Premium
Measure reliability of the Perceived Competence and Social Acceptance for Children Scale via Rasch analysis
Author(s) -
Heritage Brody,
Mancini Vincent,
Rigoli Daniela,
Piek Jan
Publication year - 2020
Publication title -
british journal of educational psychology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.557
H-Index - 95
eISSN - 2044-8279
pISSN - 0007-0998
DOI - 10.1111/bjep.12268
Subject(s) - rasch model , psychology , competence (human resources) , developmental psychology , psychometrics , construct validity , social competence , item response theory , test validity , cronbach's alpha , item analysis , clinical psychology , social psychology , applied psychology , social change , economics , economic growth
Background The self‐concept of children has an impact on later behavioural development and psychopathology; therefore, evidence of the accurate measurement of self‐concept is important. Harter and Pike's (1984, Child Development , 55 , 1969) commonly used measure of self‐concept, the Pictorial Scale of Perceived Competence and Social Acceptance for Children, has demonstrated varying evidence of its construct validity and reliability, particularly with preschool‐ or kindergarten‐aged participants. Aim To examine the measurement properties of the Acceptance and Competence measures, and to substantiate the measures’ adequacy within the latter participant age range. Sample One hundred and ninety‐seven preschool or kindergarten‐aged children ( M = 5.40 years, SD = 0.30) provided data as part of the Animal Fun study over three measurement periods. Methods The study conducted a pair of Rasch analyses on the Acceptance and Competence measures described prior, with adherence to the assumptions of univariate measurement, item fit, item invariance, and response category adequacy examined as part of these analyses. Results While the Acceptance measure demonstrated adequate reliability outside of some potentially misfitting items, the study identified several limitations for the Competence measure, including potential gender‐based and verbal IQ ‐based response biases. Conclusions Practitioners and research applications of these subscales may benefit from the response recoding recommendations provided in this study for the Acceptance measure, as they improved the measure's properties. The study provides potential solutions to item adequacy concerns, and avenues for future research, involving these measures.