Development of the physical education observation instrument using generalizability study theory.

Abstract

Abstract In this exploratory study, the rater reliability of the Physical Education Observation Instrument (PEOI) was satisfactorily established. Three raters coded the frequency of occurrences of selective teacher (9) and student (4) behaviors and one catchall category when male teachers were instructing male pupils in volleyball. A 3 Rater × 3 Teacher × 2 Occasion Generalizability Study design indicated high stability of both inter- and intra-rater scores on 13 of 14 variables from Ěp 2 .65 to .99, establishing the appropriateness of the PEOI for further observational research on teacher and student behaviors in physical education.