Measurement Invariance in Online Game Addiction Scale: Performance of Multi-Group Confirmatory Factor Analysis, Graded Multidimensional IRT, and EFA Tree

• In Event: AERA Roundtable Session Friday 8:00 am Four Seasons Ballroom 4
  In Roundtable Session: Advancing Psychometric Modeling: Innovative Techniques for Educational Assessment (Table 8)

Abstract

This study investigated the performance of three statistical approaches (i.e., MG-CFA, GR-MIRT, and EFA Tree) in estimating the measurement invariance (MI) by using one of the illustrative dataset that adopted the online game addiction (OGA) scale among high school and universities’ students. MG-CFA results showed that OGA was a valid scale to measure four dimensions, establishing three MI levels. The GR-MIRT confirmed the multidimensionality of the OGA, showing several DIFs. The EFA Tree showed no MI when fitting the four-dimension model. However, it showed multiple splits, indicating the violation of MI when fitting the unidimensional model considering four covariates (time spent playing OG, study level, gender, and age). This study expressed many practical applied and methodological dissimilarities among three methods.

Author

• Amal Alhadabi, Sultan Qaboos University