsquo;s M test confirmed the assumption of equal covariance matrices was not violated, M\u0026thinsp;=\u0026thinsp;4.747, F(6, 88111.212)\u0026thinsp;=\u0026thinsp;.771, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.593. The analysis yielded two discriminant functions, with the first function explaining 82.5% of the variance between groups (canonical R\u0026sup2; = .358), and the second function explaining an additional 17.5% (canonical R\u0026sup2; = .174). Collectively, the discriminant functions significantly differentiated the groups, Wilks\u0026rsquo; Lambda\u0026thinsp;=\u0026thinsp;.845, χ\u0026sup2;(4)\u0026thinsp;=\u0026thinsp;20.913, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001, with the second function alone approaching significance, χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;3.831, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.050.\u003c/p\u003e \u003cp\u003eThe standardized canonical discriminant function coefficients for the first function were .716 for \u003cem\u003eMetaCog\u003c/em\u003e and .658 for \u003cem\u003eCodExp\u003c/em\u003e, whereas for the second function they were \u0026minus;\u0026thinsp;.701 for \u003cem\u003eMetaCog\u003c/em\u003e and .755 for \u003cem\u003eCodExp\u003c/em\u003e. Group centroids indicated that Group 1 was characterized by scores of .130 on the first function and \u0026minus;\u0026thinsp;.182 on the second function, Group 2 by scores of \u0026minus;\u0026thinsp;.494 on the first and .092 on the second, and Group 3 by scores of .506 on the first and .244 on the second. The analysis also provided classification functions for each group, with the coefficients for \u003cem\u003eCodExp\u003c/em\u003e ranging from 4.821 to 6.415 and for \u003cem\u003eMetaCog\u003c/em\u003e from 10.672 to 11.526. These functions effectively classified the participants into their respective groups based on the predictors.\u003c/p\u003e \u003cp\u003e \u003cb\u003eIn what ways do online behaviors on a CT task differ based on preservice teachers\u0026rsquo; CT skills?\u003c/b\u003e \u003c/p\u003e \u003cp\u003eParticipants were randomly assigned to different difficulty levels, with 24\u0026ndash;30% allocated to each level. A one-way ANOVA was conducted to examine the differences across the three profiles in time spent on playground (\u003cem\u003ePG_Time\u003c/em\u003e), time spent completing the CT task (\u003cem\u003eL_Time\u003c/em\u003e), and the number of clicks while completing the CT task (\u003cem\u003eL_Clicks\u003c/em\u003e), and perceived task difficulty (\u003cem\u003eL_Diff\u003c/em\u003e). A Kruskal-Wallis H test was also conducted to determine if there were differences in perceived task difficulty \u003cem\u003e(L_Diff)\u003c/em\u003e, across the three different profiles. The mean scores and standard deviations for each variable are presented in Table\u0026nbsp;2.\u003c/p\u003e \u003cp\u003eINSERT TABLE 2 HERE\u003c/p\u003e \u003cp\u003eThe one-way ANOVA revealed that there was a statistically significant difference in \u003cem\u003eL_Time\u003c/em\u003e [\u003cem\u003eF\u003c/em\u003e(2, 99)\u0026thinsp;=\u0026thinsp;4.072, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.020] across the profiles. No significant differences were found for \u003cem\u003ePG_Time\u003c/em\u003e [\u003cem\u003eF\u003c/em\u003e(2, 121)\u0026thinsp;=\u0026thinsp;.891, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.413] or \u003cem\u003eL_Clicks\u003c/em\u003e [\u003cem\u003eF\u003c/em\u003e(2, 99)\u0026thinsp;=\u0026thinsp;1.692, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.189]. The Kruskal-Wallis H test also revealed a significant difference in the medians of \u003cem\u003eL_Diff\u003c/em\u003e across the profiles, [H(2)\u0026thinsp;=\u0026thinsp;17.271, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001], suggesting that not all groups perceived task difficulty similarly.\u003c/p\u003e \u003cp\u003ePost hoc comparisons using the LSD test indicated that the mean score for \u003cem\u003eL_Time\u003c/em\u003e for Profile 1 (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;131.42, \u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;72.92) was significantly different (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.007) than that for Profile 3 (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;87.91, \u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;55.44). No significant differences were found between Profiles 1 and 2 (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;106.65, \u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;50.79), \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.095, and between Profiles 2 and 3, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.281. Post-hoc analyses using pairwise comparisons with Bonferroni correction also indicated a significant difference in \u003cem\u003eL_Diff\u003c/em\u003e scores (H\u0026thinsp;=\u0026thinsp;28.804, SE\u0026thinsp;=\u0026thinsp;7.491, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001) between Profiles 1 and 3 and between Profiles 2 and 3 (H\u0026thinsp;=\u0026thinsp;29.380, SE\u0026thinsp;=\u0026thinsp;8.118, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001). No significant differences were found between Profiles 1 and 2 (H\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;.576, SE\u0026thinsp;=\u0026thinsp;6.833, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.933).\u003c/p\u003e \u003cp\u003eThese findings suggest that learner profiles significantly differ in terms of how long they engage with learning tasks and their perceived difficulty, with Profile 3 participants typically finding tasks less challenging and spending less time on them compared to participants belonging to the other profiles.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study examined the variations in preservice teachers\u0026rsquo; computational thinking (CT) skills, the predictive role of prior coding experience and metacognitive strategies on these skills, and how these skills influence their behavior during an online CT task. Our findings extend Shute et al.\u0026rsquo;s (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) CT model by demonstrating that metacognitive strategies and coding experience play a significant role in CT proficiency, a dynamic previously underexplored in preservice teacher education. The identification of three distinct profiles (Developing, Novice, Proficient) provides empirical support for Rom\u0026aacute;n-Gonz\u0026aacute;lez et al.\u0026rsquo;s (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) assertion that CT skill acquisition is non-linear and context-dependent. This profiling approach is novel in the CT literature and offers a replicable framework for assessing preservice teachers\u0026rsquo; readiness to integrate computational principles into pedagogy.\u003c/p\u003e \u003cp\u003eThe findings revealed significant differentiation among preservice teachers\u0026rsquo; CT skills based on a three-profile model identified through Latent Profile Analysis (LPA). The most proficient group in CT, Profile 3, displayed above-average levels of digital literacy, problem-solving, and comfort with coding, which translated into higher performance scores and a lesser perceived difficulty of the CT task, thus corroborating with the hypothesis that higher CT skills enhance problem-solving efficiency (Cheng et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Specifically, a recent systematic review revealed a close link between CT and problem solving in the 37 studies reviewed (Wu, Asmara, Huang, \u0026amp; Permata Hapsari, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This finding supports the notion that well-developed CT skills can significantly ease the cognitive load of problem-solving learning situations, a result consistent with earlier studies highlighting the importance of strong CT skills in reducing perceived task difficulty and enhancing problem-solving efficiency (Barr \u0026amp; Stephenson, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2011\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOur results also highlighted the critical role of prior coding experience and the use of metacognitive strategies in predicting varying levels of CT skills. The positive correlation between coding experience and CT skills underscores the value of practical coding knowledge in enhancing computational proficiency. This aligns with previous research by Liao et al. (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), highlighting the significance of background knowledge in CT development. As hypothesized, preservice teachers with prior coding experience also found the CT task less daunting and displayed higher problem-solving skills, emphasizing the need for integrating practical coding exercises in teacher-education programs. The study\u0026rsquo;s results also confirm the hypothesis that metacognitive strategies are crucial in navigating CT tasks effectively. The use of these strategies was strongly correlated with better problem-solving skills, echoing the findings of Pintrich et al. (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e1993\u003c/span\u003e), which emphasize the importance of strategic and reflective thinking in learning environments. This supports the conceptual framework that positions metacognitive engagement as integral to mastering CT tasks. These observations suggest that both technical skills and reflective practices are essential for fostering effective computational thinking, aligning with Shute et al.\u0026rsquo;s (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) emphasis on a pedagogical approach that integrates metacognitive skills into CT education.\u003c/p\u003e \u003cp\u003eMoreover, the relationship between time spent on training activities (i.e., the playground, PG_Time) and subsequent performance on a CT task underscores the importance of preparatory engagement in enhancing the application of CT skills. This is consistent with findings in line with educational theories that advocate for experiential learning, in which a positive correlation was demonstrated between prior engagement with relevant learning activities and improved skill application in actual problem-solving contexts (Argelag\u0026oacute;s et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThis study contributes to the theoretical understanding of computational thinking by validating and extending Shute et al.\u0026rsquo;s (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) model, particularly in the context of preservice teacher education. Our findings highlight the importance of metacognitive strategies as a mediating factor in CT skill development, a dimension that has been underexplored in existing frameworks. By demonstrating the interplay between coding experience and metacognitive engagement, this study bridges the gap between technical proficiency and reflective practice, offering a more holistic view of CT development. Furthermore, the differentiation of preservice teachers into distinct profiles (Developing, Novice, and Proficient) aligns with Rom\u0026aacute;n-Gonz\u0026aacute;lez et al.\u0026rsquo;s (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) assertion that CT skill acquisition is not uniform but varies based on individual experiences and cognitive strategies. This nuanced understanding of CT development can inform future theoretical models by emphasizing the role of metacognition and prior experience in shaping computational proficiency.\u003c/p\u003e \u003cp\u003eWhile this study offers significant insights, it also reveals important limitations that call for further research. The cross-sectional nature of our data limits our ability to discern the developmental trajectories of CT skills among preservice teachers, which would indicate how these competencies evolve over time. To address these limitations and build on the findings of this study, further research could explore longitudinal studies to track the progression of CT skills over time and the long-term impact of metacognitive strategies and coding experience on these skills. Such studies would provide deeper insights into the developmental aspects of CT skills and help in designing interventions that are more effective over the long term. Additionally, experimental and comparative studies could be conducted to directly measure the effects of specific teaching interventions on CT skills development. These studies would help clarify the causality between teaching methods and CT skills acquisition, which this study suggests but cannot definitively establish due to its methodological constraints. Additionally, while prior studies often rely on self-reports or isolated performance metrics, our mixed-methods design, combining self-assessments, behavioral analytics (e.g., clicks, time-on-task), and task performance provides a holistic view of CT development. This methodological innovation addresses critiques of oversimplified CT assessments (Grover \u0026amp; Pea, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) and sets a precedent for future research in teacher education.\u003c/p\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eImplications\u003c/h2\u003e \u003cp\u003eThis research makes a theoretical contribution by providing valuable insights into the factors influencing CT skills among preservice teachers. It also makes a practical contribution by offering a basis for enhancing educational practices and curriculum design to better prepare future educators for the increasingly digital world. Lastly, a methodological contribution was presented through the use of detailed log data from online learning platforms, which provided a rich resource for understanding learner interactions in problem-solving situations that require the application of computational thinking skills. This approach aids not only in the immediate improvement of individual learning strategies but also informs broader educational practices and curriculum development.\u003c/p\u003e \u003cp\u003eThis study leads to multiple implications. First, the integration of CT tasks into preservice teacher education can serve as a valuable diagnostic and developmental tool in educational practices and policy development. The clear delineation in CT skill levels among preservice teachers suggests a need for differentiated instructional strategies that cater to specific needs of each group, thereby enhancing overall CT proficiency before these individuals enter the professional field. These findings also underscore the importance of personalized learning interventions that can be tailored to focus on enhancing coding experience and fostering metacognitive abilities as these proved to play a pivotal role in predicting CT skills. For instance, learners in the novice group might benefit more from targeted support and incremental task levels, which could help in gradually building their confidence and skills. Furthermore, given the demonstrated importance of coding experience and metacognitive strategies, the findings advocate for the integration of structured coding tasks and metacognition training in teacher-education programs. This approach could improve CT skills uniformly across different learner profiles, equipping future educators with the necessary skills not only to tackle digital tasks but also to implement technology-driven pedagogies effectively in their teaching practices.\u003c/p\u003e \u003cp\u003eFrom a practical standpoint, this study underscores the need for teacher education programs to adopt a dual-focused approach that combines technical coding skills with metacognitive training. For instance, programs could incorporate project-based learning activities that require preservice teachers to reflect on their problem-solving processes, such as journaling or peer feedback sessions. Tools like Scratch or robotics kits could be used to provide hands-on coding experience, while metacognitive strategies could be reinforced through guided reflection prompts or self-assessment checklists. Additionally, the findings suggest the value of adaptive learning technologies that tailor tasks to individual skill levels, ensuring that novice learners receive the scaffolding they need while proficient learners are challenged with more complex problems. Policymakers could support these efforts by allocating resources for professional development workshops focused on CT integration and by incentivizing the adoption of evidence-based practices in teacher education curricula and ensuring preservice teachers graduate equipped to model CT across disciplines.\u003c/p\u003e \u003cp\u003eIn the context of 21st-century education, where computational thinking is increasingly recognized as a core competency, this study highlights the critical role of teacher preparation in meeting the demands of a technology-driven world. As K-12 curricula worldwide begin to integrate CT into subjects beyond computer science, preservice teachers must be equipped not only with technical skills but also with the ability to foster these skills in their students. Our findings suggest that teacher education programs should prioritize the development of both coding proficiency and metacognitive strategies, as these are essential for enabling future educators to model and teach CT effectively. By doing so, preservice teachers will be better prepared to create learning environments that encourage algorithmic thinking, abstraction, and problem-solving across disciplines, ultimately contributing to a more computationally literate society.\u003c/p\u003e \u003cp\u003eWhile this study provides valuable insights, it is not without limitations. First, the reliance on self-reported data for CT skills and metacognitive strategies may introduce bias, as participants\u0026rsquo; perceptions may not always align with their actual abilities. Future studies could complement self-reports with performance-based assessments to provide a more objective measure of CT proficiency. Second, the study was conducted within a single institution, which may limit the generalizability of the findings. Replicating this study in diverse educational contexts, including different countries or regions, would help determine the extent to which these findings apply across settings. Finally, the cross-sectional design of the study precludes causal inferences about the relationship between coding experience, metacognitive strategies, and CT skills. Longitudinal studies tracking preservice teachers\u0026rsquo; CT development over time could provide deeper insights into how these factors interact and evolve. Additionally, experimental studies could test the effectiveness of specific interventions, such as metacognitive training modules or coding bootcamps, in enhancing CT skills. Exploring the role of cultural and contextual factors in CT development could also yield valuable insights, particularly in non-Western educational systems where CT integration may be at an earlier stage.\u003c/p\u003e \u003cp\u003eThis study advances the field by demonstrating that CT proficiency in preservice teachers is not merely a function of technical skill but a cognitive-metacognitive interplay. By linking coding experience to reduced task difficulty and metacognition to enhanced problem-solving, we provide a roadmap for reimagining teacher education. These insights are timely, as global K-12 reforms increasingly demand educators who can seamlessly integrate CT into diverse subjects, a competency our findings show is achievable through targeted, evidence-based training\u003c/p\u003e \u003cp\u003eIn conclusion, this study highlights the critical role of prior coding experience and metacognitive strategies in shaping preservice teachers\u0026rsquo; computational thinking skills. By identifying distinct profiles of CT proficiency and demonstrating the importance of preparatory engagement and reflective practice, this research provides a foundation for designing more effective teacher education programs. As the demand for computational thinking continues to grow in K-12 education, equipping preservice teachers with the skills and strategies to teach CT effectively will be essential for preparing students to thrive in a technology-driven world. Ultimately, this study not only advances our understanding of CT development but also underscores the importance of fostering both technical and cognitive skills in future educators, paving the way for a more computationally literate and innovative society.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAcknowledgments\u003c/h2\u003e \u003cp\u003eWe extend our gratitude to [name blinded], a graduate student, for her invaluable assistance with data collection during our classroom visits. We also thank the three college students, [names blinded] whose dedication and creativity in designing the computational thinking tasks were instrumental to the success of this project.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eArgelag\u0026oacute;s E, Garcia C, Privado J, Wopereis I (2022) Fostering information problem solving skills through online task-centred instruction in higher education. Comput Educ 180:104433. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.compedu.2022.104433\u003c/span\u003e\u003cspan address=\"10.1016/j.compedu.2022.104433\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAuthor 1 Blinded for review\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAuthor 2 Blinded for review\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAuthor 3. Blinded for review\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarr V, Stephenson C (2011) Bringing computational thinking to K-12: What is involved and what is the role of the computer science education community? ACM Inroads 2(1):48\u0026ndash;54. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1145/1929887.1929905\u003c/span\u003e\u003cspan address=\"10.1145/1929887.1929905\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBers MU (2018) Coding as playground: Programming and computational thinking in the early childhood classroom. Routledge, London, UK\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBower M, Falkner K (2015), January Computational thinking, the notional machine, pre-service teachers, and research opportunities. In \u003cem\u003eProceedings of the 17th Australasian computing education conference (ACE 2015)\u003c/em\u003e (Vol. 27, p. 30). Australian Computer Society\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrennan K, Resnick M (2012), April New frameworks for studying and assessing the development of computational thinking. In \u003cem\u003eProceedings of the 2012 annual meeting of the American educational research association, Vancouver, Canada\u003c/em\u003e (Vol. 1, p. 25)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eButler D, Leahy M (2021) Developing preservice teachers\u0026rsquo; understanding of computational thinking: A constructionist approach. Br J Edu Technol 52(3):1060\u0026ndash;1077. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/bjet.13090\u003c/span\u003e\u003cspan address=\"10.1111/bjet.13090\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChan SW, Looi CK, Ho WK, Huang W, Seow P, Wu L (2021) Learning number patterns through computational thinking activities: A Rasch model analysis. Heliyon 7(9):e07922. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.heliyon.2021.e07922\u003c/span\u003e\u003cspan address=\"10.1016/j.heliyon.2021.e07922\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheng Y-P, Lai C-F, Chen Y-T, Wang W-S, Huang Y-M, Wu T-T (2023) Enhancing student's computational thinking skills with student-generated questions strategy in a game-based learning platform. Comput Educ 200:1\u0026ndash;20. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.compedu.2023.104794\u003c/span\u003e\u003cspan address=\"10.1016/j.compedu.2023.104794\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDagienė V, Futschek G (2008) Bebras international contest on informatics and computer literacy: Criteria for good tasks. In: Mittermeir RT, Sysło MM (eds) \u003cem\u003eInformatics Education - Supporting Computational Thinking\u003c/em\u003e. ISSEP 2008. Lecture Notes in Computer Science, vol 5090. Springer, Berlin, Heidelberg. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-3-540-69924-8_2\u003c/span\u003e\u003cspan address=\"10.1007/978-3-540-69924-8_2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong W, Li Y, Sun L, Liu Y (2024) Developing pre-service teachers\u0026rsquo; computational thinking: A systematic literature review. Int J Technol Des Educ 34:191\u0026ndash;127. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10798-023-09811-3\u003c/span\u003e\u003cspan address=\"10.1007/s10798-023-09811-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFlavell JH (1979) Metacognition and cognitive monitoring: A new area of cognitive\u0026ndash;developmental inquiry. Am Psychol 34(10):906\u0026ndash;911. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1037/0003-066X.34.10.906\u003c/span\u003e\u003cspan address=\"10.1037/0003-066X.34.10.906\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrover S, Pea R (2013) Computational thinking in K-12: A review of the state of the field. Educational Researcher 42(1):38\u0026ndash;43. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3102/0013189X12463051\u003c/span\u003e\u003cspan address=\"10.3102/0013189X12463051\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoyles C, Noss R A computational lens on design research. ZDM, \u003cem\u003e47\u003c/em\u003e(6), 1039\u0026ndash;1045. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11858-015-0731-2\u003c/span\u003e\u003cspan address=\"10.1007/s11858-015-0731-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJaipal-Jamani K, Angeli C (2017) Effect of robotics on elementary preservice teachers\u0026rsquo; self-efficacy, science learning, and computational thinking. J Sci Edu Technol 26(2):175\u0026ndash;192. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10956-016-9663-z\u003c/span\u003e\u003cspan address=\"10.1007/s10956-016-9663-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKotsopoulos D, Floyd L, Khan S, Namukasa IK, Somanath S, Weber J, Yiu C (2017) A pedagogical framework for computational thinking. Digit Experiences Math Educ 3(2):154\u0026ndash;171. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s40751-017-0031-2\u003c/span\u003e\u003cspan address=\"10.1007/s40751-017-0031-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiao CH, Chiang CT, Chen IC, Parker KR (2022) Exploring the relationship between computational thinking and learning satisfaction for non-STEM college students. Int J Educational Technol High Educ 19(1):43. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s41239-022-00347-5\u003c/span\u003e\u003cspan address=\"10.1186/s41239-022-00347-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Z, Gearty Z, Richard E et al (2024) Bringing computational thinking into classrooms: a systematic review on supporting teachers in integrating computational thinking into K-12 classrooms. Int J STEM Educ 11:51, 1\u0026ndash;18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40594-024-00510-6\u003c/span\u003e\u003cspan address=\"10.1186/s40594-024-00510-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMishra P, Koehler MJ (2006) Technological pedagogical content knowledge: A framework for teacher knowledge. Teachers Coll Record 108(6):1017\u0026ndash;1054. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1467-9620.2006.00684.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1467-9620.2006.00684.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMouza C, Yang H, Pan Y-C, Ozden SY, Pollock L (2017) Resetting educational technology coursework for pre-service teachers: A computational thinking approach to the development of technological pedagogical content knowledge (TPACK). Australasian J Educational Technol 33(3). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.14742/ajet.3521\u003c/span\u003e\u003cspan address=\"10.14742/ajet.3521\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePintrich PR, Smith DAF, Garcia T, McKeachie WJ (1993) Reliability and predictive validity of the Motivated Strategies for Learning Questionnaire (Mslq). Educ Psychol Meas 53(3):801\u0026ndash;813. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/0013164493053003024\u003c/span\u003e\u003cspan address=\"10.1177/0013164493053003024\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRom\u0026aacute;n-Gonz\u0026aacute;lez M, Moreno-Le\u0026oacute;n J, Robles G (2019) Combining assessment tools for a comprehensive evaluation of computational thinking interventions. In: Kong SC, Abelson H (eds) Computational Thinking Education. Springer, Singapore. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-981-13-6528-7_6\u003c/span\u003e\u003cspan address=\"10.1007/978-981-13-6528-7_6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSadik O, Ottenbreit-Leftwich AT, Nadiruzzaman H (2017) Computational thinking conceptions and misconceptions: Progression of preservice teacher thinking during computer science lesson planning. In: Rich P, Hodges C (eds) Emerging Research, Practice, and Policy on Computational Thinking. Educational Communications and Technology: Issues and Innovations. Springer, Cham. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-3-319-52691-1_14\u003c/span\u003e\u003cspan address=\"10.1007/978-3-319-52691-1_14\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShodiev H (2015) Computational thinking and simulation in teaching science and mathematics. In M. G. Cojocaru, I. S. Kotsireas, R. N. Makarov, R. V. N. Melnik, \u0026amp; H. Shodiev (Eds.), \u003cem\u003eInterdisciplinary Topics in Applied Mathematics, Modeling and Computational Science\u003c/em\u003e (pp. 405\u0026ndash;410). Springer. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-3-319-12307-3_58\u003c/span\u003e\u003cspan address=\"10.1007/978-3-319-12307-3_58\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShute VJ, Sun C, Asbell-Clarke J (2017) Demystifying computational thinking. Educational Res Rev 22:142\u0026ndash;158. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.edurev.2017.09.003\u003c/span\u003e\u003cspan address=\"10.1016/j.edurev.2017.09.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang X, Yin Y, Lin Q, Hadad R, Zhai X (2020) Assessing computational thinking: A systematic review of empirical studies. Comput Educ 148:103798. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.compedu. 2019.103798\u003c/span\u003e\u003cspan address=\"10.1016/j.compedu. 2019.103798\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeintrop D, Beheshti E, Horn M, Orton K, Jona K, Trouille L, Wilensky U (2016) Defining computational thinking for mathematics and science classrooms. J Sci Edu Technol 25(1):127\u0026ndash;147. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10956-015-9581-5\u003c/span\u003e\u003cspan address=\"10.1007/s10956-015-9581-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWing JM (2006) Computational thinking. Commun ACM 49(3):33\u0026ndash;35. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1145/1118178.1118215\u003c/span\u003e\u003cspan address=\"10.1145/1118178.1118215\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu TT, Asmara A, Huang YM, Permata Hapsari I (2024) Identification of problem-solving techniques in computational thinking studies: Systematic literature review. SAGE Open 14(2). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/21582440241249897\u003c/span\u003e\u003cspan address=\"10.1177/21582440241249897\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e\u003cstrong\u003eTable 1\u003c/strong\u003e\u003cem\u003e\u003cbr\u003e\u0026nbsp;Means and Standard Deviations for Online Behaviors\u003c/em\u003e\u003c/p\u003e\n \u003ctable border=\"0\" cellspacing=\"3\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProfile\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMean\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSD\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e95% CI for Mean\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\n \u003cp\u003e\u003cem\u003ePG_Time\u003c/em\u003e (in seconds)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e62.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e30.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[54.13, 70.42]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e60.15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e38.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[47.92, 72.38]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e70.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e30.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[58.89, 82.71]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\n \u003cp\u003e\u003cem\u003eL_Time\u003c/em\u003e (in seconds)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e131.42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e72.92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[110.24, 152.59]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e106.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e50.79\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[87.69, 125.62]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e87.91\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e55.44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[64.50, 111.31]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\n \u003cp\u003e\u003cem\u003eL_Clicks\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e49.81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e27.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[41.93, 57.69]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e51.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e28.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[41.22, 62.18]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e38.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e28.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[26.59, 50.91]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\n \u003cp\u003e\u003cem\u003eL_Diff\u003c/em\u003e (on a scale of 10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e5.42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e3.21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[4.49, 6.35]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e5.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e3.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[4.55, 7.25]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e2.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e2.82\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[1.56, 3.94]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u0026nbsp;\u003cem\u003ePG_Time:\u003c/em\u003e Time spent on playground\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003cem\u003eL_Time:\u0026nbsp;\u003c/em\u003eTime spent completing the task (Programming Lamps)\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003cem\u003eL_Clicks:\u003c/em\u003e Number of clicks while completing the task (Programming Lamps)\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003cem\u003eL_Diff:\u003c/em\u003e Perceived level of task difficulty\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2\u0026nbsp;\u003c/strong\u003e\u003cem\u003e\u0026nbsp;Means and Standard Deviations for Online Behaviors\u003c/em\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"3\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProfile\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMean\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSD\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e95% CI for Mean\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\n \u003cp\u003e\u003cem\u003ePG_Time\u003c/em\u003e (in seconds)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e62.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e30.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[54.13, 70.42]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e60.15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e38.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[47.92, 72.38]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e70.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e30.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[58.89, 82.71]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\n \u003cp\u003e\u003cem\u003eL_Time\u003c/em\u003e (in seconds)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e131.42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e72.92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[110.24, 152.59]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e106.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e50.79\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[87.69, 125.62]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e87.91\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e55.44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[64.50, 111.31]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\n \u003cp\u003e\u003cem\u003eL_Clicks\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e49.81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e27.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[41.93, 57.69]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e51.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e28.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[41.22, 62.18]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e38.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e28.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[26.59, 50.91]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\n \u003cp\u003e\u003cem\u003eL_Diff\u003c/em\u003e (on a scale of 10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e5.42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e3.21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[4.49, 6.35]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e5.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e3.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[4.55, 7.25]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 26px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 7px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e2.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 11px;\"\u003e\n \u003cp\u003e2.82\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 37px;\"\u003e\n \u003cp\u003e[1.56, 3.94]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003cp\u003e\u003cem\u003ePG_Time:\u003c/em\u003e Time spent on playground\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eL_Time:\u0026nbsp;\u003c/em\u003eTime spent completing the task (Programming Lamps)\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003cem\u003eL_Clicks:\u003c/em\u003e Number of clicks while completing the task (Programming Lamps)\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003cem\u003eL_Diff:\u003c/em\u003e Perceived level of task difficulty\u0026nbsp;\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[{"identity":"03f94f9e-77c4-4699-ac4a-81f86ed34c4c","identifier":"10.13039/501100000155","name":"Social Sciences and Humanities Research Council of Canada","awardNumber":"430-2021-00707","order_by":0}],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Université de Sherbrooke","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":"computational thinking, preservice teachers, teacher education, coding, metacognition","lastPublishedDoi":"10.21203/rs.3.rs-6283533/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6283533/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn our technology-driven world, developing computational thinking skills among preservice teachers is essential for fostering digital innovation and preparing them to effectively integrate technology into their future classrooms. This study investigates how metacognitive strategies and prior coding experience impact preservice teachers’ computational thinking skills and online problem-solving behaviors during an interactive task. A sample of preservice teachers (n=129) completed a survey assessing their computational thinking skills and metacognitive strategies and then engaged in completing an online task. Latent profile analysis using self-reported computational thinking skills identified three distinct profiles: Developing, Novice, and Proficient. Key results revealed that higher levels of self-reported computational thinking skills correlated positively with better performance on the online task, as well as lower perceptions of task difficulty. Both metacognitive strategies and prior coding experience were significant predictors of computational thinking skills profile membership. This study highlights the importance of integrating targeted coding exercises and metacognitive skill-building activities into teacher education programs to better prepare future educators to implement digital technologies effectively and confidently in their teaching practices. The findings provide actionable insights for designing teacher training that fosters computational thinking competencies essential for the modern digital era.\u003c/p\u003e","manuscriptTitle":"Profiling Preservice Teachers’ Computational Thinking: The Role of Metacognition and Coding Experience","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-03-25 07:17:57","doi":"10.21203/rs.3.rs-6283533/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":"90e69e26-f9d3-4da5-8371-cd37a2208b9c","owner":[],"postedDate":"March 25th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":46061590,"name":"Educational Psychology"}],"tags":[],"updatedAt":"2025-03-25T07:17:57+00:00","versionOfRecord":[],"versionCreatedAt":"2025-03-25 07:17:57","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6283533","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6283533","identity":"rs-6283533","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}