AUTHOR=Denovan Andrew , Dagnall Neil , Drinkwater Kenneth Graham , Escolà-Gascón Álex TITLE=The Illusory Health Beliefs Scale: preliminary validation using exploratory factor and Rasch analysis JOURNAL=Frontiers in Psychology VOLUME=15 YEAR=2024 URL=https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2024.1408734 DOI=10.3389/fpsyg.2024.1408734 ISSN=1664-1078 ABSTRACT=

Illusory health beliefs are ill-founded, erroneous notions about well-being. They are important as they can influence allied attitudes, actions, and behaviors to the detriment of personal and societal welfare. Noting this, and the prevalence of paranormal beliefs in contemporary Western society, researchers developed the Paranormal Health Beliefs Scale (PHBS). Modification of the PHBS for use with a United Kingdom-based sample resulted in the instrument broadening to incorporate illusory rather than merely paranormal health beliefs. The present study psychometrically assessed the emergent Illusory Health Beliefs Scale (IHBS). The principal objective was to validate the IHBS using a large, representative sample. Eight hundred and fifty participants (360 males, 482 females, eight non-binary) completed the IHBS alongside instruments assessing theoretically associated constructs (i.e., magical thinking, faith in scientifically unsubstantiated notions, and forms of self-referential, intuitive causation). Exploratory factor analysis revealed the existence of six meaningful IHBS dimensions: Religious/Spiritual, Superstition, Precognitive, Health Myths, Skepticism, and Health Pseudoscience. The IHBS demonstrated satisfactory reliability and convergent validity with theoretically aligned constructs. Rasch analysis at the subscale level revealed good item/person fit and item/person reliability, unidimensionality, and equivalency of items across subgroups (gender and religious affiliation). Analysis confirmed the IHBS was an effective measure of illusory health beliefs. However, researchers should undertake further work to refine the scale and evaluate its performance across different samples and time points.