Introduction

The inclusion of visually engaging but instructionally irrelevant elements—termed seductive details (Harp & Mayer, 1998)—has sparked ongoing debate regarding their pedagogical value. While such features may enhance engagement by capturing attention and stimulating situational interest, they risk diverting cognitive resources from core content, potentially impairing learning. This trade-off poses a significant challenge in designing digital instructional materials, where balancing engagement and cognitive load is essential. In response, the present meta-analysis synthesizes two decades of research on the seductive details effect, aiming to clarify under what conditions these elements hinder or support learning and to inform evidence-based design practices for effective and engaging instruction.

Literature review

Seductive details effect and mediating role of cognitive load

A central question in seductive details research is whether such elements help or hinder learning. Early studies typically reported negative effects, particularly for learners with lower working memory who struggle to filter irrelevant content (Harp & Mayer, 1997). More recent findings, however, are mixed. Some suggest seductive details enhance situational interest without improving deep learning or metacognition (Magner et al., 2014; Jager & Wiley, 2014), while others propose they act as cognitive breaks that help sustain attention and reduce fatigue (Fries et al., 2019; Szpunar et al., 2013). These inconsistencies underscore the need to explore boundary conditions such as learner characteristics, task demands, and design features. The seductive details effect is grounded in Cognitive Load Theory (CLT; Sweller et al., 1998) and the Cognitive Theory of Multimedia Learning (CTML; Mayer, 2014), which differentiate intrinsic, extraneous, and germane load. Seductive details are thought to increase extraneous load and impaired recall performance (Wesenberg et al., 2025), though the mediating effect of extraneous cognitive load is not consistently observed (Colliot & Boucheix, 2024).

Moderators variables

Recent research has moved beyond confirming the seductive details effect to investigating its boundary conditions. This reflects a broader recognition that instructional outcomes depend on learner characteristics and environmental context. Moderating factors include age (Scherer et al., 2023), prior knowledge (Magner et al., 2014), working memory capacity (Sanchez & Wiley, 2006), and motivation (Wang & Adesope, 2016). Contextual elements like instructional techniques (Eitel et al., 2019; Wang et al., 2017) and learning environments (Bender et al., 2021; Fries et al., 2019) also influence the effect. Emerging evidence further suggests that interactions between learner traits and seductive detail types shape outcomes (Wesenberg et al., 2024). Given these inconsistencies, a meta-analytic approach is needed to clarify when and for whom seductive details help or hinder learning (Mayer, 2019).

Current meta-analysis

Two prior meta-analyses examined the effects of seductive details on academic achievement. Rey (2012) reported small-to-medium negative effects on recall (d = -0.30) and transfer (d = -0.48), while Sundararajan and Adesope (2020) analyzed 68 effect sizes from 42 studies and found a similar effect (g = -0.33). However, these studies did not address effect size non-independence or test cognitive load as a mediator. Since 2020, the number of relevant studies has nearly doubled, highlighting the need for an updated analysis. This meta-analysis includes studies published between 2019 and 2025 and adopts a multilevel approach to account for dependencies among effect sizes. The study is guided by four research questions:

RQ1: What is the overall effect of seductive details on academic achievement?

RQ2: How do they affect specific learning outcomes?

RQ3: What factors moderate these effects?

RQ4: To what extent does cognitive load mediate the relationship between seductive details and learning outcomes?

Selection criteria The selected literatures should met the inclusion criteria which is reported in Table 1. Literature search and selection. As shown in Figure 1, the literature search followed the PRISMA guidelines (Page et al., 2021; Moher et al., 2009). We conducted a literature search in the electronic databases including Web of Science, PsycINFO, Scopus, ProQuest, and ERIC (Education Resources Information Center) using the search syntax shown in Table 2. Consequently, a total of 51 publications were included in this review at the end. Literature coding Before coding, the authors held detailed discussions to finalize the coding criteria. A random sample of 10 references (20%) was independently coded to ensure consistency, with disagreements resolved through discussion. The remaining 41 references were coded using the agreed scheme. Inter-coder reliability was high (Cohen’s kappa = 0.91–0.96, p < 0.001). Following Wilson and Lipsey (2001), six coding schemes were applied (see Table 3), focusing on effect sizes and study-level variables. Effect sizes were categorized by learning outcome (recall, comprehension, transfer), and 14 study-level variables were coded to support moderation and mediation analyses. Data analysis All of the analyses, including overall effect size calculation, sensitivity analysis, moderation analysis, and mediation analysis, were conducted in R Studio (R Core Team, 2017) with the metafor package (Viechtbauer, 2010) and metaSEM package (Cheung, 2015; Cheung & Cheung, 2016). In this study, the effect size represents the effects of seductive details on the academic achievements. We used Hedges’ g (Hedges, 1982) as the standardized mean difference effect size metric. For the mediating analysis, we employed a two-stage meta-analytic structural equation modeling (MASEM) approach (Cheung & Chan, 2005), using 16 effect sizes from eight studies—including four derived from raw data obtained by contacting the authors of two studies (Wang et al., 2019; 2021).

The seductive details effect (RQ1 & RQ2) The present meta-analysis confirms that seductive details have a small but significant negative effect on academic achievement (g = -0.16), aligning with earlier studies (Rey, 2012; Sundararajan & Adesope, 2020) but reporting a smaller effect. This difference likely stems from methodological improvements: our multi-level model accounted for data dependencies which were overlooked in prior pairwise meta-analyses. Across achievement types, seductive details negatively affected recall (g = -0.17), comprehension (g = -0.19), and transfer (g = -0.12). While these outcomes vary in cognitive complexity—from lower-order (recall) to higher-order (comprehension and transfer) cognitive abilities—the differences in effect size were minimal (maximum Δg = 0.07). This suggests that seductive details consistently impair learning across cognitive levels, with no strong evidence that higher-order tasks are more affected than lower-order ones. The moderating effects (RQ3) Three significant moderators emerged. First, sample size influenced effects: small (N ≤ 50, g = -0.49) and large samples (N ≥ 100, g = -0.16) showed significant negative effects, while medium-sized samples did not, possibly due to statistical power issues or bias. Second, text language mattered: English (g = -0.23) and German (g = -0.14) texts produced negative effects, while Chinese texts showed no significant impact—likely due to limited data and cultural familiarity with seductive content in English. Third, applied environment moderated outcomes: seductive details in technology-based contexts (g = -0.19) impaired learning significantly, whereas paper-and-pen settings had no effect, suggesting that digital environments may amplify distraction and cognitive overload, while traditional formats promote sustained focus. The mediating effects (RQ4) The mediation analysis supports CTML by showing that seductive details increase extraneous cognitive load, which in turn hinders academic achievement. Both direct and indirect effects were significant, suggesting other mediators like situational interest. Instructionally, designers should balance engagement with cognitive efficiency, minimizing seductive elements—especially for novices or complex topics—to reduce load and enhance deep learning.

Cheung, M. W. L. (2015). metaSEM: An R package for meta-analysis using structural equation modeling. Frontiers in Psychology, 5. https://doi.org/10.3389/fpsyg.2014.01521 Cheung, M. W. L., & Chan, W. (2005). Meta-analytic structural equation modeling: A two-stage approach. Psychological Methods, 10(1), 40–64. https://doi.org/10.1037/1082-989x.10.1.40 Cheung, M. W. L., & Cheung, S. F. (2016). Random‐effects models for Meta‐Analytic Structural Equation Modeling: Review, issues, and illustrations. Research Synthesis Methods, 7(2), 140–155. https://doi.org/10.1002/jrsm.1166 Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Lawrence Erlbaum Associates. Colliot, T., & Boucheix, J. (2024). Exploring the effects of seductive details and illustration dynamics on young children’s performance in an origami task. Journal of Computer Assisted Learning, 40, 437-451. https://doi.org/10.1111/jcal.12879 Fries, L., DeCaro, M. S., & Ramirez, G.(2019). The lure of seductive details during lecture learning. Journal of Educational Psychology, 111, 736-749. http://dx.doi.org/10.1037/edu0000301 Harp, S., & Mayer, R. E. (1997). The role of interest in learning from scientific text and illustrations: On the distinction between emotional interest and cognitive interest. Journal of Educational Psychology, 89, 92–102. https://doi.org/10.1037/0022-0663.89.1.92 Harp, S. F., & Mayer, R. E. (1998). How seductive details do their damage: A theory of cognitive interest in science learning. Journal of Educational Psychology, 90, 434–441. https://doi.org/10.1037/0022-0663.90.3.414 Hedges, L. V. (1982). Estimation of effect size from a series of independent experiments. Psychological Bulletin, 92(2), 490–499. https://doi.org/10.1037/0033-2909.92.2.490 Mayer, R. E. (2014). Cognitive theory of multimedia learning. In R. E. Mayer (Ed.), The Cambridge handbook of multimedia learning (2nd ed., pp. 43-71). Cambridge University Press. https://doi.org/10.1017/CBO9781139547369.005 Mayer, R. E. (2019). Taking a new look at seductive details. Applied Cognitive Psychology, 33, 139-141. https://doi.org/10.1002/acp.3503 Mayer, R. E. (2021). Multimedia learning (3rd ed.). Cambridge University Press. Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred reporting items for systematic reviews and meta-analyses: The Prisma statement. Journal of Clinical Epidemiology, 62(10), 1006–1012. https://doi.org/10.1016/j.jclinepi.2009.06.005 Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S., … Moher, D. (2021). The Prisma 2020 statement: An updated guideline for reporting systematic reviews. International Journal of Surgery, 88, 105906. https://doi.org/10.1016/j.ijsu.2021.105906