MNLFA as a CritQuant Psychometric Tool

• In Event: Critical Methodological Innovations in Quantitative Research

Abstract

Traditional measurement approaches, such as confirmatory factor analysis, treat the effects of social context and positionality as error rather than an integral facet of phenomena under study. Using mixed methods, the analysis presented here demonstrates the use of moderated nonlinear factor analysis (MNLFA) as a means for incorporating these factors directly into measurement models. This reflects an essential advance for CritQuant, because any approach to quantitative analysis is anchored to the perspectives and practices which generated the data being analyzed. Successful application of this strategy is illustrated using a research self-efficacy scale with Ph.D. students, where results of thematic analysis were used as a moderator within the measurement model, leading to increased invariance in measurement properties over time.

Authors

• David F. Feldon, Utah State University

• Umar Shehzad, Utah State University