A unidimensional short form of the Beck Hopelessness Scale (BHS-7) derived using item response theory

Abstract

The Beck Hopelessness Scale (BHS) is the most widely used measure of hopelessness, a key psychological construct linked with various mental health outcomes. In clinical settings, the BHS has proven a reliable tool for assessing hopelessness; however, there has been debate regarding the tool’s internal consistency among non-clinical populations. Most studies assessing the dimensionality of the BHS have relied on the use of classical test theory (CTT). The length of the BHS has also prompted concerns over its practicality. The BHS-9 was developed to address these critiques and formulated based on psychiatrically hospitalized adult patients. The current study investigates the dimensionality of the BHS-9 among a non-clinical sample using item response theory (Mokken scale analysis and Rasch) and CTT. The results confirm that the BHS-9 is essentially unidimensional. However, a salient finding was that Item 6 violated invariant item ordering. An exploratory factor analysis of the remaining eight items found that the items accounted for 48.05% of the variance. Further exploratory factor analyses, removing one item at a time, showed that the removal of item 18 would increase variance explained > 50%. The revised BHS-7 was found to be unidimensional and maintained strong internal consistency and criterion-related validity. This revised tool effectively captures the essence of hopelessness among a non-clinical population and presents a more refined option for the assessment of this construct.