Development of the Psychological Insight Questionnaire among a sample of people who have consumed psilocybin or LSD

Earlier research has produced several broad instruments to quantify subjective psychedelic effects (for example, the States of Consciousness Questionnaire, Hallucinogen Rating Scale, and Altered States of Consciousness Rating Scale), and shorter targeted measures have since been derived to capture specific experiential domains such as mystical-type experiences (Mystical Experiences Questionnaire, MEQ) and challenging experiences (Challenging Experience Questionnaire, CEQ). However, the authors note that no psychometrically validated, multi-item measure specifically targets acute psychologically insightful experiences—realizations or discoveries about one’s personality, relationships, patterns of behaviour or emotions—that can occur during a discrete psychedelic episode. Existing single- or few-item constructs of insight (for instance, the “insightfulness” sub-factor within Oceanic Boundlessness or single items in persisting-effects questionnaires) and measures of emotional breakthrough capture parts of this domain but do not assess its breadth. Davis and colleagues designed this study to develop and psychometrically evaluate a brief self-report instrument, the Psychological Insight Questionnaire (PIQ), intended to measure acute insight experiences occasioned by classic psychedelics. The primary aims were to generate candidate items, test the instrument’s factor structure and internal reliability, and evaluate construct, predictive and incremental validity against established measures of mystical and challenging experiences as well as retrospective reports of changes in psychological flexibility and well-being. The authors positioned the PIQ as a tool for future clinical and laboratory research on mechanisms of psychedelic-related change.

Item development began with consultation among psilocybin session facilitators and iterative revision by members of the study team, yielding an initial pool of 38 candidate items that was reduced to a 28-item draft after removing redundant or unclear items. Respondents rated the intensity with which they experienced each candidate insight item during a single most-insightful psychedelic experience on a six-point scale from 0 ("No; not at all") to 5 ("Extremely (more than ever before in my life)"). The final PIQ and scoring are reported as provided in the supplementary material (not included here). Data were collected via an anonymous internet-based cross-sectional survey conducted June 2018–February 2019. Inclusion criteria were age 18+; fluent English; having taken a single psychedelic that produced moderate-to-strong psychoactive effects; and attributing a gained psychological insight to that experience. Recruitment used online postings, e-mail announcements and an article on Motherboard/VICE. Of 15,470 people who clicked the advertisement, 3177 began the survey; after exclusions for polysubstance use (n=996), use of psychedelics other than psilocybin or LSD (n=398), and inconsistent responses (n=122), the final analytic sample was 1661. The sample was predominantly young (mean age 22.9, SD=8.5), White (83%), male (72%) and US residents (66%). The study received institutional review board approval; no incentives were offered. Alongside the PIQ, respondents completed established measures to evaluate construct and predictive validity: the 30-item MEQ (mean score used; alpha=0.95), the 26-item CEQ (mean score used; alpha=0.95), the five-item Session Impacts Scale (SIS) insight subscale (0–4 scale; alpha=0.84), the Acceptance and Action Questionnaire II (AAQ-II) administered twice to derive retrospective pre–post change in psychological flexibility (reversed so higher = greater flexibility; alphas 0.91 and 0.88), and the Satisfaction With Life Scale (SWLS) administered twice to derive retrospective change in well-being (alphas 0.86 and 0.82). Change scores were computed as post minus pre means for AAQ-II and SWLS and used in subsequent analyses. For scale development, the 1661 respondents were stratified into matched exploratory and confirmatory subsamples (after dropping 250 observations to ensure adequate cell sizes), yielding 1411 responses for factor analyses (exploratory n=652; confirmatory n=759). The exploratory stratum underwent principal components analysis (PCA) with varimax rotation; parallel analysis and scree inspection guided factor extraction. Items with loadings <0.4 or cross-loadings ≥0.4 were removed iteratively to achieve a simple structure. Internal consistency was assessed with Cronbach’s alpha, corrected item-total correlations and average inter-item correlations. The resulting factor structure was tested in the confirmatory stratum via confirmatory factor analysis (CFA) with standard fit indices (RMSEA, SRMR, CFI). Construct validity was examined with Pearson correlations and linearity tests. Incremental validity was assessed using stepwise linear regression models predicting retrospective changes in psychological flexibility and well-being, first entering MEQ and CEQ scores, then adding PIQ subscale scores to evaluate added explanatory variance.

Scale structure and reliability: Principal component analysis in the exploratory subsample indicated two components, after removal of items that failed loading criteria, resulting in a final PIQ with two subscales labelled Avoidance and Maladaptive Patterns Insights (AMP) and Goals and Adaptive Patterns Insights (GAP). Parallel analysis supported a single component within each subscale. Internal consistency was high: AMP Cronbach’s alpha = 0.93 (95% CI 0.92–0.94) and GAP alpha = 0.85 (95% CI 0.84–0.87). Corrected item–total correlations ranged approximately 0.48–0.78 across subscales and average inter-item correlations were r=0.48 (AMP) and r=0.37 (GAP). No remaining item removal materially increased alpha, so all retained items were kept. In the confirmatory subsample, CFA yielded acceptable fit (RMSEA = 0.052, 90% CI 0.045–0.058; SRMR = 0.027; CFI = 0.973). The AMP and GAP subscales were strongly correlated with one another (r = 0.62, p < 0.001) and with the PIQ total score. Construct validity: Correlational analyses in the full sample (n=1661) showed weak-to-moderate associations between PIQ metrics and mystical-type and challenging experiences. Specifically, PIQ mean score correlations with MEQ were reported as moderate (PIQ r=0.38, p<0.001; AMP r=0.27, p<0.001; GAP r=0.50, p<0.001). Correlations with the CEQ were generally weak (PIQ r=0.14, p<0.001; AMP r=0.21, p<0.001; GAP r = -0.02, not significant), suggesting the PIQ taps a domain distinct from challenging experiences. The PIQ total and subscales also correlated strongly with the SIS insight subscale; the extracted text reports a strong correlation but the precise correlation coefficient for the PIQ–SIS relation is not clearly reported in the provided extraction. Predictive and incremental validity: Descriptive statistics reported for the PIQ and outcome change scores were: PIQ mean = 3.08 (SD 1.03), AMP mean = 2.90 (SD 1.22), GAP mean = 3.36 (SD 1.01); AAQ-II change mean = 1.55 (SD 1.46); SWLS change mean = 1.15 (SD 1.05). The authors report a strong correlation between PIQ (and both subscales) and retrospective increases in psychological flexibility, and a moderate correlation between PIQ (and subscales) and increases in well-being/life satisfaction, though exact correlation coefficients for these relations are not fully reported in the extract. Stepwise regression analyses assessed whether PIQ scores added predictive value beyond MEQ and CEQ. For change in psychological flexibility, adding AMP and GAP substantially increased explained variance: in the final model MEQ β = 0.09 (p < 0.001), CEQ β = -0.08 (p < 0.001), AMP β = 0.52 (p < 0.001), and GAP β = 0.04 (p = 0.147); the F change for adding the two PIQ subscales was F(2,1656) = 316.78, p < 0.001, and adjusted R2 for the model including PIQ subscales was reported as 0.31. For change in well-being and life satisfaction, the final model coefficients included MEQ β = 0.17 (p < 0.001), CEQ β = -0.04 (p = 0.104), AMP β = 0.27 (p < 0.001) and GAP β = 0.10 (p = 0.001); the F change for adding PIQ subscales was F(2,1656) = 104.55, p < 0.001. A portion of the extracted regression reporting (for example some adjusted R2 values and earlier-step model betas) is truncated in the provided text and not clearly recoverable from the extraction.

Davis and colleagues interpret their findings as supporting the PIQ as a brief, reliable measure of acute psychological insight following psilocybin or LSD experiences, with two coherent subscales that capture insight about avoidance/maladaptive patterns (AMP) and about goals/adaptive patterns (GAP). Both subscales and the total PIQ demonstrated high internal consistency, and the two-factor structure held up in confirmatory factor analysis with acceptable fit indices. Construct validity was evidenced by strong associations with a general session-insight measure (SIS) and by distinct, weaker associations with measures of mystical-type and challenging experiences, indicating that the PIQ measures aspects of the psychedelic experience not fully captured by MEQ or CEQ. The authors argue that the PIQ showed predictive promise: PIQ scores were associated with retrospective reports of increased psychological flexibility and improved life satisfaction, and they accounted for unique variance in those outcomes beyond MEQ and CEQ scores. AMP in particular showed a large unique association with changes in psychological flexibility, while both AMP and GAP contributed to predicting well-being. From this, the study team proposes that acute psychological insight may be a distinct experiential mechanism that can contribute to therapeutic change after psychedelic experiences. The discussion situates these results within the broader psychotherapy and psychedelic literature, noting concordance with prior findings linking insight to positive outcomes across different interventions but also differences in effect sizes compared to some earlier reports; the authors caution that differences in measurement, sample characteristics and substances studied may account for discrepancies. They also acknowledge that related constructs—such as the Emotional Breakthrough Inventory—may overlap with the PIQ but were not compared here because of timing and reporting differences. Important limitations are acknowledged. The cross-sectional, retrospective web-survey design may introduce recall bias, selection bias due to targeted recruitment of people who recalled insight experiences, and demographic skew (primarily young, White men), limiting generalisability. The study did not assess clinical history or contextual variables (set and setting) and lacked independent content validation of items. The authors therefore emphasise the need for prospective, longitudinal, and clinical-laboratory research with more diverse samples to confirm predictive validity, to examine how reported insight varies with time since the experience, and to compare the PIQ with other emerging measures. They suggest the PIQ may also be useful for assessing acute insight in traditional psychotherapy contexts and for advancing mechanistic research in psychedelic-assisted interventions.