A Reliability Generalization Meta-Analysis of the Survey of Attitudes Toward Statistics (SATS)

Students’ attitudes toward statistics play a crucial role in shaping their learning outcomes and long-term statistical thinking. The Survey of Attitudes Toward Statistics (SATS) has emerged as the most widely adopted tool in statistics education. This study aims to conduct the first reliability generalization meta-analysis of the SATS by synthesizing internal consistency estimates reported. Following the REGEMA (Reporting Guidelines for Meta-Analyses of Reliability Generalization Studies) guidelines, a systematic literature search was conducted across seven databases. A total of 41 eligible studies were identified, yielding 56 Cronbach’s alpha coefficients. A random-effects model was used to obtain pooled reliability estimates, and publication bias was assessed via funnel plots and Egger’s regression test. A mixed-effects meta-regression was conducted using several moderators, including geographic location and educational level. The SATS components demonstrated acceptable to excellent internal consistency. The findings confirm the SATS’ robust internal consistency and support their continued use in educational research.