Development and Validation of the East Africa Alcohol Expectancy Scale (AFEXS)

Abstract

ABSTRACT. Objective: The purpose of this study was to develop and validate the East Africa Alcohol Expectancy Scale (AFEXS), a measure of alcohol expectancies for use with HIV-positive adults in East Africa. Method: The study was conducted in several phases. The first involved development of the scale and an exploratory factor analysis (EFA) with a total of 209 HIV-positive, Ugandan men (n = 120) and women (n = 89). The second phase involved a confirmatory factor analysis (CFA) as well as validity analyses with a second, demographically similar, sample of 181 HIV-positive Ugandan men (n = 109) and women (n = 72). All participants were recruited from two ongoing cohort studies taking place at an HIV clinic in rural southwestern Uganda. Results: The EFA resulted in a 13-item, three-factor scale that explained 67% of the variance and showed excellent internal consistency (α = .89). The CFA showed good model fit with a final 11-item scale, χ 2 (41) = 68.60, p = .004; comparative fit index = .97; root mean square error of approximation = .06, 90% CI [.03, .09], that again showed excellent internal consistency and yielded the following three factors: sex-related (α = .94), release of inhibition (α = .61), and negative expectancies (α = .73). Conclusions: The AFEXS is the first psychometrically validated alcohol expectancy scale available for use in Africa. The extent to which the AFEXS can be used with other populations and with adults from countries outside of East Africa requires ongoing validation. (J. Stud. Alcohol Drugs, 76, 336-343, 2015)

Factor structure of the alcohol expectancies questionnaire among adolescents in rural Ghana

“…In its original form, this item may not have been relevant for young people in Ghana, but this requires further investigation. For example, rewording the question to 'I find it harder to say no to sexual advances', as was done in a study in Uganda [51], may be more appropriate and translatable. Indeed, findings from the Ugandan work also support the usefulness of such modifications to the expectancy questionnaire.…”

“…As an exception to this lack of African research in this domain, we note a small‐scale comparative study on drinking motives between Ugandan ( n = 26) and German ( n = 49) treatment seeking adult men, which suggests motives may be similar across these cultures [47]. With existing African research utilising alcohol expectancy measures primarily focussing on sexual expectancies [48–50] or beliefs among those with HIV [51] there have hitherto been no examinations of the full spectrum of alcohol expectancies in the African continent. Peruvian research in another low‐ and middle‐income country (LMIC) found limited evidence that factor loadings established in the West were not fully applicable in the South American context [52, 53].…”