Confirmatory Factor Analysis of Multitrait-Multimethod Self-concept Data: Between-group and Within-group Invariance...

Area of Science:

• Psychology
• Psychometrics

Background:

• Self-concept is a multidimensional construct.
• Evaluating self-concept accuracy often involves self-other agreement.
• Previous methods for analyzing multitrait-multimethod (MTMM) data have limitations.

Purpose of the Study:

• To extend confirmatory factor analysis (CFA) for analyzing MTMM data.
• To evaluate self-other agreement on multiple dimensions of self-concept.
• To assess the cross-cultural invariance of self-concept measurement.

Main Methods:

• Two studies utilized the Self Description Questionnaire III (SDQIII) with university students in Australia and Canada.
• Confirmatory factor analysis (CFA) and an extension for MTMM data were employed.
• Within-group and between-group invariance tests were conducted.

Main Results:

• Self-other agreement on 13 SDQIII factors was higher than previously reported.
• Both traditional and CFA approaches confirmed convergent and discriminant validity.
• Cross-group invariance tests supported the invariance of CFA-MTMM solutions across Australian and Canadian samples.
• Constrained models were more parsimonious and addressed ill-defined solutions.

Conclusions:

• The extended CFA-MTMM approach is effective for analyzing self-concept data.
• Self-other agreement on self-concept dimensions is robust across cultures.
• This methodology provides a practical solution for improving CFA-MTMM analyses.