Development and Validation of an Instrument to Measure Student Attitude Toward Physical Education: A Mixed-Method Approach

The purpose of this study was to develop an instrument to assess middle school students' attitude toward physical education, to provide psychometric evidence of validity and reliability of the instrument, and to demonstrate further evidence of construct validity of the attitude instrument employing a qualitative technique (student interviews). This study was conducted in multiple phases. The preliminary study involved 589 participants in three middle schools. Participants for the first and second elicitation study were 110 and 48 middle school students respectively. The pilot study utilized a convenient sample of 33 participants (grades 4-7) from a summer Sports Fitness Program. Participants for the content validity study were 35 experts in physical education pedagogy. The reliability and validity study involved 995 participants from three middle schools. Twelve participants who were randomly selected based on their observed scores (above 90th and below l0th percentiles) from the reliability and validity study participated in the qualitative investigation (student interviews). The preliminary study did not provide conclusive results. Two primary factors were extracted from the first elicitation study: enjoyment and usefulness. Three primary subfactors (PE teacher, curriculum, and peer) emerged from the second elicitation study. Results from the pilot study were used to revise the attitude instrument. The percent agreement (inter-item) among the experts for the revised attitude instrument was.97. Reliability results from the reliability and validity study demonstrated high internal consistency (G-C alpha above.80) among the items for two primary subfactors (PE teacher and curriculum). The G-C alpha coefficient for the peer subfactor was below.70. Confirmatory factor analysis results indicated that the hypothesized factor structure was not a good fit to the observed data. Results from a follow-up confirmatory factor analysis without the peer subfactor showed that imposing the hypothesized factor structure on the sample data produced an acceptable fit. The fit indexes values of the GFI (.86), AGFI (.82), RMSR (.08) and RMSEA (.08) were within the acceptable range of model fit. Student interview data indicated a clear distinction between the attitudes of the high and low groups for factor 1 (enjoyment) and factor 2 (usefulness). In addition, interview data matched the self-report data from the attitude instrument. Construct validity, therefore, was established both psychometrically and through student interviews for this instrument.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Kinesiology and Community Health

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2015

Author dc:creator

• Subramaniam, Prithwi Raj

Contributors dc:contributor

• Stephen J. Silverman

• Psychology, Psychometrics

Language dc:language

eng

Identifier

(MiAaPQ)AAI9904600

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/86413