Reliability Generalization Meta-Analysis of Mathematics Anxiety Scale for Primary School Students

The purpose of this study was to estimate the mean value for the reliability coefficients reported by the studies using the Mathematics Anxiety Scale for Primary School Students (MASPSS) and to examine the sources of the variation of the reliability coefficients reported in each study. A reliability generalization meta-analysis study was conducted by combining the Cronbach's alpha values of 34 studies that met the inclusion criteria. The Cronbach's alpha values used in the studies were converted to the transformed coefficient values by applying the Bonett transformation, and the analyses were carried out under the random-effects model. The mean Cronbach's alpha value of the MASPSS across 34 studies was found to be .855 (95% CI: .841-.869), and this result was statistically significant (p<.01). According to the results, there was a lack of publication bias in this meta-analysis study. Moderator analyses were conducted to explain the possible sources of heterogeneity across the individual studies. Findings revealed that the Cronbach's alpha estimates did not show any statistical differences based on publication year, female percentage, publication type and research method variables. It was found that the sample type affected the estimation of Cronbach's alpha coefficient. In addition, suggestions were made for psychometric studies that would use the MASPSS in the future.