Factor structure of the de Jong Gierveld Loneliness Scale: An ESEM approach

Abstract

This research aims to examine the reliability, convergent validity, and measurement invariance of the de Jong Gierveld Loneliness Scale (DJGLS). The study focused especially on the examination of the model-data fit of various competitive factor structures in a young adult sample. The results demonstrate that the bifactor-ESEM model shows a high model-data fit according to CFI and RMSEA. In this case, it has been determined that the cross-loadings defined by the bifactor-ESEM model have an increasing effect on the model-data fit. Also with the bifactor-ESEM model, DJGLS has one highly reliable general factor and two irrelevant subfactors. Metric measurement invariance according to gender was provided. DJGLS scores were correlated moderately and highly, and were statistically significant with external variables. Finally, it can be said that DJGLS is a measurement tool with construct and convergent validity and reliability in the young adult sample. In addition, DJGLS is essentially a uni-dimensional scale and shows the best model-data fit in the bifactor-ESEM model.