The impact of alternative assessment on non-affective learning outcomes: a meta-analysis of experimental and quasi-experimental studies

Abstract

Abstract:

With the increasing interest of researchers in alternative assessment methods, their widespread adoption, and the abundance of related research, this study aims to compile and analyze the findings of these studies comprehensively to arrive at a holistic conclusion and provide a comprehensive overview of their results. Specifically, this study seeks to reveal the overall impact of using alternative assessment on non-affective learning outcomes. It also investigates the varying impact sizes of different types of alternative assessment (self-assessment, peer assessment, and technology-based assessment) on these non-affective learning outcomes.

Therefore, the researcher relied on meta-analysis, collecting numerous published research papers and studies according to a set of criteria obtained from a literature review, and using the PRISMA framework. After identifying the research sample, the risk of bias was assessed using Cochrane tools, excluding studies classified as high risk. A precise analysis card, developed by the researcher after confirming its validity and reliability, was then used. This card was initially used for qualitative analysis of the research and then converted to quantitative values ??to facilitate the use of the data in the meta-analysis.

The results showed that the overall effect size of using alternative assessment on non-affective learning outcomes was moderate (0.44). The results also showed no statistically significant differences between effect sizes attributable to the assessment method used; that is, the results did not reveal differences in effect sizes between self-assessment, peer assessment, and technology-supported assessment on non-affective learning outcomes. Therefore, there remains an unexplained variability between the studies, requiring further research to explore the moderating factors.