Pub Date : 2025-12-08DOI: 10.1016/j.edurev.2025.100757
Quentin Brouhier , Isabel Raemdonck , Virginie März
Although professional development has been widely studied, the professional learning of late-career teachers remains an underexplored area. This mixed-methods systematic review synthesises findings from 12 peer-reviewed studies published between 2000 and 2024, focusing on teachers aged 50 and above. The analysis, structured around the Job Demands-Resources model, examined how professional learning is defined, experienced and supported in the later stages of teaching careers. The findings revealed significant inconsistencies in how late-career teachers are defined, whether by age, experience or perceived expertise, highlighting the lack of a coherent conceptualisation for this group. Similarly, professional learning was operationalised in various ways across studies, ranging from formal, structured programmes to informal, workplace-embedded practices. The review identifies late-career teachers' preference for socially embedded and collaborative learning opportunities, such as mentoring, dialogue and reflective practice. However, mentoring late-career teachers is typically seen as a one-way relationship where only the mentee learns, rather than as a shared learning experience in which both mentor and mentee grow together. Using the job demands-resources model, the review categorises key job demands (e.g., workload, role ambiguity) and resources (e.g., collegial support, autonomy, self-efficacy) that influence late-career teachers’ engagement in learning. It highlights a notable gap in tailored interventions addressing their specific needs. The review concludes with both theoretical and practical implications, advocating for more nuanced, context-sensitive approaches to professional learning that value the continued growth, contributions and well-being of teachers in the later stages of their careers.
{"title":"Professional learning of late-career teachers: A systematic literature review","authors":"Quentin Brouhier , Isabel Raemdonck , Virginie März","doi":"10.1016/j.edurev.2025.100757","DOIUrl":"10.1016/j.edurev.2025.100757","url":null,"abstract":"<div><div>Although professional development has been widely studied, the professional learning of late-career teachers remains an underexplored area. This mixed-methods systematic review synthesises findings from 12 peer-reviewed studies published between 2000 and 2024, focusing on teachers aged 50 and above. The analysis, structured around the Job Demands-Resources model, examined how professional learning is defined, experienced and supported in the later stages of teaching careers. The findings revealed significant inconsistencies in how late-career teachers are defined, whether by age, experience or perceived expertise, highlighting the lack of a coherent conceptualisation for this group. Similarly, professional learning was operationalised in various ways across studies, ranging from formal, structured programmes to informal, workplace-embedded practices. The review identifies late-career teachers' preference for socially embedded and collaborative learning opportunities, such as mentoring, dialogue and reflective practice. However, mentoring late-career teachers is typically seen as a one-way relationship where only the mentee learns, rather than as a shared learning experience in which both mentor and mentee grow together. Using the job demands-resources model, the review categorises key job demands (e.g., workload, role ambiguity) and resources (e.g., collegial support, autonomy, self-efficacy) that influence late-career teachers’ engagement in learning. It highlights a notable gap in tailored interventions addressing their specific needs. The review concludes with both theoretical and practical implications, advocating for more nuanced, context-sensitive approaches to professional learning that value the continued growth, contributions and well-being of teachers in the later stages of their careers.</div></div>","PeriodicalId":48125,"journal":{"name":"Educational Research Review","volume":"50 ","pages":"Article 100757"},"PeriodicalIF":10.6,"publicationDate":"2025-12-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145732704","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-06DOI: 10.1016/j.edurev.2025.100758
Callula Killingly, Sandra Matheson, Laura Bentley, Elizabeth Swanson
{"title":"Interventions for students with reading difficulties in Grades 4-12: A systematic review and meta-analysis","authors":"Callula Killingly, Sandra Matheson, Laura Bentley, Elizabeth Swanson","doi":"10.1016/j.edurev.2025.100758","DOIUrl":"https://doi.org/10.1016/j.edurev.2025.100758","url":null,"abstract":"","PeriodicalId":48125,"journal":{"name":"Educational Research Review","volume":"242 1","pages":""},"PeriodicalIF":11.7,"publicationDate":"2025-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145696611","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01DOI: 10.1016/j.edurev.2025.100748
Sebastian P. Suggate, Viktoria L. Karle, Tanja Kipfelsberger, Heidrun Stoeger
Evidence suggests that fine motor skills (FMS) relate to academic and cognitive development; however, findings are unclear, strewn across multiple disciplines, and lack adequate synthesis. We conducted the first comprehensive meta-analysis examining the links between different FMS facets (i.e., dexterity, speed, graphomotor, bimanual, general) and a broad range of academic-cognitive skills (i.e., sub-categories of reading, writing, mathematics, and cognition). A literature search identified 21,225 articles resulting in 118 eligible correlational studies (j = 143, k = 1110, N = 79,856). FMS exhibited significant and moderate relations with academic-cognitive skills (r = .329, d = .697). Graphomotor skills and writing showed the largest effect sizes. Mediation analysis suggested that cognitive skills mediated the link between FMS and academic skills. Overall, findings suggest that FMS share cognitive processes with academic skills, but that FMS are still related functionally to academic skills, especially writing. The discussion focuses on educational implications, moving beyond establishing if links exist to investigating why.
有证据表明，精细运动技能（FMS）与学业和认知发展有关；然而，研究结果不明确，散布在多个学科中，缺乏充分的综合。我们进行了第一次全面的荟萃分析，研究了FMS不同方面（即灵巧性、速度、书写运动、双手操作、一般）和广泛的学术认知技能（即阅读、写作、数学和认知的子类别）之间的联系。文献检索确定了21225篇文章，得到118项符合条件的相关研究（j = 143, k = 1110, N = 79856）。FMS与学业认知技能表现出显著和中等的关系（r = .329, d = .697）。笔迹运动技能和写作表现出最大的效应值。中介分析表明，认知技能在FMS与学业技能之间起中介作用。总体而言，研究结果表明，FMS与学术技能共享认知过程，但FMS在功能上仍与学术技能相关，尤其是写作。讨论的重点是教育意义，而不是确定是否存在联系，而是调查原因。
{"title":"Keep the hands in mind: A meta-analysis of correlations between fine motor skills and reading, writing, mathematics, and cognitive development in children and adolescents","authors":"Sebastian P. Suggate, Viktoria L. Karle, Tanja Kipfelsberger, Heidrun Stoeger","doi":"10.1016/j.edurev.2025.100748","DOIUrl":"10.1016/j.edurev.2025.100748","url":null,"abstract":"<div><div>Evidence suggests that fine motor skills (FMS) relate to academic and cognitive development; however, findings are unclear, strewn across multiple disciplines, and lack adequate synthesis. We conducted the first comprehensive meta-analysis examining the links between different FMS facets (i.e., dexterity, speed, graphomotor, bimanual, general) and a broad range of academic-cognitive skills (i.e., sub-categories of reading, writing, mathematics, and cognition). A literature search identified 21,225 articles resulting in 118 eligible correlational studies (<em>j</em> = 143, <em>k</em> = 1110, <em>N</em> = 79,856). FMS exhibited significant and moderate relations with academic-cognitive skills (<em>r</em> = .329, <em>d</em> = .697). Graphomotor skills and writing showed the largest effect sizes. Mediation analysis suggested that cognitive skills mediated the link between FMS and academic skills. Overall, findings suggest that FMS share cognitive processes with academic skills, but that FMS are still related functionally to academic skills, especially writing. The discussion focuses on educational implications, moving beyond establishing <em>if</em> links exist to investigating <em>why</em>.</div></div>","PeriodicalId":48125,"journal":{"name":"Educational Research Review","volume":"49 ","pages":"Article 100748"},"PeriodicalIF":10.6,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145575630","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01DOI: 10.1016/j.edurev.2025.100738
Nikoletta Mária Gulya , Anikó Fehérvári
The potential for relationships between students with disabilities and their typically developing peers is largely shaped by the attitudes of the typically developed students. This research aims to enhance understanding of these attitudes through a meta-analysis of results measured with the Chedoke-McMaster Attitudes towards Children with Handicaps Scale (CATCH). We examined changes in these attitudes from 1986 to the present, considering both temporal and geographic distribution, as well as the correlation of student gender, relationships with peers with disabilities, and the presence of a family member with a disability with these attitudes. A total of 40 studies were included in the research, analysing their quantitative data through weighted regression, t-tests, and meta-analysis. Our findings reveal a negative trend in students' average attitudes over the past 38 years. Gender shows a small positive effect on attitudes, while having a peer with a disability and having a family member with a disability exert a medium positive effect. Our results indicate that educational interventions should focus on knowledge sharing about learners with disabilities to provide mainstream students with relevant information, and additionally, create more opportunities for interaction between typical and disabled peers to foster understanding and friendships.
{"title":"The extent and change of attitudes towards students with disabilities: A meta-analysis of results measured with the Chedoke-McMaster attitudes towards children with handicaps scale","authors":"Nikoletta Mária Gulya , Anikó Fehérvári","doi":"10.1016/j.edurev.2025.100738","DOIUrl":"10.1016/j.edurev.2025.100738","url":null,"abstract":"<div><div>The potential for relationships between students with disabilities and their typically developing peers is largely shaped by the attitudes of the typically developed students. This research aims to enhance understanding of these attitudes through a meta-analysis of results measured with the Chedoke-McMaster Attitudes towards Children with Handicaps Scale (CATCH). We examined changes in these attitudes from 1986 to the present, considering both temporal and geographic distribution, as well as the correlation of student gender, relationships with peers with disabilities, and the presence of a family member with a disability with these attitudes. A total of 40 studies were included in the research, analysing their quantitative data through weighted regression, t-tests, and meta-analysis. Our findings reveal a negative trend in students' average attitudes over the past 38 years. Gender shows a small positive effect on attitudes, while having a peer with a disability and having a family member with a disability exert a medium positive effect. Our results indicate that educational interventions should focus on knowledge sharing about learners with disabilities to provide mainstream students with relevant information, and additionally, create more opportunities for interaction between typical and disabled peers to foster understanding and friendships.</div></div>","PeriodicalId":48125,"journal":{"name":"Educational Research Review","volume":"49 ","pages":"Article 100738"},"PeriodicalIF":10.6,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145416087","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01DOI: 10.1016/j.edurev.2025.100737
Shuzhen Chen, Alan C.K. Cheung
Generative artificial intelligence (Gen-AI) is transforming higher education by enhancing students' learning outcomes. Noteworthy, the magnitude of effect sizes has not reached consensus, reflecting substantial variability arising from contextual conditions and methodological approaches. To bridge this gap, grounded in activity theory-mobile computer-supported collaborative learning (AT-MCSCL) framework, this meta-analysis comprehensively synthesized 57 studies and 97 estimations and confirmed that Gen-AI had large effect size on university students’ learning outcomes (g+ = 0.804), particularly in language skills (g+ = 2.331), academic achievement (g+ = 0.633), affective-motivational status (g+ = 0.617), and higher-order thinking (g+ = 0.580), except for a lack of statistically significant effect on metacognition (g+ = 0.078). Meanwhile, this study revealed moderators related to learner, tool, roles, rules and contextual features: (a) freshmen and language learners benefited the most; (b) outcomes measured by tests, studies with medium-sized samples (40–100), medium (4–12 weeks) intervention, and published in journals yielded the higher effect sizes; (c) studies conducted in classroom settings, Middle East, low ICT levels, low socioeconomic status, and higher power distance produced higher effect sizes. This study provides more robust evidence to effectiveness of Gen-AI in higher education and elucidates contextual heterogeneity comprehensively, offering reliable insights for Gen-AI integration in diverse learning environments.
{"title":"Effect of generative artificial intelligence on university students learning outcomes: A systematic review and meta-analysis","authors":"Shuzhen Chen, Alan C.K. Cheung","doi":"10.1016/j.edurev.2025.100737","DOIUrl":"10.1016/j.edurev.2025.100737","url":null,"abstract":"<div><div>Generative artificial intelligence (Gen-AI) is transforming higher education by enhancing students' learning outcomes. Noteworthy, the magnitude of effect sizes has not reached consensus, reflecting substantial variability arising from contextual conditions and methodological approaches. To bridge this gap, grounded in activity theory-mobile computer-supported collaborative learning (AT-MCSCL) framework, this meta-analysis comprehensively synthesized 57 studies and 97 estimations and confirmed that Gen-AI had large effect size on university students’ learning outcomes (<em>g</em><sup>+</sup> = 0.804), particularly in language skills (<em>g</em><sup>+</sup> = 2.331), academic achievement (<em>g</em><sup>+</sup> = 0.633), affective-motivational status (<em>g</em><sup>+</sup> = 0.617), and higher-order thinking (<em>g</em><sup>+</sup> = 0.580), except for a lack of statistically significant effect on metacognition (<em>g</em><sup>+</sup> = 0.078). Meanwhile, this study revealed moderators related to learner, tool, roles, rules and contextual features: (a) freshmen and language learners benefited the most; (b) outcomes measured by tests, studies with medium-sized samples (40–100), medium (4–12 weeks) intervention, and published in journals yielded the higher effect sizes; (c) studies conducted in classroom settings, Middle East, low ICT levels, low socioeconomic status, and higher power distance produced higher effect sizes. This study provides more robust evidence to effectiveness of Gen-AI in higher education and elucidates contextual heterogeneity comprehensively, offering reliable insights for Gen-AI integration in diverse learning environments.</div></div>","PeriodicalId":48125,"journal":{"name":"Educational Research Review","volume":"49 ","pages":"Article 100737"},"PeriodicalIF":10.6,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145465209","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01DOI: 10.1016/j.edurev.2025.100741
Xiaofan Liu, Baichang Zhong
Advances in Generative Artificial Intelligence (GenAI) have sparked widespread debate in education. According to a survey by the Digital Education Council (DEC), 86 percent of students use AI in their learning, with ChatGPT being the most widely used tool. Given that students constitute a considerable proportion of GenAI users, concerns arise regarding the potential misuse of this technology by students themselves. Consequently, researchers and educators should rethink how GenAI tools interacts with curriculum needs and pedagogical strategies, particularly through frameworks such as TPACK, to enhance student learning while mitigating potential risks. However, the state-of-the-art of integrating GenAI into student learning remains unclear. To address this gap, this study systematically reviewed high-quality empirical studies published after 30 November 2022 (release date of ChatGPT) on integrating GenAI into student learning from a TPACK perspective. The findings aim to provide an evidence-based foundation for future research and instructional design. For the former, we analyzed the 71 studies from the perspectives of publishing year, country/region of corresponding author, authorship type, sample size, learning stage, educational setting, duration, research method and focus, as well as the moderating effect of key research design variables on learning outcomes. For the latter, we analyzed the 71 studies from the perspectives of learning content, learning tool, pedagogical strategy, assessment method, and learning outcome, as well as the moderating effect of key instructional design variables on learning outcomes. Moreover, we also summarized the recommendations of these studies for future research and instruction. The empirical evidence suggested that GenAI has the potential to enhance student learning in cognitive, affective and behavioral domains. However, deficiencies in research and instructional design still remain, including short duration, small sample sizes, and incomplete pedagogical scaffolding, etc. Based on the review's findings, this study also identifies two critical priorities for future research and instruction.
{"title":"Integrating generative Artificial Intelligence into student learning: A systematic review from a TPACK perspective","authors":"Xiaofan Liu, Baichang Zhong","doi":"10.1016/j.edurev.2025.100741","DOIUrl":"10.1016/j.edurev.2025.100741","url":null,"abstract":"<div><div>Advances in Generative Artificial Intelligence (GenAI) have sparked widespread debate in education. According to a survey by the Digital Education Council (DEC), 86 percent of students use AI in their learning, with ChatGPT being the most widely used tool. Given that students constitute a considerable proportion of GenAI users, concerns arise regarding the potential misuse of this technology by students themselves. Consequently, researchers and educators should rethink how GenAI tools interacts with curriculum needs and pedagogical strategies, particularly through frameworks such as TPACK, to enhance student learning while mitigating potential risks. However, the state-of-the-art of integrating GenAI into student learning remains unclear. To address this gap, this study systematically reviewed high-quality empirical studies published after 30 November 2022 (release date of ChatGPT) on integrating GenAI into student learning from a TPACK perspective. The findings aim to provide an evidence-based foundation for future research and instructional design. For the former, we analyzed the 71 studies from the perspectives of publishing year, country/region of corresponding author, authorship type, sample size, learning stage, educational setting, duration, research method and focus, as well as the moderating effect of key research design variables on learning outcomes. For the latter, we analyzed the 71 studies from the perspectives of learning content, learning tool, pedagogical strategy, assessment method, and learning outcome, as well as the moderating effect of key instructional design variables on learning outcomes. Moreover, we also summarized the recommendations of these studies for future research and instruction. The empirical evidence suggested that GenAI has the potential to enhance student learning in cognitive, affective and behavioral domains. However, deficiencies in research and instructional design still remain, including short duration, small sample sizes, and incomplete pedagogical scaffolding, etc. Based on the review's findings, this study also identifies two critical priorities for future research and instruction.</div></div>","PeriodicalId":48125,"journal":{"name":"Educational Research Review","volume":"49 ","pages":"Article 100741"},"PeriodicalIF":10.6,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145434381","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01DOI: 10.1016/j.edurev.2025.100746
Shunan Zhang , Yincen Li , Guangji Gan , Stevany , Sirui Pang , Jang Hyun Kim
According to the social agency theory (Moreno et al., 2001), the social cues of artificial intelligence-powered pedagogical agents (AIPAs) hold detectable potential for enhancing learning. However, previous studies have reported inconsistent results regarding the effectiveness of these social cues. This study utilized a multilevel meta-analysis to integrate the results of 33 experimental studies on the social cues of AIPAs and explore their effectiveness. The results indicated that the social cues of AIPAs were associated with statistically detectable positive effects on cognitive learning outcomes (k = 66, g = 0.359, 95 % confidence interval (CI) = [0.191–0.526]), affective learning outcomes (k = 85, g = 0.519, 95 % CI = [0.226–0.812]), and behavioral learning outcomes (k = 23, g = 0.458, 95 % CI = [0.169–0.746]). No statistically detectable change in cognitive load was observed (k = 11, g = 0.039, 95 % CI = [-0.153–0.230]). Additionally, 16 potential moderating variables related to AIPAs and non-AIPAs were identified and discussed. For cognitive learning outcomes, the type of the pedagogical agent (F = 5.420, p = .007, pattern: virtual agents with visual representation > physically embodied robots > virtual agents without visual representation), the image of the pedagogical agent (F = 7.782, p < .001, pattern: anthropomorphized > semi-anthropomorphized > non-anthropomorphized), gender of the pedagogical agent (F = 9.309, p < .001, pattern: N/A > male > female), instructional setting (F = 9.869, p < .001, classroom-based > laboratory > online), interaction type (F = 19.195, p < .001, pattern: one-to-many > one-to-one), and educational level (F = 7.134, p < .001, pattern: K–12 > university <N/A < preschool) showed statistically detectable differences in effect sizes. For affective learning outcomes, only the publication type yielded a statistically detectable difference (F = 4.803, p = .031, pattern: conference papers > journal papers). The findings of this study provide insights into the utilization and design of social cues in AIPAs for educational purposes.
根据社会代理理论（Moreno et al., 2001），人工智能驱动的教学代理（aipa）的社会线索具有可检测的增强学习的潜力。然而，之前的研究报告了关于这些社会线索的有效性的不一致的结果。本研究采用多水平荟萃分析的方法，对33项关于aipa社会线索的实验研究结果进行整合，探讨其有效性。结果表明，aipa的社会线索对认知学习结果（k = 66, g = 0.359, 95%可信区间(CI) =[0.191-0.526]）、情感学习结果（k = 85, g = 0.519, 95% CI =[0.226-0.812]）和行为学习结果（k = 23, g = 0.458, 95% CI =[0.169-0.746]）具有统计学上可检测的积极影响。认知负荷未见统计学上可检测的变化（k = 11, g = 0.039, 95% CI =[-0.153-0.230]）。此外，确定并讨论了与aipa和非aipa相关的16个潜在调节变量。对于认知学习成果，教学主体的类型（F = 5.420, p = .007，模式：具有视觉表征的虚拟主体>；物理体现的机器人>；没有视觉表征的虚拟主体>）、教学主体的形象（F = 7.782, p < .001，模式：拟人化>；半拟人化>；非拟人化）、教学主体的性别(F = 9.309, p < .001，模式：N/A >；男>；女)、教学环境（F = 9.869, p >；实验室>；在线）、互动类型（F = 19.195, p < .001，模式：一对多>；一对一）、教育水平（F = 7.134, p < .001，模式：K-12 >；大学<；N/A <；学前）的效应量差异具有统计学意义。对于情感学习结果，只有出版类型产生了统计学上可检测的差异（F = 4.803, p = 0.031，模式：会议论文>；期刊论文）。本研究的结果为以教育为目的的aipa中社会线索的使用和设计提供了见解。
{"title":"Effects of social cues of artificial intelligence-powered pedagogical agents: A multilevel meta-analysis","authors":"Shunan Zhang , Yincen Li , Guangji Gan , Stevany , Sirui Pang , Jang Hyun Kim","doi":"10.1016/j.edurev.2025.100746","DOIUrl":"10.1016/j.edurev.2025.100746","url":null,"abstract":"<div><div>According to the social agency theory (Moreno et al., 2001), the social cues of artificial intelligence-powered pedagogical agents (AIPAs) hold detectable potential for enhancing learning. However, previous studies have reported inconsistent results regarding the effectiveness of these social cues. This study utilized a multilevel meta-analysis to integrate the results of 33 experimental studies on the social cues of AIPAs and explore their effectiveness. The results indicated that the social cues of AIPAs were associated with statistically detectable positive effects on cognitive learning outcomes (k = 66, g = 0.359, 95 % confidence interval (CI) = [0.191–0.526]), affective learning outcomes (k = 85, g = 0.519, 95 % CI = [0.226–0.812]), and behavioral learning outcomes (k = 23, g = 0.458, 95 % CI = [0.169–0.746]). No statistically detectable change in cognitive load was observed (k = 11, g = 0.039, 95 % CI = [-0.153–0.230]). Additionally, 16 potential moderating variables related to AIPAs and non-AIPAs were identified and discussed. For cognitive learning outcomes, the type of the pedagogical agent (F = 5.420, p = .007, pattern: virtual agents with visual representation > physically embodied robots > virtual agents without visual representation), the image of the pedagogical agent (F = 7.782, p < .001, pattern: anthropomorphized > semi-anthropomorphized > non-anthropomorphized), gender of the pedagogical agent (F = 9.309, p < .001, pattern: N/A > male > female), instructional setting (F = 9.869, p < .001, classroom-based > laboratory > online), interaction type (F = 19.195, p < .001, pattern: one-to-many > one-to-one), and educational level (F = 7.134, p < .001, pattern: K–12 > university <N/A < preschool) showed statistically detectable differences in effect sizes. For affective learning outcomes, only the publication type yielded a statistically detectable difference (F = 4.803, p = .031, pattern: conference papers > journal papers). The findings of this study provide insights into the utilization and design of social cues in AIPAs for educational purposes.</div></div>","PeriodicalId":48125,"journal":{"name":"Educational Research Review","volume":"49 ","pages":"Article 100746"},"PeriodicalIF":10.6,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145519755","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01DOI: 10.1016/j.edurev.2025.100747
Weiwei Liu, Valentina C. Tassone, Pascal Frank, Renate Wesselink, Harm J.A. Biemans
Transdisciplinary learning is increasingly recognized as a critical educational approach for preparing future change agents in sustainability higher education. However, existing research remains fragmented, particularly regarding the subjectively experienced learning challenges and the mediating role of agency in addressing them. This scoping literature review addressed this gap by mapping the learning challenges encountered and managed by three key participant groups - students, educational staff, and societal partners - in transdisciplinary sustainability education. Analyzing 36 studies published between 2006 and 2024, the review highlights a notable underrepresentation of the learning experiences of educational staff and societal partners. It identified challenges based on the four aspects of transdisciplinary learning for sustainability - responding to sustainability issues, engaging non-academic partners, integrating disciplinary knowledge and capacity, and incorporating multiple perspectives - as well as three general challenges: navigating uneven relationships, managing high time demands, and adapting to unfamiliarity with transdisciplinarity. These challenges encompass both emotional and cognitive dimensions, with uncertainty emerging as a central feature stemming from the complexity of sustainability issues and the unfamiliarity of the transdisciplinary learning process. In response to these challenges, participants demonstrate agency through actions aimed at personal development (toward the self), empowering other (through others) and building intersubjectivity (toward groups). These actions demonstrate participants’ capacity to foster intrapersonal changes and interpersonal relationships. The findings offer valuable insights for instructors and course designers seeking to cultivate agency among all categories of participants in transdisciplinary sustainability education.
{"title":"The role of human agency in dealing with challenges in transdisciplinary learning for sustainability in higher education: A scoping literature review","authors":"Weiwei Liu, Valentina C. Tassone, Pascal Frank, Renate Wesselink, Harm J.A. Biemans","doi":"10.1016/j.edurev.2025.100747","DOIUrl":"10.1016/j.edurev.2025.100747","url":null,"abstract":"<div><div>Transdisciplinary learning is increasingly recognized as a critical educational approach for preparing future change agents in sustainability higher education. However, existing research remains fragmented, particularly regarding the subjectively experienced learning challenges and the mediating role of agency in addressing them. This scoping literature review addressed this gap by mapping the learning challenges encountered and managed by three key participant groups - students, educational staff, and societal partners - in transdisciplinary sustainability education. Analyzing 36 studies published between 2006 and 2024, the review highlights a notable underrepresentation of the learning experiences of educational staff and societal partners. It identified challenges based on the four aspects of transdisciplinary learning for sustainability - responding to sustainability issues, engaging non-academic partners, integrating disciplinary knowledge and capacity, and incorporating multiple perspectives - as well as three general challenges: navigating uneven relationships, managing high time demands, and adapting to unfamiliarity with transdisciplinarity. These challenges encompass both emotional and cognitive dimensions, with uncertainty emerging as a central feature stemming from the complexity of sustainability issues and the unfamiliarity of the transdisciplinary learning process. In response to these challenges, participants demonstrate agency through actions aimed at personal development (toward the self), empowering other (through others) and building intersubjectivity (toward groups). These actions demonstrate participants’ capacity to foster intrapersonal changes and interpersonal relationships. The findings offer valuable insights for instructors and course designers seeking to cultivate agency among all categories of participants in transdisciplinary sustainability education.</div></div>","PeriodicalId":48125,"journal":{"name":"Educational Research Review","volume":"49 ","pages":"Article 100747"},"PeriodicalIF":10.6,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145575661","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01DOI: 10.1016/j.edurev.2025.100745
Zhujun Jiang, Feng-Kuang Chiang
Text–picture reading is fundamental to science education, and eye-tracking research provides unique insights into the cognitive and metacognitive mechanisms underlying this process. This systematic review synthesized 66 studies, including 16 studies (82 effect sizes) incorporated into a meta-analysis, to present a comprehensive overview of the field. First, the review examined methodological characteristics concerning disciplines focus, participant demographics, and stimuli types. Within K–12 contexts, biology emerged as the most frequently investigated discipline, whereas chemistry remained underrepresented. Elementary school students were the most commonly studied population, and diagrams were the predominant stimulus type. Areas of interest (AOIs) were typically defined by a text–picture division, with finer-grained categorizations encompassing structural layout, functional elements, semantic content, and text–picture correspondence. Second, the review synthesized findings on learners’ processing patterns. These studies suggested a text-preference effect, with high-level learners generally showing more selective attention, strategic cross-modal integration, and metacognitive control, whereas low-level learners demonstrated localized fixations and insufficient self-regulation, requiring external scaffolds for effective integration. Third, the meta-analysis confirmed that high-level learners focused more on pictures than did low-level learners, as reflected in fixation counts (g = 0.473) and fixation proportions (g = 0.365), and performed more saccades between text and pictures (g = 0.420). In contrast, low-level learners focused more on text, as reflected by their fixation proportions (g = −0.255). Overall, this review synthesizes methodological insights and empirical evidence on science reading, providing informed directions for advancing research design, instructional practices, and learner development in multimedia science education.
{"title":"A systematic review and meta-analysis of research on eye-tracking of K–12 students during scientific text–picture reading","authors":"Zhujun Jiang, Feng-Kuang Chiang","doi":"10.1016/j.edurev.2025.100745","DOIUrl":"10.1016/j.edurev.2025.100745","url":null,"abstract":"<div><div>Text–picture reading is fundamental to science education, and eye-tracking research provides unique insights into the cognitive and metacognitive mechanisms underlying this process. This systematic review synthesized 66 studies, including 16 studies (82 effect sizes) incorporated into a meta-analysis, to present a comprehensive overview of the field. First, the review examined methodological characteristics concerning disciplines focus, participant demographics, and stimuli types. Within K–12 contexts, biology emerged as the most frequently investigated discipline, whereas chemistry remained underrepresented. Elementary school students were the most commonly studied population, and diagrams were the predominant stimulus type. Areas of interest (AOIs) were typically defined by a text–picture division, with finer-grained categorizations encompassing structural layout, functional elements, semantic content, and text–picture correspondence. Second, the review synthesized findings on learners’ processing patterns. These studies suggested a text-preference effect, with high-level learners generally showing more selective attention, strategic cross-modal integration, and metacognitive control, whereas low-level learners demonstrated localized fixations and insufficient self-regulation, requiring external scaffolds for effective integration. Third, the meta-analysis confirmed that high-level learners focused more on pictures than did low-level learners, as reflected in fixation counts (g = 0.473) and fixation proportions (g = 0.365), and performed more saccades between text and pictures (g = 0.420). In contrast, low-level learners focused more on text, as reflected by their fixation proportions (g = −0.255). Overall, this review synthesizes methodological insights and empirical evidence on science reading, providing informed directions for advancing research design, instructional practices, and learner development in multimedia science education.</div></div>","PeriodicalId":48125,"journal":{"name":"Educational Research Review","volume":"49 ","pages":"Article 100745"},"PeriodicalIF":10.6,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145531158","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01DOI: 10.1016/j.edurev.2025.100740
Sølvi Mausethagen, Hege Hermansen, Hilde Afdal, Marte Lorentzen, Hanna Holmeide
This article presents a systematic critical review of the research literature on ‘research use’ in education. The concept of research use has gradually developed with increased policy expectations that teachers should engage in externally produced research. With such expectations, it is important to ask how existing research positions the teaching profession in relation to ‘research’ and its ‘use’. The characteristics of these relations have implications for how the professional knowledge and practices of teachers are conceptualised. Based on a full-text analysis of 34 articles published in renowned databases and peer-reviewed English scholarly journals, we find that research on research use is characterised by similar problem framings and results over a 30-year period. Based on the analysis, we argue that there is a need for more profession-sensitive conceptualisations of research use in education and for more complex understandings of what it entails to employ a system perspective on research use. We argue that the next generation of studies on research use must be developed to include a conception of the teaching profession and of teachers' professional knowledge.
{"title":"A systematic critical review of research on ‘research use’ in education: Towards more profession-sensitive conceptualisations","authors":"Sølvi Mausethagen, Hege Hermansen, Hilde Afdal, Marte Lorentzen, Hanna Holmeide","doi":"10.1016/j.edurev.2025.100740","DOIUrl":"10.1016/j.edurev.2025.100740","url":null,"abstract":"<div><div>This article presents a systematic critical review of the research literature on ‘research use’ in education. The concept of research use has gradually developed with increased policy expectations that teachers should engage in externally produced research. With such expectations, it is important to ask how existing research positions the teaching profession in relation to ‘research’ and its ‘use’. The characteristics of these relations have implications for how the professional knowledge and practices of teachers are conceptualised. Based on a full-text analysis of 34 articles published in renowned databases and peer-reviewed English scholarly journals, we find that research on research use is characterised by similar problem framings and results over a 30-year period. Based on the analysis, we argue that there is a need for more profession-sensitive conceptualisations of research use in education and for more complex understandings of what it entails to employ a system perspective on research use. We argue that the next generation of studies on research use must be developed to include a conception of the teaching profession and of teachers' professional knowledge.</div></div>","PeriodicalId":48125,"journal":{"name":"Educational Research Review","volume":"49 ","pages":"Article 100740"},"PeriodicalIF":10.6,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145397731","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号