Sleep-Related Cannabis Expectancies Questionnaire (SR-CEQ): Replication and Psychometric Validation among College Students using Cannabis for Sleep Aid

Emerging evidence suggests that cannabis is commonly used to aid sleep among college students. Although outcome expectancies have been associated with the progression of cannabis use, sleep-related expectancies have not been included in widely-used cannabis expectancy measures. Recently, the Sleep-Related Cannabis Expectancies Questionnaire (SR-CEQ; Goodhines et al., 2020) was developed and initial evidence for its 2-factor structure was obtained in a general college sample (including non-cannabis users). However, the SR-CEQ’s associations with sleep and cannabis use behaviors among cannabis sleep aid users is unknown. This study aimed to replicate the previous factor structure and test construct and concurrent validity of the SR-CEQ among college students using cannabis for sleep aid. Method: Cross-sectional data were drawn from 94 college students reporting at least bimonthly cannabis use for sleep aid. Five multivariate outliers on the SR-CEQ were excluded, resulting in an analytic sample of 89 (Mage=19.92 [SD=1.19; range=18-22]; 66% female; 72% White, 12% Multiracial, 7% Asian, 5% Black or African-American, 1% self-reported Other and 3% did not disclose; 14% Hispanic/Latinx). Students completed an online survey of sleep and substance use behaviors. A confirmatory factor analysis (CFA) replicated the 2-factor structure (Positive and Negative Sleep-Related Cannabis Expectancies), bivariate correlations tested associations with related constructs (sleep and cannabis use behaviors/beliefs), and independent-samples t-tests further explicated relevant group differences. Results: After dropping item 5 (factor loading .05). Students who used cannabis more frequently in general (≥36 of 60 days, per median split) reported more positive sleep-related cannabis expectancies (t[86]=1.99, p=.05, Cohen’s d=0.42). Negative sleep-related cannabis expectancies (α=.80) were not associated with any cannabis or sleep variables assessed (ps>.05). Negative sleep-related cannabis expectancies were marginally lower among students with greater frequency of general cannabis use (t[87]=-1.89, p=.06, Cohen’s d=0.40) and cannabis use for sleep aid (≥3 times/week, per median split; t[87]=-1.87, p=.06, Cohen’s d=0.40). Further, greater negative sleep-related cannabis expectancies were reported among male (versus female) students (t[87]=2.30, p=.02, Cohen’s d=0.51). Conclusion: Overall, replication of this 2-factor structure showed good fit to the data and both subscales demonstrated good internal consistency. Although replication is needed, results suggest that college students using cannabis for sleep aid may have less negative sleep-related expectancies about sleep. Positive sleep-related cannabis expectancies were associated with dysfunctional beliefs about sleep, but not sleep behaviors or cannabis use. Current novel findings extend existing knowledge of general non-sleep related cannabis expectancies among cannabis users in terms of cannabis use correlates. Findings can help identify at-risk students and modifiable risk factors that can be targeted to minimize harm with cannabis sleep aid use. Future research is needed among larger samples to (a) assess generalizability to varied populations and (b) clarify temporal sequencing of potential consequences through longitudinal designs.