Factor Indicator Reliability and Fit Indices: More Evidence Against Golden Rule Cutoffs in Latent Variable Models

• In Event: 61.084 - Criteria and Bias in Measuring Latent Constructs and Classes

Abstract

In applications of structural equation modeling with latent variables, data-model fit assessment is of great importance to the modeling process and model interpretation. The most widely used cut-off criteria for different fit indices, developed in Hu and Bentler (1999), have frequently been treated as “golden rules” for adjudicating adequate model fit in applied studies. Recent literature, however, has questioned the generalizability of these cut-off criteria across different model types and different types of misspecifications than those examined by Hu and Bentler (1999). The current paper adds to these concerns from the perspective of factor indicator reliability, specifically investigating the extent to which fit index cut-offs are affected when factor indicator reliability differs from values used in previous methodological studies.

Authors

• Daniel McNeish, University of Maryland

• Ji An, University of Maryland - College Park

• Gregory R. Hancock, University of Maryland