The Australian Sheep-Goat Scale: An evaluation of factor structure and convergent validity

Drinkwater, K, Denovan, A, Dagnall, N and Parker, A (2018) The Australian Sheep-Goat Scale: An evaluation of factor structure and convergent validity. Frontiers in Psychology, 9. pp. 1-14. ISSN 1664-1078

Preview

Published Version Available under License Creative Commons Attribution. Download (3MB) | Preview

Abstract

The Australian Sheep-Goat Scale (ASGS) is a commonly used measure of belief in the paranormal. The scale contains items that index extrasensory perception (ESP), psychokinesis (PK), and life after death (LAD). Although, research employs the ASGS as both a general (unidimensional) and factorial (multidimensional) measure, few studies have examined the appropriateness of these solutions. Accordingly, the present paper tested the psychometric integrity of the ASGS via two studies. Study 1 assessed ASGS factorial structure using confirmatory factor analysis. To achieve this, merging of ASGS data from previously published studies and ongoing work created a heterogeneous sample of 1,601 responses. Analysis revealed that a two-factor bifactor model best explained ASGS organization. This comprised a general overarching factor incorporating two subfactors (ESP and PK). Factor loadings and omega reliability supported a unidimensional structure for the most part. Removal of LAD items improved model fit because the factor added unnecessary complexity and undermined scale psychometric integrity. Study 2, using a supplementary composite sample of 320 respondents, assessed the convergent validity of the emergent ASGS model against a recently published Revised Paranormal Belief Scale (RPBS) bifactor solution. Comparison revealed high convergent validity. The general ASGS factor, despite deriving from only psi-related dimensions (ESP and PK) predicted RPBS scores. This finding indicated that ASGS brevity relative to the RPBS is advantageous when assessing general belief in the paranormal. The ASGS, notwithstanding limited construct content, functions as an effective measure of paranormal belief. Additionally, Study 2 replicated the bifactor structure identified in Study 1 and invariance testing supported invariance of form, factor loadings and item intercepts for this solution across Studies 1 and 2.