Measuring Implicit Attitudes of High School Students Toward Physical Education

Background/Purpose: If an attitude is measured without the participant’s filtering, or without full consideration so that it has elements of automatic expression it is considered implicit.  Implicit attitudes have been found to be separate constructs, compared to explicit attitudes, in physical activity and individuals’ exercise identities.  The purpose of this study was to evaluate the validity of scores and practicality of measuring students’ implicit attitudes with web-based tools for high school physical education (HSPE).

Method: The scores for the implicit instruments were validated through extensive, multiple pilot studies and using the Theory of Planned Behavior to guide the development of constructs and items. Two different types of implicit measures (the Single Category Implicit Association Test [SCIAT] and the Go / No-go Association Test [GNAT] were employed to measure implicit attitudes toward HSPE.  Both were used with piloted photographs of HSPE classes to create two physical education instruments (SCIATPE and GNATPE) and elicited automatic attitude responses.  A widely used and previously validated measure of students’ explicit attitudes (SAtPE) was also utilized to facilitate comparison with explicit attitude.  Demographic questions and all measures were completed consecutively in school computer labs under the supervision of HSPE teachers utilizing a script.  The final validation study included students (N = 434) in seven high schools.

Analysis/Results: A hypothesized model of overall attitude toward physical education containing implicit and explicit latent variables was tested using structural equation modeling (SEM) with AMOS.  All three measures had scores that were reliable (SCIATPE: α = .829; GNATPE: α = .922; and SAtPE: α = .953) using the computerized system.  PCA performed with scores from both implicit measures and the SAtPE’s four subcomponent scores yielded a clean two component Varimax-rotated solution (Determinant = .014, KMO = .764, Bartlett’s (15) = 1063.436, p < .001) indicating implicit attitude is a unique construct.  An eight variable SEM analysis utilizing four explicit subcomponent scores and two sub-scores from each of the implicit measures produced good fit statistics (X²(16) = 29.830, p = .019; CFI = .989, RMSEA = .045, MECVI = .220) indicating both implicit instruments provide valid scores.

Conclusions: These results indicate that implicit and explicit attitudes are separate constructs for HSPE students.  Both forms of attitude can be measured during a single physical education period via the web.  Valid scores for measures of both forms of attitude may be a significant step in understanding and ameliorating students’ overall attitudes toward HSPE.