Scheduled for Research Consortium Pedagogy II Poster Session, Thursday, April 27, 2006, 1:15 PM - 2:45 PM, Convention Center: Exhibit Hall Poster Area I


Revalidation of Physical Education Teacher Attitudes Toward Fitness Test Scales

Xiaofen Keating, California State University, Los Angeles, Lake Forest, CA, Robert H. Ferguson, University of Arkansas, Fayetteville, AR, Jianmin Guan, Texas-San Antonio/Univ Of, San Antonio, TX, Li Chen, Delaware State University, Dover, DE and Dwan Bridges, California State University, Los Angeles, Los Angeles, CA

Reliable and valid instruments are important to the quality of research on youth fitness testing in school-based physical education programs. The Physical Education Teacher Attitudes toward Fitness Tests Scale (PETAFTS) (Keating & Silverman, 2004) has been used to examine teacher attitudes toward fitness tests. The scale consisted of 16 items measuring affective and cognitive components: affects toward fitness tests, and beliefs in the usefulness of fitness test results. There were two subdomains in the affective component (i.e., affects of fitness test implementation, and affects of using test results). A 7-point Likert-type scale was used to assess the strength of attitudes. The purpose of the study was to re-validate the PETAFTS in order to utilize the scale with more confidence. The 16-item scale was re-tested using 469 full-time physical education teachers from two southwestern states. Alpha was computed for the reliability and confirmative factor analysis was used to examine the construct validity. Alpha (.66) for the subdomain of affects of using fitness test results was not acceptable, nor were the major indices for confirmative factor analysis with 2-independent factors using the entire 16 items (comparative fit index [CFI] = .81, the Tucker-Lewis Index [TLI] = .74, and the root mean squared error of approximation [RMSEA] = .10). After deleting the subdomain of affects of using fitness test results (i.e., 4 items in total), the alphas for the total scale (.89), the affective (.84) and cognitive (.78) components were acceptable. The confirmative factor analysis indices were also within acceptable range (i.e., CFI = .93, TLI = .90, and RMSEA = .08). The data from the study suggest that the items in the subdomain of affects of using fitness test results in the affective component might be cross-loading items with those in the cognitive component. The 12-item scale with the affective and cognitive components demonstrated robust reliability and validity.
Keyword(s): assessment, exercise/fitness, physical activity

Back to the 2006 AAHPERD National Convention and Exposition