基于学习分析的干预能提升学生学习效果吗?——基于49篇实验与准实验研究的元分析

Can Learning Analysis Interventions Improve Learning Outcomes?A Meta-Analysis Based on 49 Experimental and Quasi-Experimental Studies

干预是学习分析的一个重要环节,基于学习分析的干预能提升学生学习效果吗?国内外研究者已开展大量相关实证研究,但研究结论尚存在分歧。采用元分析方法,对2010〜2021年间49篇与基于学习分析的干预主题相关的实验与准实验研究文献进行量化统计分析,研究结果表明:基于学习分析的干预对学习效果具有中等偏上的正向影响,且相较于学业成绩和认知技能,情感态度的学习效果不明显;从干预规模来看,群体干预的效果最佳;从支持学习分析技术的学习环境而言,干预对学习效果均具有正向促进作用,但各组间在统计学上不存在显著差异;从干预策略而言,呈现可视化图表进行反馈的干预策略作用效果最为显著。基于上述发现,建议通过优化情感状态识别系统、在个人干预中应用进阶式干预及整合多种干预策略,提升干预效果。

As an important part of learning analysis,intervention is crucial to improve learning outcomes.However,can interventions based on learning analysis improve student learning outcomes?Domestic and foreign researchers have carried out a large number of empirical studies,while research conclusions have not yet been unified.In view of this,this study uses meta-analysis method to quantify and statistically analyze 49 experimental and quasi-experimental research articles related to learning analysis intervention from 2010 to 2021.It is found that the combined effect value is 0.499,which indicated that the intervention had a medium to upper positive impact on learning performance,and compared with academic performance and cognitive skills,the learning outcome of emotional attitude was not obvious.From the perspective of intervention scale,the group intervention has the best effect.In terms of learning environments that support learning analysis technologies,the intervention had a positive promoting effect on learning outcomes,and there was no significant difference between the groups.In terms of intervention strategies,the intervention strategy that presented visual chart for feedback had the most significant effect.Based on the above findings,it is recommended to optimize the affective state recognition system,to apply advanced intervention in individual intervention,and to integrate multiple intervention strategies to improve the effect of intervention.