The factor structure of the Duke University Religion Index: Continuing the conversation.

Abstract

The purpose of this psychometric study was to continue the conversation among Psychology of Religion and Spirituality researchers regarding the optimal factor structure of the widely used Duke University Religion Index (DUREL). Single-factor, two-factor, and three-factor models of the DUREL are compared in two samples of previously unstudied populations: U.S. community-dwelling adults and U.S. communitydwelling older adults. Special attention is given to adjusting estimation methods for multivariate nonnormality. A stratified random sample of 404 U.S. adults and a convenience sample of 428 communitydwelling older adults completed the DUREL. Based on previous theoretical and empirical work, a series of confirmatory factor analyses were conducted to compare single-factor, two-factor, and three-factor models of the data in both samples. The two-and three-factor models provided the best representation of the underlying structure of the DUREL in both samples. A two-or three-factor measurement model of the DUREL is recommended for use in structural models with U.S. adult and older adult samples. A caveat regarding the use of a three-factor model is discussed.