Measuring Spirituality as Personal Belief in Supernatural Forces

Is the Character Strength Inventory-Spirituality subscale a brief, reliable and valid measure?

Abstract

Critical evaluations of commonly used spirituality measures find many wanting— with most lacking the properties required of scientific measures. Common deficiencies include using non-representative development samples, a failure to satisfy normality assumptions, and the confounding of related yet-distinct constructs (e.g. religiosity, wellbeing, civility, prosociality, virtues, etc). The current paper utilizes two studies to examine the psychometric properties of the 7-item Character Strength Inventory-Spirituality (CSI-Spirit; Isaacowitz, Seligman and Valiant 2003). Factor analytic investigations (exploratory and confirmatory) suggest that six items reliably (Cronbach’s a > .70) capture a single latent construct that accounts for around 45% of the variance in responses. This truncated CSI-Spirit appears normally distributed and uni-dimensional. Item difficulty (as reflected by mean scores on items) varies and total scores converge meaningfully with religious affiliation and measures of religiosity, spirituality, paranormal beliefs, wellbeing, agreeableness and conscientiousness. In summary, the CSI-Spirit appears statistically robust and its brevity makes it ideal for individual assessment (in psychological practice) and large scale socioepidemiological research purposes.