Theory Precedes Practice: Simulation-Based Learning Enhances Long-Term Recall, but Prior Text-Based Learning Enhances Its Benefits

Theory Precedes Practice: Simulation-Based Learning Enhances Long-Term Recall, but Prior Text-Based Learning Enhances Its Benefits | 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 Article Theory Precedes Practice: Simulation-Based Learning Enhances Long-Term Recall, but Prior Text-Based Learning Enhances Its Benefits Angélique Lebert, Oscar Vilarroya This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6253202/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 24 Nov, 2025 Read the published version in npj Science of Learning → Version 1 posted 14 You are reading this latest preprint version Abstract Simulation-Based Learning (SBL) is widely used in medical and STEM education, offering immersive, embodied, and interactive experiences. However, its implementation often introduces variability in control conditions, instructional design, and a reliance on between-subjects comparisons, making it difficult to isolate its specific contributions to learning. This study used a within-subjects randomized controlled design (N=88) to evaluate the effects of SBL on knowledge retention, retrieval efficiency, and confidence calibration. Participants, naïve to the learning content, learned two counterbalanced fictitious clinical cases via either a live-actor simulation or a structured text-based format. Retention was assessed one month later through video-based and written evaluations measuring accuracy, reaction time, and confidence. SBL led to significantly faster reaction times (Estimate = 0.066, 95% CI [0.03, 0.10], SE = 0.017, t = 3.90, p < 0.001) and higher recall accuracy (Estimate = -0.456, SE = 0.097, z = -4.71, p < 0.001, OR = 0.63, 95% CI [0.52, 0.77]) compared to text-based learning. An order effect emerged: learning first via text enhanced subsequent SBL performance, whereas the reverse sequence impaired text-based retention (Estimate = –0.837, SE = 0.343, z = –2.44, p = 0.015; OR = 0.43, 95% CI [0.22, 0.85]). Mental imagery ability influenced retrieval accuracy, with higher imagery scores predicting greater accuracy overall (Estimate = 0.266, SE = 0.114, z = 2.34, p = 0.019; OR = 1.30, 95% CI [1.04, 1.63]). A significant interaction between imagery ability and modality showed that this effect was more pronounced in the text-based condition (Estimate = -0.186, SE = 0.058, z = -3.22, p = 0.001; OR = 0.83, 95% CI [0.74, 0.93]). Confidence ratings further highlighted SBL’s advantages, with participants in the SBL condition being three times more likely to report absolute confidence (Estimate = –1.14, SE = 0.08, z = –13.68, p < 0.001; OR = 0.32, 95% CI [0.27, 0.38]). Moreover, in the SBL condition, confidence was more closely aligned with actual accuracy. This study provides empirical evidence supporting the benefits of SBL over traditional text-based learning for the acquisition and long-term retention of clinical knowledge. While SBL enhances learning, our results suggest that structured, text-based methods can also yield strong retention outcomes, particularly for item identification and sequential recall. These findings clarify the role of SBL’s immersive, embodied, and interactive elements in shaping learning while highlighting the impact of instructional sequencing and individual differences in imagery ability. Additionally, they underscore the potential benefit of SBL in aligning self-confidence with accuracy. By isolating specific SBL features, this study refines our understanding of its effects on knowledge acquisition, retrieval, and self-confidence alignment. This refined understanding allows for better-informed design of SBL interventions and offers insights that can be applied to non-SBL learning environments. Biological sciences/Psychology Biological sciences/Psychology/Human behaviour Simulation-Based Learning simulation training mental imagery Instructional sequencing text-based learning immersive learning interactive learning Figures Figure 1 Figure 2 Figure 3 Figure 4 I. Introduction Simulation-Based Learning (SBL) is an educational approach that uses realistic, interactive, and immersive simulations to enhance learning and skill acquisition 1 , 2 . By replicating real-world scenarios, it offers hands-on experiences and serves as an alternative to traditional learning methods. SBL has been widely applied in specialized fields such as STEM (science, technology, engineering, and mathematics) and medical education to develop procedural skills, communication abilities, decision-making, and the application of theoretical knowledge in controlled environments 3 , 4 . Three key aspects of SBL are believed to enhance learning. First, it provides a realistic experience by immersing learners in scenarios featuring real or virtual actors, tools, and contexts that closely resemble those in which the acquired skills or knowledge will be applied 4 . Second, SBL is typically designed to create a sense of immersion, fully engaging a learner’s attention and perception and making them feel deeply involved or absorbed 1 . Finally, SBL introduces an embodied and interactive dimension that strengthens learning by requiring active engagement with the environment, material objects, and other individuals. The embodied aspect grounds learning in real experiences, while the interactive component dynamically reinforces cognitive processes through action and response rather than passive reception 5 . Research on SBL has primarily been conducted in structured educational settings, such as classrooms, professional training programs, and hospital-based instruction 6 . These environments integrate SBL with relevant content, materials, and instructional methods, often incorporating complementary support such as lectures, guided reflections, worked examples, or feedback during or after simulations, facilitating the impact of SBL interventions 7 , 4 , 6 . While these elements likely enhance learning, their specific contributions are not always systematically controlled, making it difficult to isolate the unique role of SBL in the learning process 8 , 9 , 7 , 4 , 6 , 10 , 11 . Moreover, while some SBL studies employ randomized controlled trials (RCTs), practical constraints often shape sample sizes, comparison conditions, requiring a balance between ecological validity and experimental control. Consequently, comparison groups and control conditions may differ in important aspects, complicating attributions of learning differences to SBL itself. Variability of the control group 8 , 9 content exposure 7 , learning duration, instructional setting 4 , 6 , assessment analyses and methods across conditions 12 , 7 , 6 , 10 , potential biases from non-blinded assessments (e.g., by experts) 7 , 8 , and limited assessments of long-term retention 7 , 8 further complicates interpretation. Additionally, the widespread use of between-subjects designs provides insight into overall learning effects but introduces inter-individual variability. In sum, while prior research highlights SBL’s potential, methodological complexities make it challenging to assess its true impact. Variations in instructional design, engagement levels, and assessment procedures raise questions about whether observed performance differences genuinely reflect the benefits of SBL or stem from extraneous factors. Addressing these limitations through a rigorously controlled study with a robust experimental design in a lab setting is crucial for accurately determining SBL’s advantages over traditional learning methods. This study examines the effectiveness of a simulation-based learning (SBL) intervention compared to text-based learning in acquiring and retaining novel information. Participants, who were naïve to the learning content, learned two distinct and fictitious clinical cases with structured symptom sequences, either through interactive simulation with an actor or through a text-based method. The experimental design systematically controlled for order effects and task biases by counterbalancing conditions and symptom presentations. To ensure that the acquired knowledge was entirely novel and free from preexisting biases, we designed the study around a clinical case that was intentionally rendered fictitious. This deliberate separation—a genuine clinical scenario presented as a fictitious case—ensured that participants encountered completely new information. Clinical cases, widely used in SBL, provided us with an optimal balance between acquiring specific knowledge and enabling its representation through bodily movements. The structured, sequential symptom presentation mirrors realistic clinical encounters while enhancing the representability of the learned information. In addition, observational learning has been shown to improve retention of relational structures, such as sequential item patterns 13 , 14 , hence the decision to use an actor simulating the cases. Moreover, the actor not only presented the clinical case but also engaged participants through controlled physical interactions, ensuring an active learning process rather than passive observation. Research shows that touch affects memory recognition 15 , 16 , gesturing during encoding enhances recall similarly to action 17 , and body posture facilitates autobiographical memory retrieval 18 . More broadly, embodied effects, such as the “enactment effect” 19 , have shown to enhance retention and recall 20 . In sum, this protocol allowed us to integrate realism and embodied interactivity—key features of SBL—immersing participants in a lifelike learning experience. Long-term memory was assessed using item recognition and serial recall tasks, which are widely employed to measure retrieval strength and temporal organization 21 – 24 . These methods have also been used to investigate the effects of simulation, embodiment, and embedded learning 20 . Learning outcomes were assessed using identical objective measures across both conditions, ensuring that retrieval mode did not influence results. The study also included both video-based and written evaluations to assess whether knowledge acquired through SBL effectively transfers across different retrieval contexts. Beyond learning and retention, we examine the impact of an imagery questionnaire on accuracy and reaction time across different retrieval modes. Specifically, we explore whether an individual's perceived capacity for mental imagery influences retrieval accuracy, particularly in the written evaluation modality. Prior research has shown that mental imagery can substitute for direct perceptual experience by engaging neural mechanisms similar to actual perception 25 . This internal simulation not only allows individuals with stronger imagery abilities to reconstruct learned information more accurately and efficiently in contexts lacking external cues but also strengthens recall for abstract concepts when linked to sensorimotor experiences 26 . Consequently, internal simulation appears capable of compensating for the absence of direct interaction, thereby reinforcing structured learning. Based on the proposed role of realism, immersion, embodiment, and interactivity in Simulation-Based Learning (SBL), we hypothesize that participants will exhibit faster reaction times, higher accuracy, and greater confidence in the SBL condition compared to the text-based learning condition. Additionally, due to the within-subjects design and counterbalancing of conditions, and a delayed one-month assessment, we do not anticipate a systematic effect of learning order on performance. Furthermore, given the distinct cognitive processes involved in retrieving information from video versus text, we expect differences in reaction time and accuracy between these modalities. Specifically, we hypothesize that retrieval via video may lead to faster reaction times and higher accuracy compared to written modality retrieval, as video provides a more direct representation of the learned clinical cases, reducing the need for reconstructive processing. Additionally, we predict that individuals with higher imagery capacity may exhibit faster reaction time and accuracy, particularly in the written modality where internal simulation may compensate for the lack of external sensory cues. Finally, we hypothesize that participants will report higher absolute confidence in their responses in the simulation condition compared to text-based learning. We propose that SBL’s immersive, interactive, and embodied design enhances sensorimotor engagement and internal simulation, fostering a robust internal feedback mechanism that reinforces both prediction and its verification. In sum, by systematically comparing SBL and text-based learning while controlling for prior knowledge and potential confounds, this study contributes to a broader understanding of how embodied and interactive learning experiences shape knowledge acquisition and retention. II. Method The presented study was not preregistered. This research was approved by the Ethics Committee at the Hospital del Mar Research Institute (Ref : 2022/10237/I). All participants signed a consent form before participating in the study and received monetary compensation for their time. 1. Participants Participants ranged in age from 18 to 35 years, with no history of neurological disorders and no prior experience or studies in the medical, psychiatric, or psychological fields (a complete list of inclusion and exclusion criteria is available in the Supplementary Material 1). They were recruited through convenience sampling by posting flyers on several university campuses. Sample Size and Power Analysis : A power analysis was conducted using G*Power to determine the required sample size, ensuring adequate power to detect anticipated effects (Supplementary Material 2). Ultimately, 88 participants (n = 58 females, 28 males, 5 non-binary, mean age ± SD = 22.98 ± 4.50) were included in the final analysis after accounting for exclusions. One participant was excluded due to a software issue, and 12 participants were excluded following the actors' video quality check (see the Procedure section for more details). 2. Materials Equipment and Software All online questionnaires were administered via the Qualtrics platform, and evaluations were conducted using OpenSesame ( version 4.0 27 ) on a Windows computer with a 23-inch Dell monitor (1920x1200 pixels), along with a keyboard and mouse. To facilitate manual responses, two stickers were placed on the keyboard, clearly indicating the designated keys for answering. SBL Room Setup A single room was used for both sessions, ensuring consistency in the learning environment. The chairs for the participant and the actor were positioned one meter apart—marked on the floor with tape—aligning with typical social spacing norms 28,29 . A clock was placed on a table visible to both the participant and the actor, helping them manage the timing of each simulation session. Finally, two 360-degree pivot cameras (AOSU - C2E) were installed to record both the actor and the participant. This setup enabled real-time monitoring by the experimenter and allowed for later review if needed, thereby ensuring the simulation sessions were conducted as intended. Simulation Actors Professional actors were recruited for both the pilot and main studies. These actors regularly collaborate with Hospital del Mar to conduct simulation-based learning (SBL) sessions for medical students. Over the course of several weeks, they participated in monitored and recorded training sessions, alongside regular meetings with the study’s supervisors. The primary objectives, before starting the experiment, were to ensure: ● Mastery of the clinical cases and associated symptoms, ● Consistent intensity and portrayal of symptoms across all actors, ● Minimal improvisation—restricted to prearranged anecdotes about how the symptoms affect the patient’s daily life. Stimuli To refine the clinical cases, determine the number of sessions, and calibrate the complexity of the evaluation task, a pilot study was conducted prior to the main study (see Supplementary Material 3). Symptom Design and Presentation . With the guidance of an external psychiatrist (Dr. Daniel Bergé), two distinct clinical cases were developed, each featuring a fictitious syndrome characterized by three core symptoms. These symptoms were designed to involve the same bodily part and a similar bodily movement. Each case began with a triggering situation, followed by the sequential presentation of Symptom 1, Symptom 2, and Symptom 3. Additionally, each symptom was paired with a corresponding control symptom. To ensure consistency, the level of difficulty and the written length of the symptoms were standardized across the two clinical cases, maintaining a uniform level of complexity. Evaluation Materials To evaluate symptom recall, all possible pairs were created, including combinations of symptoms, control symptoms, and mixed pairs (symptom + control symptom), resulting in 15 total combinations. Of these, only two sequences were correct (Symptom 1 → Symptom 2 and Symptom 2 → Symptom 3). Participants underwent evaluations in two distinct modalities: - Written Modality : Each combination displayed the first symptom above a downward arrow, and the second symptom below. - Video Modality : All symptoms and control symptoms were recorded with a different actor—one who did not take part in the SBL sessions—so that participants would not recognize them. Each video was standardized for elements such as attire, lighting, camera angle, and duration. During the combined video sequence, the first symptom was shown, followed by an arrow, and then the second symptom. Self-Report Measures Previous research has shown that social interactions can be influenced by changes in depression 30 , anxiety 31 , and emotional regulation strategies 32 . To account for these variables in our sample, participants completed several self-report measures: ● World Health Organisation–Five Well-Being Index (WHO-5) , assessing general well-being 33 . ● State Trait Anxiety Scales (STAI) , evaluating anxiety 34 . ● Scale of Positive and Negative Experience (SPANE) , measuring affect 35 . Additional questionnaires assessed socio-demographic factors, potential mental health diagnoses, medication use, and current stress levels (0–10 scale). However, because analyzing how these traits might influence behavior lies beyond our scope, we provide only descriptive statistics and correlations for these measures in Supplementary Material 4. Finally, given that previous research has shown that higher levels of embodied mental imagery can enhance learning outcomes 26 , we included a brief scenario-based measure of imagery, which asked participants to rate their ability to imagine key sensorimotor features (e.g., appearance, movement, voice) of a nonclinical figure and perceived physical contact with the patient in the clinical case. These ratings were combined into a single 10-point index reflecting the overall richness of participants’ mental simulation. 3. Procedure Overall Experimental Design This study employed a mixed design, incorporating both within-subjects and between-subjects factors. Participants were randomly assigned to one of two groups, ensuring systematic counterbalancing of learning conditions (SBL vs. text-based) and clinical cases. Two fictitious syndromes were developed in consultation with an external psychiatrist, each characterized by three core symptoms affecting the same bodily region and involving similar motor components. Each case followed a structured sequence: an initial triggering situation, followed by the progressive presentation of Symptom 1, Symptom 2, and Symptom 3. To assess symptom recall, participants completed an evaluation phase in two distinct modalities: written and video-based. The written modality presented symptom pairs in a structured format, while the video modality displayed pre-recorded symptom demonstrations by an unfamiliar actor. Each evaluation required participants to judge whether presented symptom sequences were correct, reinforcing the study’s focus on both item identification and sequential recall. To control for learning order effects, both symptom sequences and assessment modalities were counterbalanced across participants, ensuring that no systematic bias influenced the results. Participants were randomly assigned to one of two groups, and the simulation/control conditions as well as the two cases were systematically counterbalanced to minimize order effects and any task-specific biases. Within each modality (written or video), symptom-combination presentations were randomized, and the sequence of modalities themselves was counterbalanced across participants. To indicate whether a displayed combination was correct, participants pressed one of two marked keys (left or right). Moreover, the mapping of these keys (left/right = correct/incorrect) was reversed for half of the participants to avoid response biases. To ensure further balance, half of the participants completed the simulation condition during the first week and the control condition during the second week, while the remaining half did the opposite. Within each subgroup, half of the participants were assigned to case 1 initially and then switched to case 2 for the following week (and condition), whereas the other half followed the reverse order. This design guaranteed that both the conditions and pathologies were evenly distributed among participants. SBL Condition On the day of the experiment, each participant provided written informed consent upon arrival. They subsequently completed a series of questionnaires, including a sociodemographic questionnaire, the WHO-5, ERQ, STAI-Y (trait), stress assessments, and the imagery representation questionnaire, all administered via the Qualtrics platform. Participants were given a plastic sheet containing guiding questions (see Supplementary Material 5) designed to help them gather essential clinical information about the patient's symptoms. They were instructed to ask these questions during the interview, with the flexibility to reformulate them as needed. To ensure greater fluency, participants were given approximately ten minutes before the start of the experiment to familiarize themselves with the full set of questions. Additionally, they were informed that the patient did not display any severe or violent symptoms. Prior to entering the room (where the patient was already seated), the experimenter instructed participants to identify and memorize the patient’s symptom sequence through an interview. They were informed the patient might engage in tactile contact and that the sessions would be recorded; however, they were advised not to focus on the camera. Each session had to last between 10 and 12 minutes and around the 12-minute mark, the experimenter would knock on the door to signal the end of the session, while also monitoring via live video to confirm that the actor had finished describing the symptoms. Participants then completed STAI-Y (state) and SPANE questionnaires. Twenty-four hours later, the same procedure was repeated with the same actor to reinforce and deepen the information. After this second session, participants again completed the STAI-Y (state) and SPANE questionnaires, followed by the case evaluation in two counterbalanced formats: video and written. Control Condition Participants were tasked with identifying and memorizing a sequence of symptoms from a patient described in a text displayed on the screen. The video accompanying the text described one of the two clinical cases and mirrored the structure and information provided during the SBL sessions. The text appeared progressively on the screen while being narrated by an unfamiliar voice, ensuring the viewing duration was standardized at 10 minutes, comparable to the SBL sessions. Participants were instructed to remain focused throughout the presentation, without pausing or interruptions. Following this session, they completed the STAI-Y (anxiety) and SPANE (affect) questionnaires via the Qualtrics platform. Twenty-four hours later, participants took part in a second session that presented more details and information about the same clinical case, emphasizing the symptoms and their impact on the patient’s everyday life. After finishing this second session, participants once again filled out the anxiety-state and affect questionnaires and then carried out the evaluation of the pathology in both video and written modalities. Initial Baseline Assessment and one month Evaluation To ensure that participants engaged with the material comparably across conditions before progressing to the long-term retention phase, we conducted an initial baseline assessment immediately after the learning phase. This assessment served as a methodological check to confirm that there were no systematic differences between conditions, clinical cases, or counterbalancing factors. During the initial baseline assessment, participants were seated at a viewing distance of approximately 60 cm from the screen, in line with recommendations for comfortable vision 36 , with audio levels standardized across all participants. The experimenter provided verbal instructions, informing participants that they would view either written symptoms (written-modality evaluation) or short video clips (video-modality evaluation) on the screen. Participants were instructed to pay close attention to the order in which the symptoms appeared. Following this, participants read detailed on-screen instructions explaining that they would need to determine whether the displayed symptom sequence corresponded to the patient they had encountered during the previous two sessions (SBL or control condition). Each trial began with a fixation cross displayed at the center of the screen for 2 seconds. The symptom sequence was then presented for 10 seconds in the written modality or for 15–20 seconds in the video modality. After the sequence disappeared, a response screen prompted participants to indicate as quickly and accurately as possible whether the sequence was correct by pressing the designated key. Participants were then asked to rate their confidence in their response on a scale from 1 to 10. A 3-second interval separated each trial. The order of evaluations was counterbalanced, with some participants completing the written-modality evaluation first, followed by the video-modality evaluation, and others completing the evaluations in the reverse order. Measures reflecting performance speed (i.e., response time between the display of the answer screen and the keypress), accuracy (i.e., correct or incorrect keypress) and participants' self-reported confidence (rated using the mouse) were recorded for each trial. One-Month Session One month after the last session, participants performed an evaluation in both the video and written formats. The key difference in this session was that participants were required to identify which correct symptom sequences they had observed in any of the previous sessions (SBL and control combined). They were instructed to respond both rapidly and accurately, and to remain focused, as they could not pause the trial or revise their answers. After completing this evaluation, participants filled out the STAI-Y (state) and SPANE questionnaires for the final time. Quality Check of Videos of the simulation sessions. Whenever there was uncertainty about a given session’s adherence to instructions, we conducted a quality check of the video recordings to identify any potential issues related to noncompliance with the experimental protocol. For instance, some exclusions were due to the actor not performing the interaction correctly with the participant, such as not properly engaging in the sensorimotor interaction. 4. Data processing and statistical analyses All inferential tests were performed using RStudio (version 4.1.2 37 ). Using Tukey's method to detect extreme outliers (±3 × IQR), we replaced outlier trials with the overall mean reaction time calculated without including outliers. In the baseline evaluation session, 2.53% of the trials were replaced, while in the one-month evaluation session, 5.26% were replaced using this method 38,39 . All significance tests were two-sided, considered significant when p < 0.05. For repeated-measures analyses, if the sphericity assumption was violated, Greenhouse–Geisser corrections were applied. Effect sizes are reported as partial eta-squared (or eta-squared, as appropriate). We began by assessing whether demographic variables and counterbalancing factors influenced our primary outcomes in our initial Baseline Assessment. Using the aov function ( afex library), we examined the effects of gender, actor portraying the patient, lateralization of correct/incorrect responses on the keyboard, and written vs. video evaluation order on reaction time (RT) and accuracy. These analyses revealed no significant effects. We also confirmed that neither the Order of conditions (simulation first vs. control first), the Condition (simulation vs. control), nor the Clinical case (P1 vs. P2) had any observable impact on these measures. Detailed statistical outputs from this baseline assessment are provided in the Supplementary Material 6. Subsequently, our main interest lay in examining performance one month later, focusing on RT, accuracy, and confidence. To address deviations from normality in reaction time, we applied a logarithmic transformation and conducted standard diagnostic checks (e.g., residual normality, linearity, and homogeneity of variance) via the easystats package ( easystats library) in R. We employed linear mixed-effects models using lmer ( lmerTest package) and generalized linear mixed-effects models using glmer ( lme4 package) where appropriate. Model estimates and 95% confidence intervals for the fixed effects were extracted using the tidy function from the broom.mixed package. Separately, marginal means—with their standard errors and 95% confidence intervals computed via Satterthwaite-corrected degrees of freedom—were derived using the emmeans package for visualization. We generated the figures with ggplot2 . For RT, we fit a LME model with log-transformed RT as the dependent variable and Condition (simulation vs. control) and Modality (written vs. video) as fixed effects. We also included the Order (simulation-control vs control-simulation) and Clinical cases (P1 vs P2) to account for potential confounding factors. Moreover, we incorporated a random intercept to control for subject-specific differences (Formula 1). Reaction Time ~ Condition + Modality + Order + Clinical case + (1|ID) Accuracy (0 = incorrect, 1 = correct) was analyzed via a generalized linear mixed-effects model (binomial link). We used Condition and Modality as fixed effects and we also included the Order of conditions and the Clinical case to account for possible confounding factors. Lastly, to account for subject and stimuli specific differences, we incorporated random intercepts into the models (Formula 2) Accuracy ~ Condition + Modality + Order + Clinical case + (1 | ID) + (1 | STIM) Confidence was categorized into absolute (100%) certainty versus anything below 100%. Separating them thus allowed us to capture potential qualitative differences between those who were entirely certain and those who expressed any level of uncertainty, reflecting a distinctly different cognitive state. Given the large cluster of self-reported scores at 100%, this approach also balanced the data for logistic regression. To ensure the reliability of our logistic regression model and in line with the Events Per Variable (EPV) recommendations 40,41 , we confirmed that our model maintained an adequate EPV and that predictor variables were well-balanced, eliminating the need for penalized likelihood corrections (for further details, see the Supplementary Material 7). Hence, we fitted GLME model using confidence as the dependent variable and Condition as well as Modality as fixed effects. To account for confounding factors, we also included the Order of conditions and the Clinical case, and we incorporated random intercepts to control for subject and stimuli specific differences (Formula 3). Confidence ~ Condition + Modality + Order + Clinical case + (1 | ID) + (1 | STIM) We employed the DHARMa package to evaluate model adequacy in both accuracy and confidence analyses. Residual plots, dispersion tests, checks for zero-inflation, and outlier detection indicated no major violations of model assumptions. We further verified our classification approach by examining the ROC curve and computing the AUC for accuracy and confidence, confirming the appropriateness of these binary outcomes. Complete statistical outputs, diagnostic plots, and the corresponding R code are provided in the Supplementary Material. Finally, we examined the effect of the imagery score on reaction time and accuracy during the one-month evaluation. Due to missing responses, this analysis was conducted on a sample of 82 participants (n = 56 females, 22 males, 4 non-binaries; mean age ± SD = 22.96 ± 4.52). The imagery score was standardized (mean-centered and scaled) to facilitate interpretation and to ensure comparability across participants. We first fit a similar LME model using log-transformed RT as the dependent variable and an interaction between Modality and the Imagery score as fixed effects (Formula 4): Reaction Time ~ Condition + Modality * Imagery score + Order + Clinical case + (1|ID) Then, we fit a similar GLME model with accuracy as the dependent variable and an interaction between Modality et the imagery score as fixed factors (Formula 5) Accuracy ~ Condition + Modality * Imagery score + Order + Clinical case + (1 | ID) + (1 | STIM) III. Results Effects of Condition, Order, Modality, and Imagery on Reaction Time Following the approach described in the Statistical Analysis section (Formula 1), we fit a linear mixed-effects model to investigate the effects of Condition (simulation vs. control), Modality (written vs. video), Order (simulation-control vs control-simulation), and Clinical case (P1 vs P2) on log-transformed reaction times. The model converged successfully, and diagnostic checks (e.g., residual normality, homogeneity of variance, and collinearity) indicated a satisfactory model fit. We observed a significant main effect of Condition, with participants in the control condition exhibiting longer reaction times than those in the simulation condition (Estimate = 0.066, 95% CI [0.03, 0.10], SE = 0.017, t = 3.90, p < 0.001). There was also a significant main effect of Order (Estimate = 0.169, 95% CI [0.02, 0.32], SE = 0.074, t = 2.27, p = 0.026), indicating that participants tested in the simulation-control sequence responded more slowly than those in the control-simulation sequence. Neither Modality (Estimate = 0.027, 95% CI [–0.01, 0.06], SE = 0.017, t = 1.57, p = 0.116) nor Clinical case (Estimate = 0.043, 95% CI [–0.10, 0.19], SE = 0.074, t = 0.58, p = 0.566) reached statistical significance. The fixed effects explained approximately 1.7% of the variance in log-RT (marginal R² = 0.017), whereas including random intercepts for participants increased the explained variance to about 24.4% (conditional R² = 0.244). Relation between Reaction time and Imagery score We focus here solely on the results concerning the imagery score. For the full model results, please refer to the Supplementary Material 8 (Table S4). We did not observe a significant main effect of the imagery score on reaction times (Estimate = -0.037, 95% CI [-0.08, 0.01], SE = 0.023, t = -1.57, p = 0.121), suggesting that individual differences in imagery ability alone did not influence response times. However, we observed a significant interaction between Modality and the imagery score, indicating that the effect of the modality of presentation on reaction times varied as a function of participants’ ability to generate mental imagery (Estimate = -0.061, 95% CI [-0.08, -0.04], SE = 0.010, t = -5.88, p < 0.001). Effects of Condition, Order, Modality, and Imagery on Accuracy Using the generalized linear mixed-effects model described in the Statistical Analysis section (Formula 2), we examined how Condition (simulation vs. control), Modality (written vs. video), Order (simulation-control vs control-simulation), and Clinical case (P1 vs P2) influenced the likelihood of a correct response. The model converged successfully, diagnostics indicated no issues with overdispersion or zero-inflation, and outlier tests did not reveal any problematic observations. We observed a significant main effect of Condition, indicating that participants in the simulation condition exhibited higher accuracy than those in the control condition (Estimate = -0.456, SE = 0.097, z = -4.71, p < 0.001, corresponding to OR = 0.63, 95% CI [0.52, 0.77]). A main effect of Order also emerged, showing that participants tested in the simulation-control sequence had lower accuracy than those in the control-simulation sequence (Estimate = –0.837, SE = 0.343, z = –2.44, p = 0.015; OR = 0.43, 95% CI [0.22, 0.85]). Neither Modality (Estimate = -0.059, SE = 0.096, z = -0.61, p = 0.539, OR = 0.94, 95% CI [0.78, 1.14]) nor Clinical case (Estimate = 0.203, SE = 0.338, z = 0.60, p = 0.548, OR = 1.23, 95% CI [0.63, 2.38]) reached statistical significance. Model fit was further supported by an Area Under the Curve (AUC) of 0.901, reflecting an excellent discriminative ability. Regarding effect sizes, the marginal R² was approximately 3.2%, while the conditional R² rose to around 55%. Relation between Accuracy and Imagery score We focus here solely on the results concerning the imagery score. For the full model results, please refer to the Supplementary Material 8 (Table S5). Participants with higher imagery scores exhibited greater accuracy overall (Estimate = 0.266, SE = 0.114, z = 2.34, p = 0.019; OR = 1.30, 95% CI [1.04, 1.63]). Moreover, a significant interaction between imagery score and modality indicated that the relationship between imagery score and accuracy varied across modalities ( Estimate = -0.186, SE = 0.058, z = -3.22, p = 0.001; OR = 0.83, 95% CI [0.74, 0.93]). Specifically, the positive effect of imagery score on accuracy was more pronounced in the written condition than in the video condition. Effects of Condition, Order and Modality on Confidence Using the generalized linear mixed-effects model described in the Statistical Analysis section (Formula 3), we examined how Condition (simulation vs. control), Modality (written vs. video), Order (simulation-control vs control-simulation), and Clinical case (P1 vs P2) impacted participants' confidence in their responses. Diagnostic checks showed no indications of singularity, over/under-dispersion, zero-inflation, or problematic outliers. A significant main effect of Condition emerged, indicating that participants in the control condition were substantially less likely to report absolute confidence compared to those in the simulation condition (Estimate = –1.14, SE = 0.08, z = –13.68, p < 0.001; OR = 0.32, 95% CI [0.27, 0.38]). Modality showed a marginal effect, with participants in the video condition being slightly more likely to report the highest certainty compared to those in the written condition (Estimate = 0.15, SE = 0.08, z = 1.90, p = 0.058; OR = 1.16, 95% CI [1.00, 1.36]). Neither Order (Estimate = -0.35, SE = 0.57, z = -0.61, p = 0.541; OR = 0.70, 95% CI [0.23, 2.17]) nor Clinical case (Estimate = -0.36, SE = 0.57, z = -0.63, p = 0.527; OR = 0.70, 95% CI [0.23, 2.13]) reached statistical significance. The model showed excellent predictive accuracy, as reflected by an AUC of 0.91, indicating a high ability to discriminate between confidence levels. Regarding effect sizes, the marginal R² was approximately 3.5%, whereas the conditional R² rose to about 71.1%. IV. Discussion This study, designed as a randomized controlled trial using a within-subjects design, examined the impact of a simulation-based protocol on the acquisition and long-term retention of new knowledge by naïve learners. A key strength of this study lies in its careful design, which controlled for confounding factors and employed a counterbalanced approach, ensuring robust comparisons between learning conditions. By doing so, we were able to isolate the specific contribution of simulation-based, immersive, and embodied interaction to long-term learning outcomes. Participants engaged with an actor who portrayed a clinical case, experiencing the symptoms dynamically through physical interaction, while a matched text-based learning condition provided identical information in a non-interactive, passive format. One month after learning, participants exhibited significantly faster reaction times and higher accuracy when recalling clinical symptom sequences in the simulation-based condition compared to the text-based condition. Interestingly, an order effect emerged: when participants began with the control condition, their subsequent performance in the simulation condition was enhanced, whereas starting with the simulation condition did not provide the same benefit for the control condition. This suggests a potential beneficial transfer effect that enhances performance in the simulation condition when it follows the text-based condition, while the control condition undergoes a disadvantageous transfer effect when it follows the simulation condition. Beyond these effects, we also observed that participants were significantly more likely to report absolute confidence in the SBL condition. Moreover, confidence and accuracy were closely aligned in the SBL condition, with higher confidence accompanying higher accuracy. This contrasts with the control condition, where accuracy was relatively high, but probability of absolute confidence remained moderate, potentially leading to a more tentative application of knowledge. Our findings support previous research demonstrating the benefits of SBL in medical and STEM education 3,4 , although they moderate the advantages of SBL over more traditional text-based methods in learning outcomes. Consistent with prior work, SBL appears to improve knowledge retention, reaction time, and retrieval accuracy, reinforcing the idea that immersive, embodied, and interactive experiences create stronger memory traces than passive learning methods 42,43 . Furthermore, research emphasizing the role of embodiment and interactivity in cognitive processes 5 suggests that learning through active engagement with a simulated patient reinforces memory processes, especially by leveraging sensorimotor integration, a mechanism shown to facilitate encoding and retrieval 44,45 . However, our study also provides a more nuanced view of SBL’s effectiveness by isolating its specific contributions through a counterbalanced within-subjects design. While SBL demonstrated an advantage, text-based learning can still effectively encode and retrieve information, at least in the domain of knowledge acquisition where information is systematically organized and sequentially presented. The observed differences between conditions might nevertheless have been moderated by the role of mental imagery and the assessment modality. On the one hand, the simulation condition might have been disadvantaged by the retrieval assessment being presented in a video format, attenuating the immersive and embodied benefits of the original learning experience. This aligns with prior research suggesting that the format of retrieval assessments can impact memory performance, particularly when there is a mismatch between learning and retrieval conditions 12 . Conversely, participants with higher imagery capacity may have been able to mentally simulate the clinical case in the written evaluation modality, facilitating their retrieval process. Mental imagery has been shown to engage neural mechanisms akin to actual perception, supporting perceptual learning without direct sensory input 25 . Additionally, it strengthens recall for abstract concepts when linked to sensorimotor experiences 26 . This suggests that internal simulation can compensate for the absence of direct interaction. An additional key finding of this study is the effect of condition order. For SBL, an initial exposure to the material in a structured, traditional format may facilitate later embodied learning. This finding confirms previous research 46–48 suggesting that simulation-based learning benefits from prior contextualization or introduction, which may help structure the learning experience before engagement in an immersive and interactive environment. Conversely, for text-based learning, experiencing an immersive, interactive session beforehand may reduce engagement or disrupt retention, possibly because the shift from an active, sensory-rich experience to a more passive format is less conducive to sustained cognitive engagement. This aligns with prior research indicating that immersive, embodied learning can lead to deeper but more rigid knowledge acquisition, potentially hindering subsequent learning through a more abstract and static format such as text-based instruction 49,50 . Immersive learning also places high demands on cognitive resources, engaging sensorimotor, emotional, and social processes. From a cognitive load perspective, this intense engagement early in learning may create a form of anchoring or cognitive rigidity, making it harder to adapt to less interactive and more abstract formats 49 , mirroring the proactive interference observed by Saylık (2021) 50 . Consequently, when working memory processing is hindered by cognitive overload, the consolidation, retrieval, and transfer of information become significantly more challenging, highlighting the advantage of a sequential, low-to-high complexity approach that progressively builds learners' long-term memory stores 48 . While this remains speculative, future research should explore whether such order effects are driven by differences in attentional or cognitive load across modalities, from proactive interference, or shifts in engagement 20 across learning modalities. We also observed that participants in the simulation condition were three times more likely to report absolute confidence than those in the control condition. One possible explanation may stem from the immersive nature of SBL, which fosters deeper sensorimotor engagement and internal simulation. As suggested by the sensorimotor model of memory, this process may facilitate the reactivation of sensorimotor patterns associated with the learned material. Specifically, the richer sensory and contextual cues provided by SBL, compared to the control condition, may act as markers of a more robust internal feedback mechanism, strengthening the link between knowledge verification and retrieval 20 . In addition, in our results, confidence and accuracy were closely aligned in the simulation condition, with higher confidence accompanying higher accuracy. This alignment suggests that SBL not only enhances knowledge retention but also reinforces learners' trust in their acquired knowledge. In contrast, the control condition exhibited a mismatch, where accuracy was relatively high, but absolute confidence probability remained moderate. This discrepancy may indicate a more tentative application of knowledge, as learners might hesitate despite correctly recalling information. While confidence is generally beneficial 51–53 , its relationship with accuracy is complex 54 . When confidence and competence are well-calibrated, learners are more likely to apply their knowledge effectively 51 . However, when confidence outpaces accuracy, overconfidence can lead to errors and misjudgments 54,55 . Therefore, SBL can be seen as a method that may foster both accuracy and self-confidence, creating a learning environment where knowledge is not only acquired but also more efficiently applied than with more traditional methods of learning. This combination of accuracy and confidence may be particularly valuable in settings where both competence and the conviction to act on one's knowledge are essential. Moreover, by aligning confidence with performance, SBL may help mitigate the risks of overconfidence. Future studies should aim to replicate these findings and further investigate the mechanisms underlying the observed alignment between high absolute confidence probability and accuracy in SBL. Specifically, research should explore how sensorimotor engagement and internal simulation contribute to learners' self-assessment and whether the enhanced sensory and contextual cues in SBL function as an internal feedback mechanism that strengthens learning outcomes. Additionally, examining the conditions under which confidence and accuracy remain well-calibrated in SBL, as well as identifying potential boundary effects, would provide deeper insights into the robustness and generalizability of these effects. Interestingly, our results demonstrate that despite the advantages of SBL, text-based learning effectively facilitated item identification and sequential recall of realistic, embodied knowledge, enabling learners to transfer this information to recognition tasks involving realistic symptom presentations. Related to this, our findings suggest a possible role of mental imagery in knowledge retrieval when confronted with its realistic depiction. Participants with stronger imagery abilities performed better, suggesting that the ability to mentally simulate content may be an important predictor of retrieval efficiency. This finding extends beyond SBL and may have implications for understanding individual differences in learning across various modalities, such as virtual reality training, medical case studies, or language learning, where visualization plays a key role in comprehension and recall. Regarding limitations, a strength, but also a weakness, of this study is that it specifically targeted the realistic, immersive, embodied, and interactive dimensions of SBL in a lab setting. However, SBL also encompasses other crucial components, such as strategic decision-making, dynamic feedback, and iterative practice 56 . These elements, central to experiential learning theories 57–59 , contribute significantly to SBL’s pedagogical value. While our study did not capture the full spectrum of SBL’s educational benefits, isolating its embodied and interactive aspects is an important step toward understanding the mechanisms underlying its effectiveness. Additionally, although the study demonstrated an advantage of SBL over text-based learning within a one-month retention period, further research is needed to assess the long-term durability of these effects. Future studies should examine whether the observed benefits persist beyond one month and whether different retrieval modalities influence retention over extended periods. Another limitation concerns the nature of the retrieval assessments. Since knowledge retention was tested using video assessments rather than direct interactions with a live actor, the transferability of learning from SBL to real-world settings may have been attenuated. It is possible that a more immersive retrieval condition, such as role-playing with a live patient, would have amplified the benefits of SBL by preserving its embodied and interactive elements. Finally, the order effect, while intriguing, requires further validation. The fact that simulation benefited from an initial text-based session while text-based learning was negatively impacted by starting with simulation suggests that cognitive flexibility, engagement shifts, or proactive interference may play a role. Future studies should explore whether this effect generalizes to other educational contexts and whether instructional sequencing can be strategically optimized to enhance learning outcomes. Conclusion This study provides empirical evidence supporting the benefits of simulation-based learning over traditional text-based learning for the acquisition and long-term retention of clinical knowledge. While SBL enhances learning, our results suggest that structured, text-based methods can also yield strong retention outcomes, particularly for item identification and sequential recall. By isolating the effects of realism, immersion, embodiment, and interactivity in SBL, our study provides a more precise characterization of the mechanisms that drive its effectiveness. Identifying key factors that enhance learning outcomes enables future learning protocols, whether simulation-based or not, to incorporate principles of interactivity and embodiment, maximizing their effectiveness. This refined understanding allows for better-informed design of SBL interventions and offers insights that can be transferred to non-SBL learning environments. Future research building on these findings will help create more tailored, evidence-based approaches to enhance learning through immersive and interactive experiences. Declarations Data availability All data are available at the following OSF repository: https://osf.io/ub6ce/?view_only=6ea1d137e1bf41079bdc9b20cb8df7fe Code availability Data preprocessing and analysis scripts are available at the following OSF repository: https://osf.io/ub6ce/?view_only=6ea1d137e1bf41079bdc9b20cb8df7fe Acknowledgements We thank Laura Alsina, Mila Markovic, and Anna Yue for their assistance during the experimental sessions, as well as the actors Mar Casas, Berta Graells, Mirena Nafarrete, Arnau Armengol and Carla Vilaró for their dedication. Special thanks to Dr. Meritxell Girvent and Dr. Carolina Pérez García for their assistance in designing the simulation protocol, and to Dr. Daniel Bergé for his contribution to developing the fictitious clinical syndromes. This work was supported by Fundación Tatiana. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. Author information These authors jointly supervised this work: Angélique Lebert and Óscar Vilarroya Authors and Affiliations Unitat de Recerca en Neurociència Cognitiva Departament de Psiquiatria i Medicina Legal Universitat Autònoma de Barcelona Spain and Hospital del Mar Research Institute, Barcelona Spain Angélique Lebert and Óscar Vilarroya Contributions (to complete/adapt) AL: Conceptualization, Methodology, Software, Formal Analysis, Investigation, Writing. OV: Conceptualization, Methodology, Supervision, Project administration, Resources, Funding acquisition, Writing. Corresponding authors Correspondence to Óscar Vilarroya References Frasson, C. & Blanchard, E. G. Simulation-Based Learning. in Encyclopedia of the Sciences of Learning (ed. Seel, N. M.) 3076–3080 (Springer US, Boston, MA, 2012). doi:10.1007/978-1-4419-1428-6_129. Gormley, G. J. & Murphy, P. When I say … simulation. Medical Education 57 , 1182–1183 (2023). Chernikova, O. et al. Simulation-Based Learning in Higher Education: A Meta-Analysis. Review of Educational Research 90 , 499–541 (2020). Lateef, F. Simulation-based learning: Just like the real thing. J Emerg Trauma Shock 3 , 348–352 (2010). Lebert, A. & Vilarroya, Ó. The links between experiential learning and 4E cognition. Annals of the New York Academy of Sciences 1541 , 37–52 (2024). Lee, C. A., Pais, K., Kelling, S. & Anderson, O. S. A scoping review to understand simulation used in interprofessional education. Journal of Interprofessional Education & Practice 13 , 15–23 (2018). Kaplonyi, J. et al. Understanding the impact of simulated patients on health care learners’ communication skills: a systematic review. Medical Education 51 , 1209–1219 (2017). Alharbi, A., Nurfianti, A., Mullen, R. F., McClure, J. D. & Miller, W. H. The effectiveness of simulation-based learning (SBL) on students’ knowledge and skills in nursing programs: a systematic review. BMC Medical Education 24 , 1099 (2024). Clark, A. D. et al. The Effect of 3-Dimensional Simulation on Neurosurgical Skill Acquisition and Surgical Performance: A Review of the Literature. Journal of Surgical Education 74 , 828–836 (2017). von Wendt, C. E. A. & Niemi-Murola, L. Simulation in Interprofessional Clinical Education: Exploring Validated Nontechnical Skills Measurement Tools. Simulation in Healthcare 13 , 131 (2018). McGaghie, W. C., Issenberg, S. B., Cohen, E. R., Barsuk, J. H. & Wayne, D. B. Does Simulation-Based Medical Education With Deliberate Practice Yield Better Results Than Traditional Clinical Education? A Meta-Analytic Comparative Review of the Evidence. Academic Medicine 86 , 706 (2011). Hofmann, R., Curran, S. & Dickens, S. Models and measures of learning outcomes for non-technical skills in simulation-based medical education: Findings from an integrated scoping review of research and content analysis of curricular learning objectives. Studies in Educational Evaluation 71 , 101093 (2021). Manzi, A. & Nigro, G. Long-term memory for performed and observed actions: Retrieval awareness and source monitoring. Memory 16 , 595–603 (2008). Schult, J., Stülpnagel, R. von & Steffens, M. C. Enactment versus Observation: Item-Specific and Relational Processing in Goal-Directed Action Sequences (and Lists of Single Actions). PLOS ONE 9 , e99985 (2014). Convertino, G. et al. Positive and negative touch differentially modulate metacognitive memory judgements for emotional stimuli. British Journal of Psychology 116 , 34–51 (2025). Della Longa, L., Gliga, T. & Farroni, T. Tune to touch: Affective touch enhances learning of face identity in 4-month-old infants. Developmental Cognitive Neuroscience 35 , 42–46 (2019). Cook, S. W., Mitchell, Z. & Goldin-Meadow, S. Gesturing makes learning last. Cognition 106 , 1047–1058 (2008). Dijkstra, K., Kaschak, M. P. & Zwaan, R. A. Body posture facilitates retrieval of autobiographical memories. Cognition 102 , 139–149 (2007). Roberts, B. R. T., MacLeod, C. M. & Fernandes, M. A. The enactment effect: A systematic review and meta-analysis of behavioral, neuroimaging, and patient studies. Psychol Bull 148 , 397–434 (2022). Ianì, F. Embodied memories: Reviewing the role of the body in memory processes. Psychon Bull Rev 26 , 1747–1766 (2019). Humphreys, M. S., Hockley, W. E. & Chalmers, K. A. Recognition memory: The probe, the returned signal, and the decision. Psychon Bull Rev 31 , 568–598 (2024). Hurlstone, M. J. Serial Recall. in The Oxford Handbook of Human Memory, Two Volume Pack: Foundations and Applications (eds. Kahana, M. J. & Wagner, A. D.) 0 (Oxford University Press, 2024). doi:10.1093/oxfordhb/9780190917982.013.28. Logan, G. D. & Cox, G. E. Serial order depends on item-dependent and item-independent contexts. Psychological Review 130 , 1672–1687 (2023). Mack, M. L. & Palmeri, T. J. Discrimination, Recognition, and Classification. in The Oxford Handbook of Human Memory, Two Volume Pack: Foundations and Applications (eds. Kahana, M. J. & Wagner, A. D.) 0 (Oxford University Press, 2024). doi:10.1093/oxfordhb/9780190917982.013.30. Tartaglia, E. M., Bamert, L., Herzog, M. H. & Mast, F. W. Perceptual learning of motion discrimination by mental imagery. Journal of Vision 12 , 14 (2012). Marre, Q., Huet, N. & Labeye, E. Embodied mental imagery improves memory. Quarterly Journal of Experimental Psychology 74 , 1396–1405 (2021). Mathôt, S., Schreij, D. & Theeuwes, J. OpenSesame: An open-source, graphical experiment builder for the social sciences. Behav Res 44 , 314–324 (2012). Hecht, H., Welsch, R., Viehoff, J. & Longo, M. R. The shape of personal space. Acta Psychologica 193 , 113–122 (2019). Lebert, A., Vergilino-Perez, D. & Chaby, L. Keeping distance or getting closer: How others’ emotions shape approach-avoidance postural behaviors and preferred interpersonal distance. PLoS One 19 , e0298069 (2024). Grèzes, J., Risch, N., Courtet, P., Olié, E. & Mennella, R. Depression and approach-avoidance decisions to emotional displays: The role of anhedonia. Behav Res Ther 164 , 104306 (2023). Knyazev, G. G., Bocharov, A. V., Slobodskaya, H. R. & Ryabichenko, T. I. Personality-linked biases in perception of emotional facial expressions. Personality and Individual Differences 44 , 1093–1104 (2008). Butler, E. A. et al. The social consequences of expressive suppression. Emotion 3 , 48–67 (2003). Topp, C. W., Østergaard, S. D., Søndergaard, S. & Bech, P. The WHO-5 Well-Being Index: A Systematic Review of the Literature. Psychother Psychosom 84 , 167–176 (2015). Buela-Casa, G., Guillén-Riquelme, A. & Seisdedos Cubero, N. Cuestionario de Ansiedad Estado-Rasgo: Adaptación espanola. Madrid: TEA Ediciones (2011). Espejo, B., Checa, I., Perales-Puchalt, J. & Lisón, J. F. Validation and Measurement Invariance of the Scale of Positive and Negative Experience (SPANE) in a Spanish General Sample. International Journal of Environmental Research and Public Health 17 , 8359 (2020). Rempel, D., Willms, K., Anshel, J., Jaschinski, W. & Sheedy, J. The Effects of Visual Display Distance on Eye Accommodation, Head Posture, and Vision and Neck Symptoms. Hum Factors 49 , 830–838 (2007). Team, Rs. RStudio: integrated development for R. RStudio, PBC. (2020). Tukey, J. W. Exploratory data analysis. Reading/Addison-Wesley (1977). Zellner, A., Richardson, A. M., Lidbury, B. A., Hobson, P. & Badrick, T. An Investigation into Outlier Elimination and Calculation Methods in the Determination of Reference Intervals using Serum Immunoglobulin A as a Model Data Collection. Preprint at https://doi.org/10.48550/arXiv.1907.09637 (2019). Peduzzi, P., Concato, J., Kemper, E., Holford, T. R. & Feinstein, A. R. A simulation study of the number of events per variable in logistic regression analysis. Journal of Clinical Epidemiology 49 , 1373–1379 (1996). Van Smeden, M. et al. Sample size for binary logistic prediction models: Beyond events per variable criteria. Stat Methods Med Res 28 , 2455–2474 (2019). Brodt, S. et al. Rapid and independent memory formation in the parietal cortex. Proceedings of the National Academy of Sciences 113 , 13251–13256 (2016). Kumar, N., Manning, T. F. & Ostry, D. J. Somatosensory cortex participates in the consolidation of human motor memory. PLoS Biol 17 , e3000469 (2019). Kruger, J. S., Kruger, D. J. & Suzuki, R. Assessing the Effectiveness of Experiential Learning in a Student-Run Free Clinic. Pedagogy in Health Promotion 1 , 91–94 (2015). Waddell, A., Lennox, A., Spassova, G. & Bragge, P. Barriers and facilitators to shared decision-making in hospitals from policy to practice: a systematic review. Implementation Science 16 , 74 (2021). Eppich, W. & Cheng, A. Promoting Excellence and Reflective Learning in Simulation (PEARLS): Development and Rationale for a Blended Approach to Health Care Simulation Debriefing. Simulation in Healthcare 10 , 106 (2015). Rudolph, J. W., Raemer, D. B. & Simon, R. Establishing a Safe Container for Learning in Simulation: The Role of the Presimulation Briefing. Simulation in Healthcare 9 , 339 (2014). Fraser, K. L., Ayres, P. & Sweller, J. Cognitive load theory for the design of medical simulations. Simulation in Healthcare 10 , 295–307 (2015). Huang, C. L., Luo, Y. F., Yang, S. C., Lu, C. M. & Chen, A.-S. Influence of Students’ Learning Style, Sense of Presence, and Cognitive Load on Learning Outcomes in an Immersive Virtual Reality Learning Environment. Journal of Educational Computing Research 58 , 596–615 (2020). Saylik, R., Castiello, S. & Murphy, R. A. The role of emotional interference on learning in an emotional probabilistic Go/No-Go task. Dusunen Adam: Journal of Psychiatry & Neurological Sciences 34 , (2021). Colquitt, J. A., LePine, J. A. & Noe, R. A. Toward an integrative theory of training motivation: A meta-analytic path analysis of 20 years of research. Journal of Applied Psychology 85 , 678–707 (2000). Judge, T. A. & Bono, J. E. Relationship of core self-evaluations traits—self-esteem, generalized self-efficacy, locus of control, and emotional stability—with job satisfaction and job performance: A meta-analysis. Journal of Applied Psychology 86 , 80–92 (2001). Richardson, M., Abraham, C. & Bond, R. Psychological correlates of university students’ academic performance: A systematic review and meta-analysis. Psychological Bulletin 138 , 353–387 (2012). Koriat, A. The self-consistency model of subjective confidence. Psychological Review 119 , 80–113 (2012). Dunning, D. Chapter five - The Dunning–Kruger Effect: On Being Ignorant of One’s Own Ignorance. in Advances in Experimental Social Psychology (eds. Olson, J. M. & Zanna, M. P.) vol. 44 247–296 (Academic Press, 2011). Pazin Filho, A. & Romano, M. M. D. SIMULAÇÃO: ASPECTOS CONCEITUAIS. in Medicina (Ribeirão Preto) vol. 40 167–170 (2007). Fenwick, T. J. Expanding Conceptions of Experiential Learning: A Review of the Five Contemporary Perspectives on Cognition. Adult Education Quarterly 50 , 243–272 (2000). Itin, C. M. Reasserting the Philosophy of Experiential Education as a Vehicle for Change in the 21st Century. Journal of Experiential Education 22 , 91–98 (1999). Kolb, D. A. Experiential Learning: Experience as the Source of Learning and Development . (FT Press, 2014). Additional Declarations No competing interests reported. Supplementary Files LebertVilarroyaSupplementaryinformation.docx Cite Share Download PDF Status: Published Journal Publication published 24 Nov, 2025 Read the published version in npj Science of Learning → Version 1 posted Editorial decision: Revision requested 22 Sep, 2025 Reviews received at journal 22 Sep, 2025 Reviewers agreed at journal 11 Aug, 2025 Reviewers agreed at journal 24 Jul, 2025 Reviews received at journal 29 Apr, 2025 Reviews received at journal 25 Apr, 2025 Reviewers agreed at journal 24 Apr, 2025 Reviews received at journal 22 Apr, 2025 Reviewers agreed at journal 28 Mar, 2025 Reviewers agreed at journal 26 Mar, 2025 Reviewers invited by journal 26 Mar, 2025 Editor assigned by journal 26 Mar, 2025 Submission checks completed at journal 19 Mar, 2025 First submitted to journal 18 Mar, 2025 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-6253202","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":438763960,"identity":"9ba9e380-05ce-4aa1-996b-d1885cb4efdf","order_by":0,"name":"Angélique Lebert","email":"","orcid":"","institution":"Universitat Autonoma de Barcelona","correspondingAuthor":false,"prefix":"","firstName":"Angélique","middleName":"","lastName":"Lebert","suffix":""},{"id":438763961,"identity":"603439fa-3144-4f4f-a8f0-63d6c79ffbfe","order_by":1,"name":"Oscar Vilarroya","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA9klEQVRIiWNgGAWjYBADHj4oQw7IJlILG5RhDNViQFgPTEtiAyEt/PyHH36uYDgsw8Z+9tmHj2126RuOnz3A8KPmD04tkjPSjCXPMBzmYeNJN545sy05d8OZvATGnmO4bTG4wcMg2cCQBvRLGjMzbxtz7oYDOQbMDGx4tJw/w/wTrIX/GUhLfbrB+TdALf/waDmQwwa0xYaHTQJsy+EEgxtAWxjbcGsB+sXMssEApOUZM+OMc8cNZ954l3Cwt88YpxZgiD2+2VAhYc/Pn8bM8KGsWp7vfO7BBz++yeHUAnUelGaExs4BAuqRAe7YGAWjYBSMghEMAIX0SHxST7rxAAAAAElFTkSuQmCC","orcid":"","institution":"Universitat Autonoma de Barcelona","correspondingAuthor":true,"prefix":"","firstName":"Oscar","middleName":"","lastName":"Vilarroya","suffix":""}],"badges":[],"createdAt":"2025-03-18 12:38:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6253202/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6253202/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41539-025-00380-9","type":"published","date":"2025-11-24T15:57:13+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":80665765,"identity":"9b210399-11f7-4217-a540-ac9f34067939","added_by":"auto","created_at":"2025-04-15 17:32:39","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":261745,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSchematic representation of the experimental procedure\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eThis figure illustrates the arrangement of the simulation room (Unity environment, http://www.unity3D.com), alongside the presentation of clinical cases in the control condition across the two learning sessions. During each evaluation phase, participants were asked to judge whether a displayed symptom sequence (video or written format) was correct and to rate their confidence in their response. Both conditions and clinical cases were systematically counterbalanced across participants to ensure balanced exposure.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6253202/v1/ec4ff5df9b8c1b681184fe45.png"},{"id":80665243,"identity":"8dba2d73-fbfc-47d4-b889-b0d616d4e67f","added_by":"auto","created_at":"2025-04-15 17:24:39","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":70895,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eReaction Time\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cstrong\u003eA:\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Mean log-transformed reaction times (RT) by Condition (simulation in color vs. control in gray); error bars represent within‐subject 95% confidence intervals. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003eB:\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Mean log-transformed RTs by Condition and Order (simulation–control in color vs. control–simulation in gray); error bars indicate within‐subject 95% confidence intervals.\u003c/em\u003e\u003cem\u003e\u003cstrong\u003e C:\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Predicted log-transformed RTs as a function of participants’ centered imagery score and Modality (video in color vs. written in gray), with shaded regions representing 95% confidence intervals. Estimates are derived from linear mixed-effects models including random intercepts for participants.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6253202/v1/d81a63377aea08bd7499cbb6.png"},{"id":80665244,"identity":"f3117b55-65a6-44e5-b7f9-49b9a6ff0127","added_by":"auto","created_at":"2025-04-15 17:24:39","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":80110,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAccuracy\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cstrong\u003eA:\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Predicted probability of correct responses (accuracy) by Condition (simulation in color vs. control in gray); error bars represent within‐subject 95% confidence intervals. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003eB:\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Predicted accuracy by Condition and Order (simulation–control in color vs. control–simulation in gray); error bars indicate within‐subject 95% confidence intervals. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003eC: \u003c/strong\u003e\u003c/em\u003e\u003cem\u003ePredicted accuracy as a function of participants’ centered imagery score and Modality (video in color vs written in gray), with shaded regions representing 95% confidence intervals. Estimates are derived from generalized linear mixed-effects models including random intercepts for participants and for stimuli.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6253202/v1/a2ceb8cab1867ba2bc96b70e.png"},{"id":80665764,"identity":"e875d36e-a3c7-40f3-b2c5-48dbc9f6d56d","added_by":"auto","created_at":"2025-04-15 17:32:39","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":36317,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConfidence outcome\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cstrong\u003ePanels A-B\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e display results for Confidence. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003eA:\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e Mean confidence ratings by Condition (simulation case in color vs. control case in gray); vertical bars indicate within‐subject 95% confidence intervals. \u003c/em\u003e\u003cem\u003e\u003cstrong\u003eA2\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e. Mean confidence ratings by Condition (simulation case in color vs. control case in gray) and Modality; vertical bars indicate within‐subject 95% confidence intervals.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6253202/v1/daef9d8979b22167a2ba798e.png"},{"id":97178765,"identity":"1c6ad910-6908-4988-b949-e56951e25f05","added_by":"auto","created_at":"2025-12-01 16:13:27","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1571211,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6253202/v1/3952f758-5255-4d9f-b85e-af9391e4e947.pdf"},{"id":80665240,"identity":"72754388-30e3-4cec-a634-2d67088e4fa5","added_by":"auto","created_at":"2025-04-15 17:24:39","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":51254,"visible":true,"origin":"","legend":"","description":"","filename":"LebertVilarroyaSupplementaryinformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-6253202/v1/3529f273c9e9344910063921.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Theory Precedes Practice: \r\nSimulation-Based Learning Enhances Long-Term Recall, but Prior Text-Based Learning Enhances Its Benefits","fulltext":[{"header":"I. Introduction","content":"\u003cp\u003eSimulation-Based Learning (SBL) is an educational approach that uses realistic, interactive, and immersive simulations to enhance learning and skill acquisition\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. By replicating real-world scenarios, it offers hands-on experiences and serves as an alternative to traditional learning methods. SBL has been widely applied in specialized fields such as STEM (science, technology, engineering, and mathematics) and medical education to develop procedural skills, communication abilities, decision-making, and the application of theoretical knowledge in controlled environments\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThree key aspects of SBL are believed to enhance learning. First, it provides a realistic experience by immersing learners in scenarios featuring real or virtual actors, tools, and contexts that closely resemble those in which the acquired skills or knowledge will be applied\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Second, SBL is typically designed to create a sense of immersion, fully engaging a learner\u0026rsquo;s attention and perception and making them feel deeply involved or absorbed\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Finally, SBL introduces an embodied and interactive dimension that strengthens learning by requiring active engagement with the environment, material objects, and other individuals. The embodied aspect grounds learning in real experiences, while the interactive component dynamically reinforces cognitive processes through action and response rather than passive reception\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eResearch on SBL has primarily been conducted in structured educational settings, such as classrooms, professional training programs, and hospital-based instruction\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. These environments integrate SBL with relevant content, materials, and instructional methods, often incorporating complementary support such as lectures, guided reflections, worked examples, or feedback during or after simulations, facilitating the impact of SBL interventions\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. While these elements likely enhance learning, their specific contributions are not always systematically controlled, making it difficult to isolate the unique role of SBL in the learning process\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e .\u003c/p\u003e \u003cp\u003eMoreover, while some SBL studies employ randomized controlled trials (RCTs), practical constraints often shape sample sizes, comparison conditions, requiring a balance between ecological validity and experimental control. Consequently, comparison groups and control conditions may differ in important aspects, complicating attributions of learning differences to SBL itself. Variability of the control group\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e content exposure\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e, learning duration, instructional setting\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e, assessment analyses and methods across conditions\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e, potential biases from non-blinded assessments (e.g., by experts)\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e, and limited assessments of long-term retention\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e further complicates interpretation. Additionally, the widespread use of between-subjects designs provides insight into overall learning effects but introduces inter-individual variability.\u003c/p\u003e \u003cp\u003eIn sum, while prior research highlights SBL\u0026rsquo;s potential, methodological complexities make it challenging to assess its true impact. Variations in instructional design, engagement levels, and assessment procedures raise questions about whether observed performance differences genuinely reflect the benefits of SBL or stem from extraneous factors. Addressing these limitations through a rigorously controlled study with a robust experimental design in a lab setting is crucial for accurately determining SBL\u0026rsquo;s advantages over traditional learning methods.\u003c/p\u003e \u003cp\u003eThis study examines the effectiveness of a simulation-based learning (SBL) intervention compared to text-based learning in acquiring and retaining novel information. Participants, who were na\u0026iuml;ve to the learning content, learned two distinct and fictitious clinical cases with structured symptom sequences, either through interactive simulation with an actor or through a text-based method. The experimental design systematically controlled for order effects and task biases by counterbalancing conditions and symptom presentations.\u003c/p\u003e \u003cp\u003eTo ensure that the acquired knowledge was entirely novel and free from preexisting biases, we designed the study around a clinical case that was intentionally rendered fictitious. This deliberate separation\u0026mdash;a genuine clinical scenario presented as a fictitious case\u0026mdash;ensured that participants encountered completely new information.\u003c/p\u003e \u003cp\u003eClinical cases, widely used in SBL, provided us with an optimal balance between acquiring specific knowledge and enabling its representation through bodily movements. The structured, sequential symptom presentation mirrors realistic clinical encounters while enhancing the representability of the learned information. In addition, observational learning has been shown to improve retention of relational structures, such as sequential item patterns\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, hence the decision to use an actor simulating the cases. Moreover, the actor not only presented the clinical case but also engaged participants through controlled physical interactions, ensuring an active learning process rather than passive observation. Research shows that touch affects memory recognition\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, gesturing during encoding enhances recall similarly to action\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, and body posture facilitates autobiographical memory retrieval\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. More broadly, embodied effects, such as the \u0026ldquo;enactment effect\u0026rdquo;\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, have shown to enhance retention and recall\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. In sum, this protocol allowed us to integrate realism and embodied interactivity\u0026mdash;key features of SBL\u0026mdash;immersing participants in a lifelike learning experience.\u003c/p\u003e \u003cp\u003eLong-term memory was assessed using item recognition and serial recall tasks, which are widely employed to measure retrieval strength and temporal organization\u003csup\u003e\u003cspan additionalcitationids=\"CR22 CR23\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. These methods have also been used to investigate the effects of simulation, embodiment, and embedded learning\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Learning outcomes were assessed using identical objective measures across both conditions, ensuring that retrieval mode did not influence results. The study also included both video-based and written evaluations to assess whether knowledge acquired through SBL effectively transfers across different retrieval contexts.\u003c/p\u003e \u003cp\u003eBeyond learning and retention, we examine the impact of an imagery questionnaire on accuracy and reaction time across different retrieval modes. Specifically, we explore whether an individual's perceived capacity for mental imagery influences retrieval accuracy, particularly in the written evaluation modality. Prior research has shown that mental imagery can substitute for direct perceptual experience by engaging neural mechanisms similar to actual perception\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. This internal simulation not only allows individuals with stronger imagery abilities to reconstruct learned information more accurately and efficiently in contexts lacking external cues but also strengthens recall for abstract concepts when linked to sensorimotor experiences\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Consequently, internal simulation appears capable of compensating for the absence of direct interaction, thereby reinforcing structured learning.\u003c/p\u003e \u003cp\u003eBased on the proposed role of realism, immersion, embodiment, and interactivity in Simulation-Based Learning (SBL), we hypothesize that participants will exhibit faster reaction times, higher accuracy, and greater confidence in the SBL condition compared to the text-based learning condition. Additionally, due to the within-subjects design and counterbalancing of conditions, and a delayed one-month assessment, we do not anticipate a systematic effect of learning order on performance. Furthermore, given the distinct cognitive processes involved in retrieving information from video versus text, we expect differences in reaction time and accuracy between these modalities. Specifically, we hypothesize that retrieval via video may lead to faster reaction times and higher accuracy compared to written modality retrieval, as video provides a more direct representation of the learned clinical cases, reducing the need for reconstructive processing. Additionally, we predict that individuals with higher imagery capacity may exhibit faster reaction time and accuracy, particularly in the written modality where internal simulation may compensate for the lack of external sensory cues. Finally, we hypothesize that participants will report higher absolute confidence in their responses in the simulation condition compared to text-based learning. We propose that SBL\u0026rsquo;s immersive, interactive, and embodied design enhances sensorimotor engagement and internal simulation, fostering a robust internal feedback mechanism that reinforces both prediction and its verification.\u003c/p\u003e \u003cp\u003eIn sum, by systematically comparing SBL and text-based learning while controlling for prior knowledge and potential confounds, this study contributes to a broader understanding of how embodied and interactive learning experiences shape knowledge acquisition and retention.\u003c/p\u003e"},{"header":"II. Method","content":"\u003cp\u003eThe presented study was not preregistered. This research was approved by the Ethics Committee at the Hospital del Mar Research Institute (Ref : 2022/10237/I). All participants signed a consent form before participating in the study and received monetary compensation for their time.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003e1.\u0026nbsp;\u0026nbsp;Participants\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eParticipants ranged in age from 18 to 35 years, with no history of neurological disorders and no prior experience or studies in the medical, psychiatric, or psychological fields (a complete list of inclusion and exclusion criteria is available in the Supplementary Material 1). They were recruited through convenience sampling by posting flyers on several university campuses.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSample Size and Power Analysis\u003c/strong\u003e: A power analysis was conducted using G*Power to determine the required sample size, ensuring adequate power to detect anticipated effects (Supplementary Material 2). Ultimately, 88 participants (n = 58 females, 28 males, 5 non-binary, mean age \u0026plusmn; SD = 22.98 \u0026plusmn; 4.50) were included in the final analysis after accounting for exclusions. One participant was excluded due to a software issue, and 12 participants were excluded following the actors\u0026apos; video quality check (see the Procedure section for more details).\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003e2.\u0026nbsp;\u0026nbsp;Materials\u003c/strong\u003e\u003c/h2\u003e\n\u003ch3\u003e\u003cstrong\u003eEquipment and Software\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eAll online questionnaires were administered via the Qualtrics platform, and evaluations were conducted using OpenSesame (\u003cem\u003eversion 4.0\u003c/em\u003e\u003csup\u003e27\u003c/sup\u003e) on a Windows computer with a 23-inch Dell monitor (1920x1200 pixels), along with a keyboard and mouse. To facilitate manual responses, two stickers were placed on the keyboard, clearly indicating the designated keys for answering.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003eSBL Room Setup\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eA single room was used for both sessions, ensuring consistency in the learning environment. The chairs for the participant and the actor were positioned one meter apart\u0026mdash;marked on the floor with tape\u0026mdash;aligning with typical social spacing norms\u003csup\u003e28,29\u003c/sup\u003e. A clock was placed on a table visible to both the participant and the actor, helping them manage the timing of each simulation session. Finally, two 360-degree pivot cameras (AOSU - C2E) were installed to record both the actor and the participant. This setup enabled real-time monitoring by the experimenter and allowed for later review if needed, thereby ensuring the simulation sessions were conducted as intended.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSimulation Actors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eProfessional actors were recruited for both the pilot and main studies. These actors regularly collaborate with Hospital del Mar to conduct simulation-based learning (SBL) sessions for medical students. Over the course of several weeks, they participated in monitored and recorded training sessions, alongside regular meetings with the study\u0026rsquo;s supervisors. The primary objectives, before starting the experiment, were to ensure:\u003c/p\u003e\n\u003cp\u003e●\u0026nbsp; \u0026nbsp;\u0026nbsp;Mastery of the clinical cases and associated symptoms,\u003c/p\u003e\n\u003cp\u003e●\u0026nbsp; \u0026nbsp;\u0026nbsp;Consistent intensity and portrayal of symptoms across all actors,\u003c/p\u003e\n\u003cp\u003e●\u0026nbsp; \u0026nbsp;\u0026nbsp;Minimal improvisation\u0026mdash;restricted to prearranged anecdotes about how the symptoms affect the patient\u0026rsquo;s daily life.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStimuli\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo refine the clinical cases, determine the number of sessions, and calibrate the complexity of the evaluation task, a pilot study was conducted prior to the main study (see Supplementary Material 3).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eSymptom Design and Presentation\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e.\u0026nbsp;\u003c/strong\u003eWith the guidance of an external psychiatrist (Dr. Daniel Berg\u0026eacute;), two distinct clinical cases were developed, each featuring a fictitious syndrome characterized by three core symptoms. These symptoms were designed to involve the same bodily part and a similar bodily movement. Each case began with a triggering situation, followed by the sequential presentation of Symptom 1, Symptom 2, and Symptom 3. Additionally, each symptom was paired with a corresponding control symptom. To ensure consistency, the level of difficulty and the written length of the symptoms were standardized across the two clinical cases, maintaining a uniform level of complexity.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEvaluation Materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate symptom recall, all possible pairs were created, including combinations of symptoms, control symptoms, and mixed pairs (symptom + control symptom), resulting in 15 total combinations. Of these, only two sequences were correct (Symptom 1 \u0026rarr; Symptom 2 and Symptom 2 \u0026rarr; Symptom 3). \u0026nbsp; Participants underwent evaluations in two distinct modalities:\u003c/p\u003e\n\u003cp\u003e- \u003cstrong\u003eWritten Modality\u003c/strong\u003e: Each combination displayed the first symptom above a downward arrow, and the second symptom below.\u003c/p\u003e\n\u003cp\u003e- \u003cstrong\u003eVideo Modality\u003c/strong\u003e: All symptoms and control symptoms were recorded with a different actor\u0026mdash;one who did not take part in the SBL sessions\u0026mdash;so that participants would not recognize them. Each video was standardized for elements such as attire, lighting, camera angle, and duration. During the combined video sequence, the first symptom was shown, followed by an arrow, and then the second symptom.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSelf-Report Measures\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePrevious research has shown that social interactions can be influenced by changes in depression\u003csup\u003e30\u003c/sup\u003e, anxiety\u003csup\u003e31\u003c/sup\u003e, and emotional regulation strategies\u003csup\u003e32\u003c/sup\u003e. To account for these variables in our sample, participants completed several self-report measures:\u003c/p\u003e\n\u003cp\u003e● \u003cstrong\u003eWorld Health Organisation\u0026ndash;Five Well-Being Index (WHO-5)\u003c/strong\u003e, assessing general well-being\u003csup\u003e33\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e● \u003cstrong\u003eState Trait Anxiety Scales (STAI)\u003c/strong\u003e, evaluating anxiety\u003csup\u003e34\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e● \u003cstrong\u003eScale of Positive and Negative Experience (SPANE)\u003c/strong\u003e, measuring affect\u003csup\u003e35\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eAdditional questionnaires assessed socio-demographic factors, potential mental health diagnoses, medication use, and current stress levels (0\u0026ndash;10 scale). However, because analyzing how these traits might influence behavior lies beyond our scope, we provide only descriptive statistics and correlations for these measures in Supplementary Material 4. Finally, given that previous research has shown that higher levels of embodied mental imagery can enhance learning outcomes\u003csup\u003e26\u003c/sup\u003e, we included a brief scenario-based measure of imagery, which asked participants to rate their ability to imagine key sensorimotor features (e.g., appearance, movement, voice) of a nonclinical figure and perceived physical contact with the patient in the clinical case. These ratings were combined into a single 10-point index reflecting the overall richness of participants\u0026rsquo; mental simulation.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003e3. \u0026nbsp;Procedure\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003e\u003cstrong\u003eOverall Experimental Design\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study employed a mixed design, incorporating both within-subjects and between-subjects factors. Participants were randomly assigned to one of two groups, ensuring systematic counterbalancing of learning conditions (SBL vs. text-based) and clinical cases. Two fictitious syndromes were developed in consultation with an external psychiatrist, each characterized by three core symptoms affecting the same bodily region and involving similar motor components. Each case followed a structured sequence: an initial triggering situation, followed by the progressive presentation of Symptom 1, Symptom 2, and Symptom 3. To assess symptom recall, participants completed an evaluation phase in two distinct modalities: written and video-based. The written modality presented symptom pairs in a structured format, while the video modality displayed pre-recorded symptom demonstrations by an unfamiliar actor. Each evaluation required participants to judge whether presented symptom sequences were correct, reinforcing the study\u0026rsquo;s focus on both item identification and sequential recall. To control for learning order effects, both symptom sequences and assessment modalities were counterbalanced across participants, ensuring that no systematic bias influenced the results.\u003c/p\u003e\n\u003cp\u003eParticipants were randomly assigned to one of two groups, and the simulation/control conditions as well as the two cases were systematically counterbalanced to minimize order effects and any task-specific biases. Within each modality (written or video), symptom-combination presentations were randomized, and the sequence of modalities themselves was counterbalanced across participants. To indicate whether a displayed combination was correct, participants pressed one of two marked keys (left or right). Moreover, the mapping of these keys (left/right = correct/incorrect) was reversed for half of the participants to avoid response biases. To ensure further balance, half of the participants completed the simulation condition during the first week and the control condition during the second week, while the remaining half did the opposite. Within each subgroup, half of the participants were assigned to case 1 initially and then switched to case 2 for the following week (and condition), whereas the other half followed the reverse order. This design guaranteed that both the conditions and pathologies were evenly distributed among participants.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003eSBL Condition\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eOn the day of the experiment, each participant provided written informed consent upon arrival. \u0026nbsp;They subsequently completed a series of questionnaires, including a sociodemographic questionnaire, the WHO-5, ERQ, STAI-Y (trait), stress assessments, and the imagery representation questionnaire, all administered via the Qualtrics platform.\u003c/p\u003e\n\u003cp\u003eParticipants were given a plastic sheet containing guiding questions (see Supplementary Material 5) designed to help them gather essential clinical information about the patient\u0026apos;s symptoms. They were instructed to ask these questions during the interview, with the flexibility to reformulate them as needed. To ensure greater fluency, participants were given approximately ten minutes before the start of the experiment to familiarize themselves with the full set of questions. Additionally, they were informed that the patient did not display any severe or violent symptoms.\u003c/p\u003e\n\u003cp\u003ePrior to entering the room (where the patient was already seated), the experimenter instructed participants to identify and memorize the patient\u0026rsquo;s symptom sequence through an interview. They were informed the patient might engage in tactile contact and that the sessions would be recorded; however, they were advised not to focus on the camera. Each session had to last between 10 and 12 minutes and around the 12-minute mark, the experimenter would knock on the door to signal the end of the session, while also monitoring via live video to confirm that the actor had finished describing the symptoms. Participants then completed STAI-Y (state) and SPANE questionnaires. Twenty-four hours later, the same procedure was repeated with the same actor to reinforce and deepen the information. After this second session, participants again completed the STAI-Y (state) and SPANE questionnaires, followed by the case evaluation in two counterbalanced formats: video and written.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003eControl Condition\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eParticipants were tasked with identifying and memorizing a sequence of symptoms from a patient described in a text displayed on the screen. The video accompanying the text described one of the two clinical cases and mirrored the structure and information provided during the SBL sessions. The text appeared progressively on the screen while being narrated by an unfamiliar voice, ensuring the viewing duration was standardized at 10 minutes, comparable to the SBL sessions. Participants were instructed to remain focused throughout the presentation, without pausing or interruptions. Following this session, they completed the STAI-Y (anxiety) and SPANE (affect) questionnaires via the Qualtrics platform.\u003c/p\u003e\n\u003cp\u003eTwenty-four hours later, participants took part in a second session that presented more details and information about the same clinical case, emphasizing the symptoms and their impact on the patient\u0026rsquo;s everyday life. After finishing this second session, participants once again filled out the anxiety-state and affect questionnaires and then carried out the evaluation of the pathology in both video and written modalities.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInitial Baseline Assessment and one month Evaluation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo ensure that participants engaged with the material comparably across conditions before progressing to the long-term retention phase, we conducted an initial baseline assessment immediately after the learning phase. This assessment served as a methodological check to confirm that there were no systematic differences between conditions, clinical cases, or counterbalancing factors.\u003c/p\u003e\n\u003cp\u003eDuring the initial baseline assessment, participants were seated at a viewing distance of approximately 60 cm from the screen, in line with recommendations for comfortable vision\u003csup\u003e36\u003c/sup\u003e, with audio levels standardized across all participants. The experimenter provided verbal instructions, informing participants that they would view either written symptoms (written-modality evaluation) or short video clips (video-modality evaluation) on the screen. Participants were instructed to pay close attention to the order in which the symptoms appeared. Following this, participants read detailed on-screen instructions explaining that they would need to determine whether the displayed symptom sequence corresponded to the patient they had encountered during the previous two sessions (SBL or control condition).\u003c/p\u003e\n\u003cp\u003eEach trial began with a fixation cross displayed at the center of the screen for 2 seconds. The symptom sequence was then presented for 10 seconds in the written modality or for 15\u0026ndash;20 seconds in the video modality. After the sequence disappeared, a response screen prompted participants to indicate as quickly and accurately as possible whether the sequence was correct by pressing the designated key. Participants were then asked to rate their confidence in their response on a scale from 1 to 10. A 3-second interval separated each trial. The order of evaluations was counterbalanced, with some participants completing the written-modality evaluation first, followed by the video-modality evaluation, and others completing the evaluations in the reverse order. Measures reflecting performance speed (i.e., response time between the display of the answer screen and the keypress), accuracy (i.e., correct or incorrect keypress) and participants\u0026apos; self-reported confidence (rated using the mouse) were recorded for each trial.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOne-Month Session\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOne month after the last session, participants performed an evaluation in both the video and written formats. The key difference in this session was that participants were required to identify which correct symptom sequences they had observed in any of the previous sessions (SBL and control combined). They were instructed to respond both rapidly and accurately, and to remain focused, as they could not pause the trial or revise their answers. After completing this evaluation, participants filled out the STAI-Y (state) and SPANE questionnaires for the final time.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eQuality Check of Videos of the simulation sessions.\u0026nbsp;\u003c/strong\u003eWhenever there was uncertainty about a given session\u0026rsquo;s adherence to instructions, we conducted a quality check of the video recordings to identify any potential issues related to noncompliance with the experimental protocol. For instance, some exclusions were due to the actor not performing the interaction correctly with the participant, such as not properly engaging in the sensorimotor interaction.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003e4.\u0026nbsp;\u0026nbsp;Data processing and statistical analyses\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eAll inferential tests were performed using RStudio (version \u003cem\u003e4.1.2\u003c/em\u003e\u003csup\u003e37\u003c/sup\u003e). Using Tukey\u0026apos;s method to detect extreme outliers (\u0026plusmn;3 \u0026times; IQR), we replaced outlier trials with the overall mean reaction time calculated without including outliers. In the baseline evaluation session, 2.53% of the trials were replaced, while in the one-month evaluation session, 5.26% were replaced using this method\u003csup\u003e38,39\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eAll significance tests were two-sided, considered significant when p \u0026lt; 0.05. For repeated-measures analyses, if the sphericity assumption was violated, Greenhouse\u0026ndash;Geisser corrections were applied. Effect sizes are reported as partial eta-squared (or eta-squared, as appropriate).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe began by assessing whether demographic variables and counterbalancing factors influenced our primary outcomes \u0026nbsp;in our initial Baseline Assessment. Using the \u003cem\u003eaov\u003c/em\u003e function (\u003cem\u003eafex\u003c/em\u003e library), we examined the effects of gender, actor portraying the patient, lateralization of correct/incorrect responses on the keyboard, and written vs. video evaluation order on reaction time (RT) and accuracy. These analyses revealed no significant effects. We also confirmed that neither the Order of conditions (simulation first vs. control first), the Condition (simulation vs. control), nor the Clinical case (P1 vs. P2) had any observable impact on these measures. Detailed statistical outputs from this baseline assessment are provided in the Supplementary Material 6.\u003c/p\u003e\n\u003cp\u003eSubsequently, our main interest lay in examining performance one month later, focusing on RT, accuracy, and confidence. To address deviations from normality in reaction time, we applied a logarithmic transformation and conducted standard diagnostic checks (e.g., residual normality, linearity, and homogeneity of variance) via the \u003cem\u003eeasystats\u003c/em\u003e package (\u003cem\u003eeasystats\u003c/em\u003e library) in R. We employed linear mixed-effects models using \u003cem\u003elmer\u003c/em\u003e (\u003cem\u003elmerTest\u003c/em\u003e package) and generalized linear mixed-effects models using glmer (\u003cem\u003elme4\u003c/em\u003e package) where appropriate. Model estimates and 95% confidence intervals for the fixed effects were extracted using the \u003cem\u003etidy\u003c/em\u003e function from the \u003cem\u003ebroom.mixed\u003c/em\u003e package. Separately, marginal means\u0026mdash;with their standard errors and 95% confidence intervals computed via Satterthwaite-corrected degrees of freedom\u0026mdash;were derived using the \u003cem\u003eemmeans\u003c/em\u003e package for visualization. We generated the figures with \u003cem\u003eggplot2\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003eFor RT, we fit a LME model with log-transformed RT as the dependent variable and Condition (simulation vs. control) and Modality (written vs. video) as fixed effects. We also included the Order (simulation-control vs control-simulation) and Clinical cases (P1 vs P2) to account for potential confounding factors. Moreover, we incorporated a random intercept to control for subject-specific differences (Formula 1).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eReaction Time ~ Condition + Modality + Order + Clinical case + (1|ID)\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAccuracy\u003c/strong\u003e (0 = incorrect, 1 = correct) was analyzed via a generalized linear mixed-effects model (binomial link). We used Condition and Modality as fixed effects and we also included the Order of conditions and the Clinical case to account for possible confounding factors. Lastly, to account for subject and stimuli specific differences, we incorporated random intercepts into the models (Formula 2)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAccuracy \u0026nbsp;~ Condition + Modality + Order + Clinical case + (1 | ID) + (1 | STIM)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConfidence\u003c/strong\u003e was categorized into absolute (100%) certainty versus anything below 100%. Separating them thus allowed us to capture potential qualitative differences between those who were entirely certain and those who expressed any level of uncertainty, reflecting a distinctly different cognitive state. Given the large cluster of self-reported scores at 100%, this approach also balanced the data for logistic regression. To ensure the reliability of our logistic regression model and in line with the Events Per Variable (EPV) recommendations\u003csup\u003e40,41\u003c/sup\u003e, we confirmed that our model maintained an adequate EPV and that predictor variables were well-balanced, eliminating the need for penalized likelihood corrections (for further details, see the Supplementary Material 7). Hence, we fitted GLME model using confidence as the dependent variable and Condition as well as Modality as fixed effects. To account for confounding factors, we also included the Order of conditions and the Clinical case, and we incorporated random intercepts to control for subject and stimuli specific differences (Formula 3).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConfidence ~ Condition + Modality + Order + Clinical case + (1 | ID) + (1 | STIM)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe employed the \u003cem\u003eDHARMa\u003c/em\u003e package to evaluate model adequacy in both accuracy and confidence analyses. Residual plots, dispersion tests, checks for zero-inflation, and outlier detection indicated no major violations of model assumptions. We further verified our classification approach by examining the ROC curve and computing the AUC for accuracy and confidence, confirming the appropriateness of these binary outcomes. Complete statistical outputs, diagnostic plots, and the corresponding R code are provided in the Supplementary Material.\u003c/p\u003e\n\u003cp\u003eFinally, we examined the effect of the imagery score on reaction time and accuracy during the one-month evaluation. Due to missing responses, this analysis was conducted on a sample of 82 participants (n = 56 females, 22 males, 4 non-binaries; mean age \u0026plusmn; SD = 22.96 \u0026plusmn; 4.52). The imagery score was standardized (mean-centered and scaled) to facilitate interpretation and to ensure comparability across participants.\u003c/p\u003e\n\u003cp\u003eWe first fit a similar LME model using log-transformed RT as the dependent variable and an interaction between Modality and the Imagery score as fixed effects (Formula 4):\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eReaction Time ~ Condition + Modality * Imagery score + Order + Clinical case + (1|ID)\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThen, we fit a similar GLME model with accuracy as the dependent variable and an interaction between Modality et the imagery score as fixed factors (Formula 5)\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAccuracy \u0026nbsp;~ Condition + Modality * Imagery score + Order + Clinical case + (1 | ID) + (1 | STIM)\u003c/strong\u003e\u003c/p\u003e"},{"header":"III. Results","content":"\u003cp\u003e\u003cstrong\u003e\u003cu\u003eEffects of Condition, Order, Modality, and Imagery on Reaction Time\u003c/u\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFollowing the approach described in the \u003cem\u003eStatistical Analysis\u003c/em\u003e section (Formula 1), we fit a linear mixed-effects model to investigate the effects of Condition (simulation vs. control), Modality (written vs. video), Order (simulation-control vs control-simulation), and Clinical case (P1 vs P2) on log-transformed reaction times. The model converged successfully, and diagnostic checks (e.g., residual normality, homogeneity of variance, and collinearity) indicated a satisfactory model fit.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe observed a significant main effect of Condition, with participants in the control condition exhibiting longer reaction times than those in the simulation condition (Estimate = 0.066, 95% CI [0.03, 0.10], SE = 0.017, t = 3.90, p \u0026lt; 0.001). There was also a significant main effect of Order (Estimate = 0.169, 95% CI [0.02, 0.32], SE = 0.074, t = 2.27, p = 0.026), indicating that participants tested in the simulation-control sequence responded more slowly than those in the control-simulation sequence. Neither Modality (Estimate = 0.027, 95% CI [\u0026ndash;0.01, 0.06], SE = 0.017, t = 1.57, p = 0.116) nor Clinical case (Estimate = 0.043, 95% CI [\u0026ndash;0.10, 0.19], SE = 0.074, t = 0.58, p = 0.566) reached statistical significance. The fixed effects explained approximately 1.7% of the variance in log-RT (marginal R\u0026sup2; = 0.017), whereas including random intercepts for participants increased the explained variance to about 24.4% (conditional R\u0026sup2; = 0.244).\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eRelation between Reaction time and Imagery score\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eWe focus here solely on the results concerning the imagery score. For the full model results, please refer to the Supplementary Material 8 (Table S4). We did not observe a significant main effect of the imagery score on reaction times (Estimate = -0.037, 95% CI [-0.08, 0.01], SE = 0.023, t = -1.57, p = 0.121), suggesting that individual differences in imagery ability alone did not influence response times. However, we observed a significant interaction between Modality and the imagery score, indicating that the effect of the modality of presentation on reaction times varied as a function of participants\u0026rsquo; ability to generate mental imagery (Estimate = -0.061, 95% CI [-0.08, -0.04], SE = 0.010, t = -5.88, p \u0026lt; 0.001).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cu\u003eEffects of Condition, Order, Modality, and Imagery on Accuracy\u003c/u\u003e\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eUsing the generalized linear mixed-effects model described in the \u003cem\u003eStatistical Analysis\u003c/em\u003e section (Formula 2), we examined how Condition (simulation vs. control), Modality (written vs. video), Order (simulation-control vs control-simulation), and Clinical case (P1 vs P2) influenced the likelihood of a correct response. The model converged successfully, diagnostics indicated no issues with overdispersion or zero-inflation, and outlier tests did not reveal any problematic observations.\u003c/p\u003e\n\u003cp\u003eWe observed a significant main effect of Condition, indicating that participants in the simulation condition exhibited higher accuracy than those in the control condition \u0026nbsp;(Estimate = -0.456, SE = 0.097, z = -4.71, p \u0026lt; 0.001, corresponding to OR = 0.63, 95% CI [0.52, 0.77]). A main effect of Order also emerged, showing that participants tested in the simulation-control sequence had lower accuracy than those in the control-simulation sequence (Estimate = \u0026ndash;0.837, SE = 0.343, z = \u0026ndash;2.44, p = 0.015; OR = 0.43, 95% CI [0.22, 0.85]). Neither Modality (Estimate = -0.059, SE = 0.096, z = -0.61, p = 0.539, OR = 0.94, 95% CI [0.78, 1.14]) \u0026nbsp;nor Clinical case (Estimate = 0.203, SE = 0.338, z = 0.60, p = 0.548, OR = 1.23, 95% CI [0.63, 2.38]) reached statistical significance. Model fit was further supported by an Area Under the Curve (AUC) of 0.901, reflecting an excellent discriminative ability. Regarding effect sizes, the marginal R\u0026sup2; was approximately 3.2%, while the conditional R\u0026sup2; rose to around 55%.\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eRelation between Accuracy and Imagery score\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eWe focus here solely on the results concerning the imagery score. For the full model results, please refer to the Supplementary Material 8 (Table S5). Participants with higher imagery scores exhibited greater accuracy overall (Estimate = 0.266, SE = 0.114, \u003cem\u003ez\u003c/em\u003e = 2.34, \u003cem\u003ep\u003c/em\u003e = 0.019; \u003cem\u003eOR\u003c/em\u003e = 1.30, \u0026nbsp;95% \u003cem\u003eCI\u003c/em\u003e [1.04, 1.63]). Moreover, a significant interaction between imagery score and modality indicated that the relationship between imagery score and accuracy varied across modalities (\u003cem\u003eEstimate\u003c/em\u003e = -0.186, \u003cem\u003eSE\u003c/em\u003e = 0.058, \u003cem\u003ez\u003c/em\u003e = -3.22, \u003cem\u003ep\u003c/em\u003e = 0.001; \u003cem\u003eOR\u003c/em\u003e = 0.83, 95% \u003cem\u003eCI\u003c/em\u003e [0.74, 0.93]). Specifically, the positive effect of imagery score on accuracy was more pronounced in the written condition than in the video condition.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cu\u003eEffects of Condition, Order and Modality on Confidence\u003c/u\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUsing the generalized linear mixed-effects model described in the \u003cem\u003eStatistical Analysis\u003c/em\u003e section (Formula 3), we examined how Condition (simulation vs. control), Modality (written vs. video), Order (simulation-control vs control-simulation), and Clinical case (P1 vs P2) impacted participants\u0026apos; confidence in their responses. Diagnostic checks showed no indications of singularity, over/under-dispersion, zero-inflation, or problematic outliers. A significant main effect of Condition emerged, indicating that participants in the control condition were substantially less likely to report absolute confidence compared to those in the simulation condition (Estimate = \u0026ndash;1.14, SE = 0.08, z = \u0026ndash;13.68, p \u0026lt; 0.001; OR = 0.32, 95% CI [0.27, 0.38]). Modality showed a marginal effect, with participants in the video condition being slightly more likely to report the highest certainty compared to those in the written condition (Estimate = 0.15, SE = 0.08, z = 1.90, p = 0.058; OR = 1.16, 95% CI [1.00, 1.36]). Neither Order (Estimate = -0.35, SE = 0.57, z = -0.61, p = 0.541; OR = 0.70, 95% CI [0.23, 2.17]) nor Clinical case (Estimate = -0.36, SE = 0.57, z = -0.63, p = 0.527; OR = 0.70, 95% CI [0.23, 2.13]) reached statistical significance. The model showed excellent predictive accuracy, as reflected by an AUC of 0.91, indicating a high ability to discriminate between confidence levels. Regarding effect sizes, the marginal R\u0026sup2; was approximately 3.5%, whereas the conditional R\u0026sup2; rose to about 71.1%.\u003c/p\u003e"},{"header":"IV. Discussion","content":"\u003cp\u003eThis study, designed as a randomized controlled trial using a within-subjects design, examined the impact of a simulation-based protocol on the acquisition and long-term retention of new knowledge by na\u0026iuml;ve learners. A key strength of this study lies in its careful design, which controlled for confounding factors and employed a counterbalanced approach, ensuring robust comparisons between learning conditions. By doing so, we were able to isolate the specific contribution of simulation-based, immersive, and embodied interaction to long-term learning outcomes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eParticipants engaged with an actor who portrayed a clinical case, experiencing the symptoms dynamically through physical interaction, while a matched text-based learning condition provided identical information in a non-interactive, passive format. One month after learning, participants exhibited significantly faster reaction times and higher accuracy when recalling clinical symptom sequences in the simulation-based condition compared to the text-based condition.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eInterestingly, an order effect emerged: when participants began with the control condition, their subsequent performance in the simulation condition was enhanced, whereas starting with the simulation condition did not provide the same benefit for the control condition. This suggests a potential beneficial transfer effect that enhances performance in the simulation condition when it follows the text-based condition, while the control condition undergoes a disadvantageous transfer effect when it follows the simulation condition.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBeyond these effects, we also observed that participants were significantly more likely to report absolute confidence in the SBL condition. Moreover, confidence and accuracy were closely aligned in the SBL condition, with higher confidence accompanying higher accuracy. This contrasts with the control condition, where accuracy was relatively high, but probability of absolute confidence remained moderate, potentially leading to a more tentative application of knowledge.\u003c/p\u003e\n\u003cp\u003eOur findings support previous research demonstrating the benefits of SBL in medical and STEM education\u003csup\u003e3,4\u003c/sup\u003e, although they moderate the advantages of SBL over more traditional text-based methods in learning outcomes. Consistent with prior work, SBL appears to improve knowledge retention, reaction time, and retrieval accuracy, reinforcing the idea that immersive, embodied, and interactive experiences create stronger memory traces than passive learning methods\u003csup\u003e42,43\u003c/sup\u003e. Furthermore, research emphasizing the role of embodiment and interactivity in cognitive processes\u003csup\u003e5\u003c/sup\u003e suggests that learning through active engagement with a simulated patient reinforces memory processes, especially by leveraging sensorimotor integration, a mechanism shown to facilitate encoding and retrieval\u003csup\u003e44,45\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eHowever, our study also provides a more nuanced view of SBL\u0026rsquo;s effectiveness by isolating its specific contributions through a counterbalanced within-subjects design. While SBL demonstrated an advantage, text-based learning can still effectively encode and retrieve information, at least in the domain of knowledge acquisition where information is systematically organized and sequentially presented. The observed differences between conditions might nevertheless have been moderated by the role of mental imagery and the assessment modality. On the one hand, the simulation condition might have been disadvantaged by the retrieval assessment being presented in a video format, attenuating the immersive and embodied benefits of the original learning experience. This aligns with prior research suggesting that the format of retrieval assessments can impact memory performance, particularly when there is a mismatch between learning and retrieval conditions\u003csup\u003e12\u003c/sup\u003e. Conversely, participants with higher imagery capacity may have been able to mentally simulate the clinical case in the written evaluation modality, facilitating their retrieval process. Mental imagery has been shown to engage neural mechanisms akin to actual perception, supporting perceptual learning without direct sensory input\u003csup\u003e25\u003c/sup\u003e. Additionally, it strengthens recall for abstract concepts when linked to sensorimotor experiences\u003csup\u003e26\u003c/sup\u003e. This suggests that internal simulation can compensate for the absence of direct interaction.\u003c/p\u003e\n\u003cp\u003eAn additional key finding of this study is the effect of condition order. For SBL, an initial exposure to the material in a structured, traditional format may facilitate later embodied learning. This finding confirms previous research\u003csup\u003e46\u0026ndash;48\u003c/sup\u003e suggesting that simulation-based learning benefits from prior contextualization or introduction, which may help structure the learning experience before engagement in an immersive and interactive environment. Conversely, for text-based learning, experiencing an immersive, interactive session beforehand may reduce engagement or disrupt retention, possibly because the shift from an active, sensory-rich experience to a more passive format is less conducive to sustained cognitive engagement. This aligns with prior research indicating that immersive, embodied learning can lead to deeper but more rigid knowledge acquisition, potentially hindering subsequent learning through a more abstract and static format such as text-based instruction\u003csup\u003e49,50\u003c/sup\u003e. Immersive learning also places high demands on cognitive resources, engaging sensorimotor, emotional, and social processes. From a cognitive load perspective, this intense engagement early in learning may create a form of anchoring or cognitive rigidity, making it harder to adapt to less interactive and more abstract formats\u003csup\u003e49\u003c/sup\u003e, mirroring the proactive interference observed by Saylık (2021)\u003csup\u003e50\u003c/sup\u003e. Consequently, when working memory processing is hindered by cognitive overload, the consolidation, retrieval, and transfer of information become significantly more challenging, highlighting the advantage of a sequential, low-to-high complexity approach that progressively builds learners\u0026apos; long-term memory stores\u003csup\u003e48\u003c/sup\u003e.\u0026nbsp;While this remains speculative, future research should explore whether such order effects are driven by differences in attentional or cognitive load across modalities, from proactive interference, or shifts in engagement\u003csup\u003e20\u003c/sup\u003e across learning modalities.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe also observed that participants in the simulation condition were three times more likely to report absolute confidence than those in the control condition. One possible explanation may stem from the immersive nature of SBL, which fosters deeper sensorimotor engagement and internal simulation. As suggested by the sensorimotor model of memory, this process may facilitate the reactivation of sensorimotor patterns associated with the learned material. Specifically, the richer sensory and contextual cues provided by SBL, compared to the control condition, may act as markers of a more robust internal feedback mechanism, strengthening the link between knowledge verification and retrieval\u003csup\u003e20\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eIn addition, in our results, confidence and accuracy were closely aligned in the simulation condition, with higher confidence accompanying higher accuracy. This alignment suggests that SBL not only enhances knowledge retention but also reinforces learners\u0026apos; trust in their acquired knowledge. In contrast, the control condition exhibited a mismatch, where accuracy was relatively high, but absolute confidence probability remained moderate. This discrepancy may indicate a more tentative application of knowledge, as learners might hesitate despite correctly recalling information. While confidence is generally beneficial\u003csup\u003e51\u0026ndash;53\u003c/sup\u003e, its relationship with accuracy is complex\u003csup\u003e54\u003c/sup\u003e. When confidence and competence are well-calibrated, learners are more likely to apply their knowledge effectively\u003csup\u003e51\u003c/sup\u003e. However, when confidence outpaces accuracy, overconfidence can lead to errors and misjudgments\u003csup\u003e54,55\u003c/sup\u003e. Therefore, SBL can be seen as a method that may foster both accuracy and self-confidence, creating a learning environment where knowledge is not only acquired but also more efficiently applied than with more traditional methods of learning. This combination of accuracy and confidence may be particularly valuable in settings where both competence and the conviction to act on one\u0026apos;s knowledge are essential. Moreover, by aligning confidence with performance, SBL may help mitigate the risks of overconfidence.\u003c/p\u003e\n\u003cp\u003eFuture studies should aim to replicate these findings and further investigate the mechanisms underlying the observed alignment between high absolute confidence probability and accuracy in SBL. Specifically, research should explore how sensorimotor engagement and internal simulation contribute to learners\u0026apos; self-assessment and whether the enhanced sensory and contextual cues in SBL function as an internal feedback mechanism that strengthens learning outcomes. Additionally, examining the conditions under which confidence and accuracy remain well-calibrated in SBL, as well as identifying potential boundary effects, would provide deeper insights into the robustness and generalizability of these effects.\u003c/p\u003e\n\u003cp\u003eInterestingly, our results demonstrate that despite the advantages of SBL, text-based learning effectively facilitated item identification and sequential recall of realistic, embodied knowledge, enabling learners to transfer this information to recognition tasks involving realistic symptom presentations. Related to this, our findings suggest a possible role of mental imagery in knowledge retrieval when confronted with its realistic depiction. Participants with stronger imagery abilities performed better, suggesting that the ability to mentally simulate content may be an important predictor of retrieval efficiency. This finding extends beyond SBL and may have implications for understanding individual differences in learning across various modalities, such as virtual reality training, medical case studies, or language learning, where visualization plays a key role in comprehension and recall.\u003c/p\u003e\n\u003cp\u003eRegarding limitations, a strength, but also a weakness, of this study is that it specifically targeted the realistic, immersive, embodied, and interactive dimensions of SBL in a lab setting. However, SBL also encompasses other crucial components, such as strategic decision-making, dynamic feedback, and iterative practice\u003csup\u003e56\u003c/sup\u003e. These elements, central to experiential learning theories\u003csup\u003e57\u0026ndash;59\u003c/sup\u003e, contribute significantly to SBL\u0026rsquo;s pedagogical value. While our study did not capture the full spectrum of SBL\u0026rsquo;s educational benefits, isolating its embodied and interactive aspects is an important step toward understanding the mechanisms underlying its effectiveness.\u003c/p\u003e\n\u003cp\u003eAdditionally, although the study demonstrated an advantage of SBL over text-based learning within a one-month retention period, further research is needed to assess the long-term durability of these effects. Future studies should examine whether the observed benefits persist beyond one month and whether different retrieval modalities influence retention over extended periods.\u003c/p\u003e\n\u003cp\u003eAnother limitation concerns the nature of the retrieval assessments. Since knowledge retention was tested using video assessments rather than direct interactions with a live actor, the transferability of learning from SBL to real-world settings may have been attenuated. It is possible that a more immersive retrieval condition, such as role-playing with a live patient, would have amplified the benefits of SBL by preserving its embodied and interactive elements.\u003c/p\u003e\n\u003cp\u003eFinally, the order effect, while intriguing, requires further validation. The fact that simulation benefited from an initial text-based session while text-based learning was negatively impacted by starting with simulation suggests that cognitive flexibility, engagement shifts, or proactive interference may play a role. Future studies should explore whether this effect generalizes to other educational contexts and whether instructional sequencing can be strategically optimized to enhance learning outcomes.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study provides empirical evidence supporting the benefits of simulation-based learning over traditional text-based learning for the acquisition and long-term retention of clinical knowledge. While SBL enhances learning, our results suggest that structured, text-based methods can also yield strong retention outcomes, particularly for item identification and sequential recall.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBy isolating the effects of realism, immersion, embodiment, and interactivity in SBL, our study provides a more precise characterization of the mechanisms that drive its effectiveness. Identifying key factors that enhance learning outcomes enables future learning protocols, whether simulation-based or not, to incorporate principles of interactivity and embodiment, maximizing their effectiveness.\u003c/p\u003e\n\u003cp\u003eThis refined understanding allows for better-informed design of SBL interventions and offers insights that can be transferred to non-SBL learning environments. Future research building on these findings will help create more tailored, evidence-based approaches to enhance learning through immersive and interactive experiences.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data are available at the following OSF repository: https://osf.io/ub6ce/?view_only=6ea1d137e1bf41079bdc9b20cb8df7fe \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData preprocessing and analysis scripts are available at the following OSF repository: https://osf.io/ub6ce/?view_only=6ea1d137e1bf41079bdc9b20cb8df7fe \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank Laura Alsina, Mila Markovic, and Anna Yue for their assistance during the experimental sessions, as well as the actors Mar Casas, Berta Graells, Mirena Nafarrete, Arnau Armengol and Carla Vilar\u0026oacute; for their dedication. Special thanks to Dr. Meritxell Girvent and Dr. Carolina P\u0026eacute;rez Garc\u0026iacute;a for their assistance in designing the simulation protocol, and to Dr. Daniel Berg\u0026eacute; for his contribution to developing the fictitious clinical syndromes. This work was supported by Fundaci\u0026oacute;n Tatiana. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThese authors jointly supervised this work: Ang\u0026eacute;lique Lebert and \u0026Oacute;scar Vilarroya\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors and Affiliations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUnitat de Recerca en Neuroci\u0026egrave;ncia Cognitiva Departament de Psiquiatria i Medicina Legal Universitat Aut\u0026ograve;noma de Barcelona Spain and Hospital del Mar Research Institute, Barcelona Spain\u003c/p\u003e\n\u003cp\u003eAng\u0026eacute;lique Lebert and \u0026Oacute;scar Vilarroya\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eContributions (to complete/adapt)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAL: Conceptualization, Methodology, Software, Formal Analysis, Investigation, Writing. OV: Conceptualization, Methodology, Supervision, Project administration, Resources, Funding acquisition, Writing.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorresponding authors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCorrespondence to \u0026Oacute;scar Vilarroya \u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eFrasson, C. \u0026amp; Blanchard, E. G. Simulation-Based Learning. in \u003cem\u003eEncyclopedia of the Sciences of Learning\u003c/em\u003e (ed. Seel, N. M.) 3076\u0026ndash;3080 (Springer US, Boston, MA, 2012). doi:10.1007/978-1-4419-1428-6_129.\u003c/li\u003e\n \u003cli\u003eGormley, G. J. \u0026amp; Murphy, P. When I say \u0026hellip; simulation. \u003cem\u003eMedical Education\u003c/em\u003e \u003cstrong\u003e57\u003c/strong\u003e, 1182\u0026ndash;1183 (2023).\u003c/li\u003e\n \u003cli\u003eChernikova, O. \u003cem\u003eet al.\u003c/em\u003e Simulation-Based Learning in Higher Education: A Meta-Analysis. \u003cem\u003eReview of Educational Research\u003c/em\u003e \u003cstrong\u003e90\u003c/strong\u003e, 499\u0026ndash;541 (2020).\u003c/li\u003e\n \u003cli\u003eLateef, F. Simulation-based learning: Just like the real thing. \u003cem\u003eJ Emerg Trauma Shock\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 348\u0026ndash;352 (2010).\u003c/li\u003e\n \u003cli\u003eLebert, A. \u0026amp; Vilarroya, \u0026Oacute;. The links between experiential learning and 4E cognition. \u003cem\u003eAnnals of the New York Academy of Sciences\u003c/em\u003e \u003cstrong\u003e1541\u003c/strong\u003e, 37\u0026ndash;52 (2024).\u003c/li\u003e\n \u003cli\u003eLee, C. A., Pais, K., Kelling, S. \u0026amp; Anderson, O. S. A scoping review to understand simulation used in interprofessional education. \u003cem\u003eJournal of Interprofessional Education \u0026amp; Practice\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 15\u0026ndash;23 (2018).\u003c/li\u003e\n \u003cli\u003eKaplonyi, J. \u003cem\u003eet al.\u003c/em\u003e Understanding the impact of simulated patients on health care learners\u0026rsquo; communication skills: a systematic review. \u003cem\u003eMedical Education\u003c/em\u003e \u003cstrong\u003e51\u003c/strong\u003e, 1209\u0026ndash;1219 (2017).\u003c/li\u003e\n \u003cli\u003eAlharbi, A., Nurfianti, A., Mullen, R. F., McClure, J. D. \u0026amp; Miller, W. H. The effectiveness of simulation-based learning (SBL) on students\u0026rsquo; knowledge and skills in nursing programs: a systematic review. \u003cem\u003eBMC Medical Education\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 1099 (2024).\u003c/li\u003e\n \u003cli\u003eClark, A. D. \u003cem\u003eet al.\u003c/em\u003e The Effect of 3-Dimensional Simulation on Neurosurgical Skill Acquisition and Surgical Performance: A Review of the Literature. \u003cem\u003eJournal of Surgical Education\u003c/em\u003e \u003cstrong\u003e74\u003c/strong\u003e, 828\u0026ndash;836 (2017).\u003c/li\u003e\n \u003cli\u003evon Wendt, C. E. A. \u0026amp; Niemi-Murola, L. Simulation in Interprofessional Clinical Education: Exploring Validated Nontechnical Skills Measurement Tools. \u003cem\u003eSimulation in Healthcare\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 131 (2018).\u003c/li\u003e\n \u003cli\u003eMcGaghie, W. C., Issenberg, S. B., Cohen, E. R., Barsuk, J. H. \u0026amp; Wayne, D. B. Does Simulation-Based Medical Education With Deliberate Practice Yield Better Results Than Traditional Clinical Education? A Meta-Analytic Comparative Review of the Evidence. \u003cem\u003eAcademic Medicine\u003c/em\u003e \u003cstrong\u003e86\u003c/strong\u003e, 706 (2011).\u003c/li\u003e\n \u003cli\u003eHofmann, R., Curran, S. \u0026amp; Dickens, S. Models and measures of learning outcomes for non-technical skills in simulation-based medical education: Findings from an integrated scoping review of research and content analysis of curricular learning objectives. \u003cem\u003eStudies in Educational Evaluation\u003c/em\u003e \u003cstrong\u003e71\u003c/strong\u003e, 101093 (2021).\u003c/li\u003e\n \u003cli\u003eManzi, A. \u0026amp; Nigro, G. Long-term memory for performed and observed actions: Retrieval awareness and source monitoring. \u003cem\u003eMemory\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 595\u0026ndash;603 (2008).\u003c/li\u003e\n \u003cli\u003eSchult, J., St\u0026uuml;lpnagel, R. von \u0026amp; Steffens, M. C. Enactment versus Observation: Item-Specific and Relational Processing in Goal-Directed Action Sequences (and Lists of Single Actions). \u003cem\u003ePLOS ONE\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, e99985 (2014).\u003c/li\u003e\n \u003cli\u003eConvertino, G. \u003cem\u003eet al.\u003c/em\u003e Positive and negative touch differentially modulate metacognitive memory judgements for emotional stimuli. \u003cem\u003eBritish Journal of Psychology\u003c/em\u003e \u003cstrong\u003e116\u003c/strong\u003e, 34\u0026ndash;51 (2025).\u003c/li\u003e\n \u003cli\u003eDella Longa, L., Gliga, T. \u0026amp; Farroni, T. Tune to touch: Affective touch enhances learning of face identity in 4-month-old infants. \u003cem\u003eDevelopmental Cognitive Neuroscience\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 42\u0026ndash;46 (2019).\u003c/li\u003e\n \u003cli\u003eCook, S. W., Mitchell, Z. \u0026amp; Goldin-Meadow, S. Gesturing makes learning last. \u003cem\u003eCognition\u003c/em\u003e \u003cstrong\u003e106\u003c/strong\u003e, 1047\u0026ndash;1058 (2008).\u003c/li\u003e\n \u003cli\u003eDijkstra, K., Kaschak, M. P. \u0026amp; Zwaan, R. A. Body posture facilitates retrieval of autobiographical memories. \u003cem\u003eCognition\u003c/em\u003e \u003cstrong\u003e102\u003c/strong\u003e, 139\u0026ndash;149 (2007).\u003c/li\u003e\n \u003cli\u003eRoberts, B. R. T., MacLeod, C. M. \u0026amp; Fernandes, M. A. The enactment effect: A systematic review and meta-analysis of behavioral, neuroimaging, and patient studies. \u003cem\u003ePsychol Bull\u003c/em\u003e \u003cstrong\u003e148\u003c/strong\u003e, 397\u0026ndash;434 (2022).\u003c/li\u003e\n \u003cli\u003eIan\u0026igrave;, F. Embodied memories: Reviewing the role of the body in memory processes. \u003cem\u003ePsychon Bull Rev\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 1747\u0026ndash;1766 (2019).\u003c/li\u003e\n \u003cli\u003eHumphreys, M. S., Hockley, W. E. \u0026amp; Chalmers, K. A. Recognition memory: The probe, the returned signal, and the decision. \u003cem\u003ePsychon Bull Rev\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 568\u0026ndash;598 (2024).\u003c/li\u003e\n \u003cli\u003eHurlstone, M. J. Serial Recall. in \u003cem\u003eThe Oxford Handbook of Human Memory, Two Volume Pack: Foundations and Applications\u003c/em\u003e (eds. Kahana, M. J. \u0026amp; Wagner, A. D.) 0 (Oxford University Press, 2024). doi:10.1093/oxfordhb/9780190917982.013.28.\u003c/li\u003e\n \u003cli\u003eLogan, G. D. \u0026amp; Cox, G. E. Serial order depends on item-dependent and item-independent contexts. \u003cem\u003ePsychological Review\u003c/em\u003e \u003cstrong\u003e130\u003c/strong\u003e, 1672\u0026ndash;1687 (2023).\u003c/li\u003e\n \u003cli\u003eMack, M. L. \u0026amp; Palmeri, T. J. Discrimination, Recognition, and Classification. in \u003cem\u003eThe Oxford Handbook of Human Memory, Two Volume Pack: Foundations and Applications\u003c/em\u003e (eds. Kahana, M. J. \u0026amp; Wagner, A. D.) 0 (Oxford University Press, 2024). doi:10.1093/oxfordhb/9780190917982.013.30.\u003c/li\u003e\n \u003cli\u003eTartaglia, E. M., Bamert, L., Herzog, M. H. \u0026amp; Mast, F. W. Perceptual learning of motion discrimination by mental imagery. \u003cem\u003eJournal of Vision\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 14 (2012).\u003c/li\u003e\n \u003cli\u003eMarre, Q., Huet, N. \u0026amp; Labeye, E. Embodied mental imagery improves memory. \u003cem\u003eQuarterly Journal of Experimental Psychology\u003c/em\u003e \u003cstrong\u003e74\u003c/strong\u003e, 1396\u0026ndash;1405 (2021).\u003c/li\u003e\n \u003cli\u003eMath\u0026ocirc;t, S., Schreij, D. \u0026amp; Theeuwes, J. OpenSesame: An open-source, graphical experiment builder for the social sciences. \u003cem\u003eBehav Res\u003c/em\u003e \u003cstrong\u003e44\u003c/strong\u003e, 314\u0026ndash;324 (2012).\u003c/li\u003e\n \u003cli\u003eHecht, H., Welsch, R., Viehoff, J. \u0026amp; Longo, M. R. The shape of personal space. \u003cem\u003eActa Psychologica\u003c/em\u003e \u003cstrong\u003e193\u003c/strong\u003e, 113\u0026ndash;122 (2019).\u003c/li\u003e\n \u003cli\u003eLebert, A., Vergilino-Perez, D. \u0026amp; Chaby, L. Keeping distance or getting closer: How others\u0026rsquo; emotions shape approach-avoidance postural behaviors and preferred interpersonal distance. \u003cem\u003ePLoS One\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, e0298069 (2024).\u003c/li\u003e\n \u003cli\u003eGr\u0026egrave;zes, J., Risch, N., Courtet, P., Oli\u0026eacute;, E. \u0026amp; Mennella, R. Depression and approach-avoidance decisions to emotional displays: The role of anhedonia. \u003cem\u003eBehav Res Ther\u003c/em\u003e \u003cstrong\u003e164\u003c/strong\u003e, 104306 (2023).\u003c/li\u003e\n \u003cli\u003eKnyazev, G. G., Bocharov, A. V., Slobodskaya, H. R. \u0026amp; Ryabichenko, T. I. Personality-linked biases in perception of emotional facial expressions. \u003cem\u003ePersonality and Individual Differences\u003c/em\u003e \u003cstrong\u003e44\u003c/strong\u003e, 1093\u0026ndash;1104 (2008).\u003c/li\u003e\n \u003cli\u003eButler, E. A. \u003cem\u003eet al.\u003c/em\u003e The social consequences of expressive suppression. \u003cem\u003eEmotion\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 48\u0026ndash;67 (2003).\u003c/li\u003e\n \u003cli\u003eTopp, C. W., \u0026Oslash;stergaard, S. D., S\u0026oslash;ndergaard, S. \u0026amp; Bech, P. The WHO-5 Well-Being Index: A Systematic Review of the Literature. \u003cem\u003ePsychother Psychosom\u003c/em\u003e \u003cstrong\u003e84\u003c/strong\u003e, 167\u0026ndash;176 (2015).\u003c/li\u003e\n \u003cli\u003eBuela-Casa, G., Guill\u0026eacute;n-Riquelme, A. \u0026amp; Seisdedos Cubero, N. Cuestionario de Ansiedad Estado-Rasgo: Adaptaci\u0026oacute;n espanola. \u003cem\u003eMadrid: TEA Ediciones\u003c/em\u003e (2011).\u003c/li\u003e\n \u003cli\u003eEspejo, B., Checa, I., Perales-Puchalt, J. \u0026amp; Lis\u0026oacute;n, J. F. Validation and Measurement Invariance of the Scale of Positive and Negative Experience (SPANE) in a Spanish General Sample. \u003cem\u003eInternational Journal of Environmental Research and Public Health\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 8359 (2020).\u003c/li\u003e\n \u003cli\u003eRempel, D., Willms, K., Anshel, J., Jaschinski, W. \u0026amp; Sheedy, J. The Effects of Visual Display Distance on Eye Accommodation, Head Posture, and Vision and Neck Symptoms. \u003cem\u003eHum Factors\u003c/em\u003e \u003cstrong\u003e49\u003c/strong\u003e, 830\u0026ndash;838 (2007).\u003c/li\u003e\n \u003cli\u003eTeam, Rs. RStudio: integrated development for R. RStudio, PBC. (2020).\u003c/li\u003e\n \u003cli\u003eTukey, J. W. Exploratory data analysis. \u003cem\u003eReading/Addison-Wesley\u003c/em\u003e (1977).\u003c/li\u003e\n \u003cli\u003eZellner, A., Richardson, A. M., Lidbury, B. A., Hobson, P. \u0026amp; Badrick, T. An Investigation into Outlier Elimination and Calculation Methods in the Determination of Reference Intervals using Serum Immunoglobulin A as a Model Data Collection. Preprint at https://doi.org/10.48550/arXiv.1907.09637 (2019).\u003c/li\u003e\n \u003cli\u003ePeduzzi, P., Concato, J., Kemper, E., Holford, T. R. \u0026amp; Feinstein, A. R. A simulation study of the number of events per variable in logistic regression analysis. \u003cem\u003eJournal of Clinical Epidemiology\u003c/em\u003e \u003cstrong\u003e49\u003c/strong\u003e, 1373\u0026ndash;1379 (1996).\u003c/li\u003e\n \u003cli\u003eVan Smeden, M. \u003cem\u003eet al.\u003c/em\u003e Sample size for binary logistic prediction models: Beyond events per variable criteria. \u003cem\u003eStat Methods Med Res\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 2455\u0026ndash;2474 (2019).\u003c/li\u003e\n \u003cli\u003eBrodt, S. \u003cem\u003eet al.\u003c/em\u003e Rapid and independent memory formation in the parietal cortex. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e \u003cstrong\u003e113\u003c/strong\u003e, 13251\u0026ndash;13256 (2016).\u003c/li\u003e\n \u003cli\u003eKumar, N., Manning, T. F. \u0026amp; Ostry, D. J. Somatosensory cortex participates in the consolidation of human motor memory. \u003cem\u003ePLoS Biol\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, e3000469 (2019).\u003c/li\u003e\n \u003cli\u003eKruger, J. S., Kruger, D. J. \u0026amp; Suzuki, R. Assessing the Effectiveness of Experiential Learning in a Student-Run Free Clinic. \u003cem\u003ePedagogy in Health Promotion\u003c/em\u003e \u003cstrong\u003e1\u003c/strong\u003e, 91\u0026ndash;94 (2015).\u003c/li\u003e\n \u003cli\u003eWaddell, A., Lennox, A., Spassova, G. \u0026amp; Bragge, P. Barriers and facilitators to shared decision-making in hospitals from policy to practice: a systematic review. \u003cem\u003eImplementation Science\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 74 (2021).\u003c/li\u003e\n \u003cli\u003eEppich, W. \u0026amp; Cheng, A. Promoting Excellence and Reflective Learning in Simulation (PEARLS): Development and Rationale for a Blended Approach to Health Care Simulation Debriefing. \u003cem\u003eSimulation in Healthcare\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 106 (2015).\u003c/li\u003e\n \u003cli\u003eRudolph, J. W., Raemer, D. B. \u0026amp; Simon, R. Establishing a Safe Container for Learning in Simulation: The Role of the Presimulation Briefing. \u003cem\u003eSimulation in Healthcare\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 339 (2014).\u003c/li\u003e\n \u003cli\u003eFraser, K. L., Ayres, P. \u0026amp; Sweller, J. Cognitive load theory for the design of medical simulations. \u003cem\u003eSimulation in Healthcare\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 295\u0026ndash;307 (2015).\u003c/li\u003e\n \u003cli\u003eHuang, C. L., Luo, Y. F., Yang, S. C., Lu, C. M. \u0026amp; Chen, A.-S. Influence of Students\u0026rsquo; Learning Style, Sense of Presence, and Cognitive Load on Learning Outcomes in an Immersive Virtual Reality Learning Environment. \u003cem\u003eJournal of Educational Computing Research\u003c/em\u003e \u003cstrong\u003e58\u003c/strong\u003e, 596\u0026ndash;615 (2020).\u003c/li\u003e\n \u003cli\u003eSaylik, R., Castiello, S. \u0026amp; Murphy, R. A. The role of emotional interference on learning in an emotional probabilistic Go/No-Go task. \u003cem\u003eDusunen Adam: Journal of Psychiatry \u0026amp; Neurological Sciences\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, (2021).\u003c/li\u003e\n \u003cli\u003eColquitt, J. A., LePine, J. A. \u0026amp; Noe, R. A. Toward an integrative theory of training motivation: A meta-analytic path analysis of 20 years of research. \u003cem\u003eJournal of Applied Psychology\u003c/em\u003e \u003cstrong\u003e85\u003c/strong\u003e, 678\u0026ndash;707 (2000).\u003c/li\u003e\n \u003cli\u003eJudge, T. A. \u0026amp; Bono, J. E. Relationship of core self-evaluations traits\u0026mdash;self-esteem, generalized self-efficacy, locus of control, and emotional stability\u0026mdash;with job satisfaction and job performance: A meta-analysis. \u003cem\u003eJournal of Applied Psychology\u003c/em\u003e \u003cstrong\u003e86\u003c/strong\u003e, 80\u0026ndash;92 (2001).\u003c/li\u003e\n \u003cli\u003eRichardson, M., Abraham, C. \u0026amp; Bond, R. Psychological correlates of university students\u0026rsquo; academic performance: A systematic review and meta-analysis. \u003cem\u003ePsychological Bulletin\u003c/em\u003e \u003cstrong\u003e138\u003c/strong\u003e, 353\u0026ndash;387 (2012).\u003c/li\u003e\n \u003cli\u003eKoriat, A. The self-consistency model of subjective confidence. \u003cem\u003ePsychological Review\u003c/em\u003e \u003cstrong\u003e119\u003c/strong\u003e, 80\u0026ndash;113 (2012).\u003c/li\u003e\n \u003cli\u003eDunning, D. Chapter five - The Dunning\u0026ndash;Kruger Effect: On Being Ignorant of One\u0026rsquo;s Own Ignorance. in \u003cem\u003eAdvances in Experimental Social Psychology\u003c/em\u003e (eds. Olson, J. M. \u0026amp; Zanna, M. P.) vol. 44 247\u0026ndash;296 (Academic Press, 2011).\u003c/li\u003e\n \u003cli\u003ePazin Filho, A. \u0026amp; Romano, M. M. D. SIMULA\u0026Ccedil;\u0026Atilde;O: ASPECTOS CONCEITUAIS. in \u003cem\u003eMedicina (Ribeir\u0026atilde;o Preto)\u003c/em\u003e vol. 40 167\u0026ndash;170 (2007).\u003c/li\u003e\n \u003cli\u003eFenwick, T. J. Expanding Conceptions of Experiential Learning: A Review of the Five Contemporary Perspectives on Cognition. \u003cem\u003eAdult Education Quarterly\u003c/em\u003e \u003cstrong\u003e50\u003c/strong\u003e, 243\u0026ndash;272 (2000).\u003c/li\u003e\n \u003cli\u003eItin, C. M. Reasserting the Philosophy of Experiential Education as a Vehicle for Change in the 21st Century. \u003cem\u003eJournal of Experiential Education\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 91\u0026ndash;98 (1999).\u003c/li\u003e\n \u003cli\u003eKolb, D. A. \u003cem\u003eExperiential Learning: Experience as the Source of Learning and Development\u003c/em\u003e. (FT Press, 2014).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"npj-science-of-learning","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"npjscilearn","sideBox":"Learn more about [npj Science of Learning](http://www.nature.com/npjscilearn/)","snPcode":"41539","submissionUrl":"https://mts-npjscilearn.nature.com/cgi-bin/main.plex","title":"npj Science of Learning","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Simulation-Based Learning, simulation training, mental imagery, Instructional sequencing, text-based learning, immersive learning, interactive learning","lastPublishedDoi":"10.21203/rs.3.rs-6253202/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6253202/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSimulation-Based Learning (SBL) is widely used in medical and STEM education, offering immersive, embodied, and interactive experiences. However, its implementation often introduces variability in control conditions, instructional design, and a reliance on between-subjects comparisons, making it difficult to isolate its specific contributions to learning.\u003c/p\u003e\n\u003cp\u003eThis study used a within-subjects randomized controlled design (N=88) to evaluate the effects of SBL on knowledge retention, retrieval efficiency, and confidence calibration. Participants, naïve to the learning content, learned two counterbalanced fictitious clinical cases via either a live-actor simulation or a structured text-based format. Retention was assessed one month later through video-based and written evaluations measuring accuracy, reaction time, and confidence.\u003c/p\u003e\n\u003cp\u003eSBL led to significantly faster reaction times (Estimate = 0.066, 95% CI [0.03, 0.10], SE = 0.017, t = 3.90, p \u0026lt; 0.001) and higher recall accuracy (Estimate = -0.456, SE = 0.097, z = -4.71, p \u0026lt; 0.001, OR = 0.63, 95% CI [0.52, 0.77]) compared to text-based learning. An order effect emerged: learning first via text enhanced subsequent SBL performance, whereas the reverse sequence impaired text-based retention (Estimate = –0.837, SE = 0.343, z = –2.44, p = 0.015; OR = 0.43, 95% CI [0.22, 0.85]).\u003c/p\u003e\n\u003cp\u003eMental imagery ability influenced retrieval accuracy, with higher imagery scores predicting greater accuracy overall (Estimate = 0.266, SE = 0.114, z = 2.34, p = 0.019; OR = 1.30, 95% CI [1.04, 1.63]). A significant interaction between imagery ability and modality showed that this effect was more pronounced in the text-based condition (Estimate = -0.186, SE = 0.058, z = -3.22, p = 0.001; OR = 0.83, 95% CI [0.74, 0.93]).\u003c/p\u003e\n\u003cp\u003eConfidence ratings further highlighted SBL’s advantages, with participants in the SBL condition being three times more likely to report absolute confidence (Estimate = –1.14, SE = 0.08, z = –13.68, p \u0026lt; 0.001; OR = 0.32, 95% CI [0.27, 0.38]). Moreover, in the SBL condition, confidence was more closely aligned with actual accuracy.\u003c/p\u003e\n\u003cp\u003eThis study provides empirical evidence supporting the benefits of SBL over traditional text-based learning for the acquisition and long-term retention of clinical knowledge. While SBL enhances learning, our results suggest that structured, text-based methods can also yield strong retention outcomes, particularly for item identification and sequential recall.\u003c/p\u003e\n\u003cp\u003eThese findings clarify the role of SBL’s immersive, embodied, and interactive elements in shaping learning while highlighting the impact of instructional sequencing and individual differences in imagery ability. Additionally, they underscore the potential benefit of SBL in aligning self-confidence with accuracy.\u003c/p\u003e\n\u003cp\u003eBy isolating specific SBL features, this study refines our understanding of its effects on knowledge acquisition, retrieval, and self-confidence alignment. This refined understanding allows for better-informed design of SBL interventions and offers insights that can be applied to non-SBL learning environments.\u003c/p\u003e","manuscriptTitle":"Theory Precedes Practice: \nSimulation-Based Learning Enhances Long-Term Recall, but Prior Text-Based Learning Enhances Its Benefits","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-15 17:24:34","doi":"10.21203/rs.3.rs-6253202/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-09-23T03:02:29+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-22T13:41:46+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"189152782735579399908591378814846302253","date":"2025-08-11T08:05:48+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"17692576162159711464286752820930653203","date":"2025-07-24T19:36:51+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-29T10:56:55+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-25T13:27:34+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"145800184009830953087699152855751232379","date":"2025-04-24T10:17:48+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-22T06:08:15+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"328612280645932687693178573788876697876","date":"2025-03-28T22:14:04+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"228051265097021976277777221754859420820","date":"2025-03-26T09:05:03+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-03-26T09:00:17+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-03-26T05:24:21+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-03-19T10:52:42+00:00","index":"","fulltext":""},{"type":"submitted","content":"npj Science of Learning","date":"2025-03-18T12:26:03+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"npj-science-of-learning","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"npjscilearn","sideBox":"Learn more about [npj Science of Learning](http://www.nature.com/npjscilearn/)","snPcode":"41539","submissionUrl":"https://mts-npjscilearn.nature.com/cgi-bin/main.plex","title":"npj Science of Learning","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"b22688c1-e490-4e53-8968-d44aa9575cc3","owner":[],"postedDate":"April 15th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":46721095,"name":"Biological sciences/Psychology"},{"id":46721096,"name":"Biological sciences/Psychology/Human behaviour"}],"tags":[],"updatedAt":"2025-12-01T16:07:08+00:00","versionOfRecord":{"articleIdentity":"rs-6253202","link":"https://doi.org/10.1038/s41539-025-00380-9","journal":{"identity":"npj-science-of-learning","isVorOnly":false,"title":"npj Science of Learning"},"publishedOn":"2025-11-24 15:57:13","publishedOnDateReadable":"November 24th, 2025"},"versionCreatedAt":"2025-04-15 17:24:34","video":"","vorDoi":"10.1038/s41539-025-00380-9","vorDoiUrl":"https://doi.org/10.1038/s41539-025-00380-9","workflowStages":[]},"version":"v1","identity":"rs-6253202","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6253202","identity":"rs-6253202","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}