Using Dynamic Fit Indices to assess structural equation model fit
Assessing model fit is a key element in structural equation modeling. Typically, researchers use statistics such as the CFI, TLI, RMSEA, and SRMR to investigate issues of fit. Values of these statistics are compared with cut-off values to determine whether a model fits the data or not. Research has found that standard cut-values may not be appropriate in all situations. Recently, McNeish and Wolf (2023) proposed a simulation based technique for determining optimal fit index cut-off values for assessing model fit. This post describes this approach and demonstrates how it can be applied to data using an R package developed by McNeish and Wolf.
