Fitting Ordinal Factor Analysis Models With Missing Data: A Comparison Between Pairwise Deletion and Multiple Imputation

Abstract

This study compares two missing data procedures in the context of ordinal factor analysis models: pairwise deletion (PD; the default setting in Mplus) and multiple imputation (MI). We examine which procedure demonstrates parameter estimates and model fit indices closer to those of complete data. The performance of PD and MI are compared under a wide range of conditions, including number of response categories, sample size, percent of missingness, and degree of model misfit. Results indicate that both PD and MI yield parameter estimates similar to those from analysis of complete data under conditions where the data are missing completely at random (MCAR). When the data are missing at random (MAR), PD parameter estimates are shown to be severely biased across parameter combinations in the study. When the percentage of missingness is less than 50%, MI yields parameter estimates that are similar to results from complete data. However, the fit indices (i.e., χ2, RMSEA, and WRMR) yield estimates that suggested a worse fit than results observed in complete data. We recommend that applied researchers use MI when fitting ordinal factor models with missing data. We further recommend interpreting model fit based on the TLI and CFI incremental fit indices.

Fitting latent growth models with small sample sizes and non-normal missing data

“…That is, if the sum scores were smaller than its s th percentile, missing data were created. This approach used for generating missing data is consistent with that of previous simulation studies (e.g., Shi, Lee, Fairchild, & Maydeu-Olivares, 2019).…”

A New Approach for Missing Data Imputation in Big Data Interface

“…For example, one of them, the FP-algorithm can run in parallel efficiently, which is why it can be used for Big data preprocessing. However, the time complexity of this method has to be improved [25].…”

“…. (25) Transaction T contains the sequence S if S T ⊆ and the objects included in S belong to the set T with the order relation being preserved. It is assumed that other objects between the objects in the sequence S (23) can be interpreted as a sequence of assays by one person at different times (first, the venous pressure was measured, then the pH level was measured, and finally, the hemoglobin level was defined).…”

Rubber Dam Isolation Survey (RDIS) for adhesive restorative treatments

“…32,33 According to Cohen, 31 For the dependent samples t test and the polychoric correlations, which require two-stage bivariate estimations, the missing data were handled using pairwise deletion, an appropriate method given the small amount of missingness and the missing completely at random (MCAR) mechanism observed in the data. 35 In the case of the COM-B path model, the missing data were handled using the superior full information maximum likelihood (FIML) procedure. 36 The dataset for this study is available at: https://osf.io/2vyu9 /view_on-ly=53f43 2750e 57477 0af6b b4410 61403a8.…”