On the quest for validity: Testing the factor structure and measurement invariance of the technology-dimensions in the Technological, Pedagogical, and Content Knowledge (TPACK) model

• Published in: Computers and education 2017-09, Vol.112, p.1-17
• Main Authors: Scherer, Ronny, Tondeur, Jo, Siddiq, Fazilat
• Format: Article
• Language: English
• Subjects: Gender differences; Measurement invariance; Teacher education; Technological, pedagogical, and content knowledge (TPACK); Validity
• ISSN: 0360-1315; 1873-782X
• DOI: 10.1016/j.compedu.2017.04.012

The Technological, Pedagogical, and Content Knowledge (TPACK) framework – a framework which proposes a set of knowledge domains that are essential for effective teaching with technology – has gained considerable attention in the domain of education and technology. With the efforts to conceptualize these knowledge domains comes the question to what extent they can be distinguished empirically. Hence, the present study examines a measure that assesses pre-service teachers' self-efficacy in the technology-related TPACK dimensions (“T-dimensions”). In pursuit of crafting a validity argument, we investigated its factor structure and tested it for measurement invariance across gender and educational tracks, two subgroups that may indicate considerable differences. By means of multi-group confirmatory factor analysis, the data of N = 665 pre-service teachers in 18 teacher training institutions in Flanders (Belgium) revealed a nested factor structure of the TPACK measure, which comprised a general factor and a specific factor of pre-service teachers’ technological knowledge. This factor structure was fully invariant across gender and educational tracks. Mean differences between educational tracks did not occur; yet, substantial differences were found across gender in favor of male pre-service teachers. This study sheds light on critical aspects of crafting a validity argument for the measurement of the T-dimensions in the TPACK framework and reports relevant subgroup differences. •A general TPACK factor and a specific TK can be disentangled.•The T-dimensions in the TPACK are highly correlated (ρ > 0.80).•Measurement invariance across gender and educational track can be established.•Significant gender differences in the T-dimensions exist.•In conclusion, the TK factor stands out among all T-dimensions.