Validation of an Adapted Version of the Statistical Anxiety Scale in English and Its Relationship to Attitudes Toward Statistics

Abstract

Copyright © The Author(s) 2021. The aims of this study were to validate an instrument that measured statistics anxiety and to examine how attitudes toward statistics predict statistics anxiety using the Attitudes Toward Statistics (ATS) Scale for a sample of 323 undergraduate social science majors enrolled in colleges and universities in the United States. A confirmatory factor analysis suggested retaining a revised two-factor model of the Statistical Anxiety Scale (SAS) to measure statistics anxiety, namely, help and interpretation anxiety (𝜒2𝑆𝐵 = 49.37, df = 38.13, p = .105, comparative fit index [CFI] = .959, standardized root mean square residual [SRMR] = .035, root mean square error of approximation [RMSEA] = .076). An examination of discriminant validity of the scores of the SAS with scores of the ATS subscales revealed that statistics anxiety and attitudes toward statistics are distinct constructs. Structural equational modeling was used to determine whether attitude toward course and attitude toward field were predictors of examination anxiety and asking for help anxiety. Of the two factors of the ATS scale, attitudes toward field and attitudes toward course, the latter predicted examination anxiety better than the former did, although both were moderate predictors of examination anxiety. We recommend that statistics educators consider the role of statistics anxiety as well as attitudes toward statistics and the field when designing their pedagogical approach.