Validation of the Spanish-language Cardiff Anomalous Perception Scale

Abstract

The Cardiff Anomalous Perceptions Scale (CAPS) is a psychometric measure of hallucinatory experience. It has been widely used in English and used in initial studies in Spanish but a full validation study has not yet been published. We report a validation study of the Spanish-language CAPS, conducted in both Spain and Colombia to cover both European and Latin American Spanish. The Spanish-language version of the CAPS was produced through back translation with slight modifications made for local dialects. In Spain, 329 non-clinical participants completed the CAPS along with 40 patients with psychosis. In Colombia, 190 non-clinical participants completed the CAPS along with 21 patients with psychosis. Participants completed other psychometric scales measuring psychosis-like experience to additionally test convergent and divergent validity. The Spanish-language CAPS was found to have good internal reliability. Test-retest reliability was slightly below the cut-off, although could only be tested in the Spanish non-clinical sample. The scale showed solid construct validity and a principal components analysis broadly replicated previously reported three component factor structures for the CAPS.

Factor structure and age invariance of the Cardiff Anomalous Perceptions Scale (CAPS) in healthy older and younger adults.

“…Overall, the items within this factor capture more severe distortions in everyday perceptions and frank hallucinations, experienced in external space and/or involving other (nonself) agents. Previous factor analyses of the CAPS have identified a clinical psychosis factor—comprising items assessing experiences commonly linked to clinically diagnosable psychosis disorders (Bell et al, 2006; Bell et al, 2011; Debbané et al, 2011; Kao et al, 2013; Tamayo-Agudelo et al, 2019)—that also shares strong similarities with our Factor 2. However, the clinical significance of this factor is unclear, and labeling it in this way also implies that the anomalous self-experiences described in Factor 1 are not clinically significant (which also remains unknown).…”

“…Second, the participants in the current study were predominately female (69.4%) and Anglo-Australian (78.9%), but the potential effects of gender and/or culture on anomalous perceptual experiences could not be examined. Indeed, Tamayo-Agudelo et al (2019) noted that the influence of cultural factors needs to be tested in future research using measurement invariance, and—given the mixed evidence of gender effects on the CAPS (Bell et al, 2006; Bell et al, 2011)—we also propose that measurement invariance should be tested in relation to gender. In the current study, we did not examine the extent to which other sample characteristics that may differ between age groups (e.g., education, IQ) could have influenced the findings, and so we suggest that further study will be needed to systematically assess the impact of these variables.…”

“…For each item endorsed as present, participants rate the level of associated distress, intrusiveness, and frequency on a five-point scale, giving a subscale score range of 0 to 160, with higher scores indicating a greater level of distress, intrusiveness or frequency, respectively. 1 The CAPS total score has previously shown good internal reliability (Cronbach's ␣ ϭ .87; Bell et al, 2006) and convergent validity with similar measures of hallucination proneness (Jaén-Moreno et al, 2014;Tamayo-Agudelo et al, 2019).…”

“…Subsequent studies with predominately younger adults have also found both two- and three-factor structures (Debbané et al, 2011; Jaén-Moreno, Moreno-Díaz, Luque-Luque, & Bell, 2014; Tamayo-Agudelo, Jaén-Moreno, & Luque-Luque, 2015). Although the participants in these studies differed in a number of ways (e.g., gender, culture), one important factor that may also have contributed to the variation in these factor structures is age, leading Tamayo-Agudelo et al (2019) to argue that an analysis of measurement invariance of age is required. Indeed, no studies have specifically examined the factor structure of the CAPS in older adults, and so it is unknown whether this measure is invariant across age groups.…”

“…Importantly, in addition to the varying number of factors found in previous research, there is also evidence of inconsistency in the factor loadings for the indicators of similarly labeled factors (see in the online supplemental material). For example, although a “clinical psychosis” factor has frequently been found in previous research, it has considerable variation in item composition across studies, such as: Items 3, 5, 7, 10, 11, and 31 (Bell et al, 2006); Items 2, 4, 6, 7, 8, 11, 12, 13, 14, 28, 31, and 32 (Bell et al, 2011); Items 2, 3, 4, 8, 11, 12, 13, 22, 24, and 29 (Debbané et al, 2011); Items 3, 7, 11, 13, and 31 (Kao, Wang, Lu, & Liu, 2013); and Items 2, 11, 12, 13, 28, 31, and 32 (Tamayo-Agudelo et al, 2019). Furthermore, there are a number of problems in previous factor analytical studies of the CAPS that prevent researchers from replicating these previous factor structures, such as not being able to confidently determine which items belong in which factor/s (see in the online supplemental material).…”