Factorial Invariance of Statistics Anxiety Across Gender and Student Classification

• Activity Number: 196
• Type: Contributed
• Date/Time: Monday, August 2, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistical Education
• Author(s): Eric Teman*+ and Susan Hutchinson
• Companies: University of Northern Colorado and University of Northern Colorado
• Keywords: statistical anxiety rating scale ; measurement equivalence ; multi-group confirmatory factor analysis

Abstract

Statistics anxiety has been analyzed using various measurement tools and statistical methods over several decades. The statistical anxiety ratings scale is one of the most common methods for measuring said anxiety. However, there are two major limitations with past research. First, most studies implicitly assume measurement equivalence exists between two or more comparison groups. This assumption may be invalid and should first be tested to ensure meaningful comparisons. Second, some studies do account for measurement invariance as part of the analysis, but use incorrect methods that fail to account for the ordinal nature of the data. The current study uses multi-group confirmatory factor analysis via the Mplus software to test for measurement invariance between common groups of comparison, which include gender, student classification, and academic discipline.