Abstract
Objective: To determine the effects of formative assessment interventions and strategies aimed at optimising the learning outcomes in children and adolescents.
Design: A systematic review and meta-analysis.
Data sources: After searching PsycInfo, ERIC, Education Research Complete, A+ Education and Scopus electronic university databases, there were 3689 potentially eligible studies published between January 1, 2002 to December 31, 2022.
Eligibility criteria for selecting studies: This study includes randomised controlled trials, quasi-experimental studies, and controlled trials that assessed the effect of a formative assessment-based strategy or intervention against learning outcomes in students attending either primary or secondary years of schooling (i.e., aged 5–18 years).
Results: Thirty-eight (38) studies with over 16,000 participants and 43 calculated effect sizes were included in the final analysis. The sample size adjusted mean effect across all the learning and development outcomes was of a small to medium magnitude (Hedges’ g=0.31), 95% confidence interval [CI] (0.29 to 0.39). After applying the Duval and Tweedie (2000) Trim and Fill Method to account for publication bias, the average effect size was adjusted to g=0.16. Effect sizes also varied significantly based on whether they targeted cognitive, psychomotor, or affective learning outcomes. Only a small proportion (less than 10%) of formative assessment interventions examined in this systematic review indicated a detrimental impact on student learning and development.
Conclusion: Formative assessment interventions consistently have a positive effect on student learning in both primary and secondary school settings across psychomotor, affective, and cognitive learning domains. However, the reported effect of formative assessment interventions in previous meta-analytic studies appears to have been inflated by the inclusion of studies that suffer from weaker study designs. Furthermore, the scope of studies included in this systematic review and meta-analysis showed even studies with stronger study designs were plagued with poor methodological quality and substantial publication bias. Moreover, the evidence reported is limited by consistency in intervention dosage, study design, and data collection instruments.
Design: A systematic review and meta-analysis.
Data sources: After searching PsycInfo, ERIC, Education Research Complete, A+ Education and Scopus electronic university databases, there were 3689 potentially eligible studies published between January 1, 2002 to December 31, 2022.
Eligibility criteria for selecting studies: This study includes randomised controlled trials, quasi-experimental studies, and controlled trials that assessed the effect of a formative assessment-based strategy or intervention against learning outcomes in students attending either primary or secondary years of schooling (i.e., aged 5–18 years).
Results: Thirty-eight (38) studies with over 16,000 participants and 43 calculated effect sizes were included in the final analysis. The sample size adjusted mean effect across all the learning and development outcomes was of a small to medium magnitude (Hedges’ g=0.31), 95% confidence interval [CI] (0.29 to 0.39). After applying the Duval and Tweedie (2000) Trim and Fill Method to account for publication bias, the average effect size was adjusted to g=0.16. Effect sizes also varied significantly based on whether they targeted cognitive, psychomotor, or affective learning outcomes. Only a small proportion (less than 10%) of formative assessment interventions examined in this systematic review indicated a detrimental impact on student learning and development.
Conclusion: Formative assessment interventions consistently have a positive effect on student learning in both primary and secondary school settings across psychomotor, affective, and cognitive learning domains. However, the reported effect of formative assessment interventions in previous meta-analytic studies appears to have been inflated by the inclusion of studies that suffer from weaker study designs. Furthermore, the scope of studies included in this systematic review and meta-analysis showed even studies with stronger study designs were plagued with poor methodological quality and substantial publication bias. Moreover, the evidence reported is limited by consistency in intervention dosage, study design, and data collection instruments.
| Original language | English |
|---|---|
| Qualification | Master of Science |
| Awarding Institution |
|
| Supervisors/Advisors |
|
| Award date | 7 Dec 2023 |
| Publication status | Unpublished - 2023 |
| Externally published | Yes |