Psychometric Properties of the Turkish Version of the Survey of Attitudes Toward Statistics

• Activity Number: 342
• Type: Topic Contributed
• Date/Time: Tuesday, August 6, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistical Education
• Author(s): Esma Emmioglu*+
• Companies: Simon Fraser University
• Keywords: Survey of Attitudes Toward Statistics

Abstract

This study examines the psychometric properties of the Turkish version of the Survey of Attitudes toward Statistics (SATS-36©). The SATS-36© is the most current and widely used instrument that measures students' attitudes toward statistics. It contains 36 items and six components: Affect, Cognitive Competence, Value, Difficulty, Effort, and Interest. Back translation method was used to create the Turkish version of the survey. Data were collected from (n=347, 59% female and 36% male, 5% unreported) Turkish speaking students in an English medium university in Turkey. Confirmatory factor analysis results revealed that the six factor structure had a very good fit to the data. The survey was gender invariant in terms of factor loadings, intercepts and factor variances/covariances. Internal consistency coefficients supported the high score reliability for each subscale. Overall, the study presented evidence for the score reliability and validity of the Turkish version of the SATS-36© when applied to Turkish speaking students enrolled in a Turkish university.