Cannabis and Ecstasy/MDMA: Empirical measures of creativity in recreational users

Earlier literature has often linked recreational use of psychoactive drugs with artistic or musical creativity, and some users report taking drugs to enhance creative thinking. Empirical findings have been mixed: laboratory work with cannabis suggested dose-dependent effects on behavioural creativity, while survey work found that recreational Ecstasy/MDMA users sometimes endorse increased creativity as a positive outcome. Theoretical and methodological uncertainty remains about whether and how different drugs might alter creative cognition. Jones and colleagues set out to re-examine the relationship between drug use and creativity by comparing behavioural and subjective measures of creativity in three groups: recent Ecstasy users, cannabis users (who had not used Ecstasy), and nondrug-using controls. The study used an established behavioural test of divergent thinking (the Consequences test), an objective frequency-based scoring method for rarity of responses, and two self-assessment measures; intelligence was also measured to control for possible confounding by fluid ability. No specific directional predictions were made because prior findings were inconsistent.

This was a cross-sectional, between-groups study of 45 undergraduate students at a single UK university, with three groups of 15 participants each: Ecstasy users (11 men, 4 women; mean age 23.00, SD = 2.56), cannabis users (6 men, 9 women; mean age 22.26, SD = 3.05) and nondrug controls (4 men, 11 women; mean age 21.86, SD = 2.13). Recruitment used snowball sampling and flyers. Inclusion criteria included self-reported recent Ecstasy use for the Ecstasy group (use within the previous three months) and no lifetime Ecstasy use for the cannabis and control groups. Participants reported being drug-free at testing and were paid £10; ethics approval and debriefing were provided. The extracted text does not report random sampling or biological verification of abstinence. Behavioural and self-report measures included: (1) the Consequences test of divergent thinking (five hypothetical scenarios, 2 minutes per scenario), scored by two independent, blinded judges according to the standard manual classifications (responses labelled "obvious," "remote," "duplicate" or "irrelevant"); (2) a frequency-based/objective scoring method in which responses across all participants were categorised and those occurring in less than 5% of responses were classed as "rare"; (3) a self-rated creativity item asking participants to compare their Consequences performance with peers on a five-point Likert scale; and (4) Gough's Adjective Creativity Checklist (ADCL), a trait self-report with 18 positive and 12 negative items producing a score range of -12 to +18. Fluid intelligence was assessed using 18 items from Raven's Advanced Progressive Matrices (RAPM), scored as percent correct. Two sets of independent judges performed blind ratings for standard and frequency scoring; agreement statistics were reported. Analyses were primarily ANOVAs comparing the three groups on the various creativity measures, with follow-up Tukey tests for pairwise comparisons. Gender was examined as a potential covariate and was included in some analyses; correlations among creativity measures were also computed with group partialled out. The text reports use of partial eta squared for some effect sizes and t-tests for select group comparisons (for example, cocaine use).

Participant characteristics: the three groups did not differ on RAPM fluid intelligence scores (F(2,44) = 1.124, p = .335) or years of education since age 16 (F(2,44) = 0.274, p = .762). Current alcohol and nicotine use did not differ across groups, but Ecstasy users reported significantly greater lifetime cocaine use than cannabis users (t(28) = 2.941, p < .01). Cannabis and Ecstasy users reported having used cannabis a mean of 3.33 and 3.53 days before testing, respectively; Ecstasy users reported last taking Ecstasy a mean of 7.40 days prior. The study relied on self-reported abstinence. Gender effects: men scored higher than women on the Creativity Adjective Checklist (male M = 6.85, SD = 3.81; female M = 3.87, SD = 4.05; t(43) = 2.530, p < .05) and showed lower fluency on the Consequences test in some scoring variants, but there were no significant gender differences for remote (standard scoring), rare (frequency scoring) or self-rated Consequences performance. Consequences test — standard scoring (judge-rated remote responses and fluency): there were no significant group differences in the number of remote (creative) responses (F(2,45) = 0.600, p = .553, partial eta squared = .028). Fluency (total number of responses) also did not differ between groups (F(2,45) = 1.500, p = .235, partial eta squared = .067); these null results persisted after controlling for gender. Consequences test — frequency scoring (objective rarity): a significant group effect emerged for the total number of rare responses (F(2,44) = 4.440, p < .05). Post hoc Tukey tests indicated that cannabis users produced significantly more rare responses than controls (Tukey p < .05). This difference remained when gender was included as a covariate. Frequency-scored fluency showed no group differences (F(2,44) = 1.556, p = .223). Self-assessment and trait checklist: group differences were observed on the self-rated Consequences performance (F(2,44) = 3.448, p < .05). Follow-up tests showed that Ecstasy users tended to rate their responses as more creative than controls, but this comparison narrowly missed conventional significance in the post hoc test (Tukey p = .058). The group effect on this self-rating became a trend rather than statistically significant when gender was controlled (F(2,44) = 3.160, p = .053). Scores on the ADCL did not differ between groups (F(2,44) = 0.143, p = .867). Inter-rater reliability and correlations: across all five Consequences scenarios judges agreed on 870 of 1,350 responses (64.44% agreement). The two Consequences scoring methods correlated with one another (r = .41, p < .01). Self-rated Consequences performance correlated strongly with judge-rated remote responses (r = .60, p < .001) but did not correlate significantly with frequency-defined rare scores (r = .22).

Jones and colleagues interpret the pattern of findings as indicating a dissociation between objective rarity-based creativity measures and subjective or judge-rated measures. Cannabis users produced significantly more rare responses on the objective frequency scoring of the Consequences test, whereas no group differed on judge-rated remote responses or on fluency. Ecstasy users showed a tendency to perceive their Consequences responses as more creative than controls, despite lacking objective increases on the behavioural measures. The authors note that intelligence (RAPM) and years of post-16 education did not differ between groups and therefore are unlikely to account for the observed effects. The investigators acknowledge several limitations that temper the conclusions. Groups were not fully matched for gender, although the main positive finding (cannabis > controls on rare responses) persisted when gender was statistically controlled. Abstinence before testing relied on self-report and no urinalysis was carried out, so recent substance use cannot be definitively ruled out; the cannabis users reported not having smoked for three days but the possibility of recent use remains. Raven's Matrices assess fluid intelligence but not broader aspects of general intelligence, so the authors recommend using an additional measure such as the National Adult Reading Test in future work. The Consequences test assesses divergent thinking — the production of many novel ideas from a single stimulus — rather than real-world creativity, which may involve problem identification and construction; this distinction limits generalisability. Possible neurochemical mechanisms are discussed cautiously. The authors note prior observations that acute cannabis use can elevate dopaminergic activity, which theoretically could increase arousal and reduce latent inhibition, thereby facilitating generation of uncommon ideas; nevertheless, reported abstinence and the lack of biological verification make it uncertain whether such acute effects explain the current results. Two alternative causal accounts are proposed by the researchers: more creatively inclined individuals might be more likely to use psychoactive substances, or regular cannabis use might enhance aspects of creative thinking. The authors recommend that future research use urine screening to verify abstinence, include multiple intelligence measures, consider cultural or sampling effects (this sample was drawn from a single UK university), and distinguish between divergent thinking and other forms of creativity when investigating drug–creativity relationships.