Evaluating Reasonableness Tests for Longitudinal Measurement Invariance using CFA (307899)Emily Cramer, University of Kansas Medical Center
Mauricio Garnier-Villarreal, University of Marquette
*Elizabeth Grandfield, University of Kansas Medical Center
Keywords: Longitudinal Measurement Invariance, Longitudinal CFA, Simulation
Many health researchers are interested in comparing people across groups, conditions, and/or time. To conclude differences are due to group dynamics or time, the measure(s) used must measure the same underlying construct. Classical statistical methods make this assumption without direct evaluation. Evaluating measurement invariance (MI) prior to generalizing results of a study is a matter of social justice if the study results are to be applied to health care and policy decisions. Confirmatory Factor Analysis (CFA) directly allows MI evaluation. Current MI testing recommendations are based on multiple group studies and simulations. There is a lack of literature on testing MI in longitudinal designs. Current procedures recommend researchers apply the same guidelines from multiple group to longitudinal designs. Longitudinal designs are more complex and require different recommendations. This study evaluated measurement invariance in longitudinal CFA in order to ascertain if the current guidelines are acceptable when applied to the longitudinal framework. Additional considerations and evaluation tools will be discussed.