Can outcome expectancies be measured across substances? Development and validation of a questionnaire for populations in treatment

Abstract

Purpose – The purpose of this paper is to develop a brief outcome expectancies questionnaire applicable across nicotine, alcohol, opioid and stimulant users seeking or willing to seek treatment and to assess its construct and predictive validity. Design/methodology/approach – The items were generated using semi-structured interviews. A cross-sectional study was used to determine the factor structure and internal reliability, to compare the factor structure across the groups and to assess construct validity. Scores were used to predict reduction in dependence at three-month follow-up. Findings – The qualitative study produced 98 items. For the cross-sectional study 99 nicotine, 96 alcohol, 98 opioid and 77 stimulant misusers were recruited. Factor analysis produced a two-factor (positive and negative expectancies) solution, similar across groups. A 28-item common version had scale correlations above 0.94 with the long versions of each group, and high internal consistency (Cronbach's α>0.90). The Positive expectancies sub-scale was positively correlated with urges across all groups, and negatively correlated with self-efficacy in three groups. Negative sub-scale scores were positively correlated with motivation sub-scales and self-efficacy in three groups. Urges and negative expectancies predicted reduction of dependence at three months. Research limitations/implications – The study suggested that outcome expectancies are similar across substance sub-groups. The new tool appears to have good construct and predictive validity. Further validation with larger samples is required. Originality/value – This is the first tool to measure outcome expectancies across substances, facilitating relevant research with poly-substance users.