A Rasch Analysis of The Buss-Perry Aggression Questionnaire - Short Form: An evidence from Czech adolescents sample
Abstract
Objectives. The purpose of this paper is to establish the measurement properties of The Short-Form Buss-Perry Aggression Questionnaire using for the first time the Rasch graded scale model of measurement.
Sample and setting. Responses were obtained from a representative sample of 3,159 Czech schoolchildren between 13 and 15 years of age.
The interval covers the critical transition period from childhood to early adolescence. The subjects were asked to fill in the online questionnaire anonymously via the internet during regular classes under the supervision of their teachers.
Results. The study lends support to the 12-item BPAQ-SF unidimensionality which conflicts with earlier works corroborating multidimensional four-factor models using confirmatory factor analysis.
The data reasonably conforms to the Rasch model specifications having acceptable item and person reliabilities, satisfactory item fit indices and item difficulties, and gender and age DIF of no practical significance. The alignment between item difficulty and the disposition of adolescents to endorse an item indicating a noticeable floor effect can be expected, taking into consideration that the scale is intended primarily as a screening instrument.
Study limitation. A limitation of this study is a lack of criterion for the instrument validation, analyzing a questionnaire only in isolation might paint a distorted picture.