Discovery Lab Answers

Neil Dagnall, PhD — Experiments and Data

Neil Dagnall, PhD, is a researcher based in the School of Psychology at Manchester Metropolitan University whose work sits at the intersection of parapsychology and mainstream cognitive psychology. His research program does not primarily test whether psi phenomena occur; instead, it examines the psychological, cognitive, and personality factors that shape paranormal belief, anomalous experience, and related constructs. His most frequent collaborators across the sources retrieved for this question are Kenneth Drinkwater and Andrew Denovan, with additional co-authors including Álex Escolà-Gascón, Danny Powell, Nick Neave, and Claire Murphy-Morgan.

His ESP-Nexus profile is at Neil Dagnall, and his frequent collaborator is profiled at Kenneth Drinkwater, PhD.

Experiments

The sources retrieved for this question document several distinct lines of empirical work.

Cognitive styles and paranormal belief vs. scientific skepticism ^[1]: Dagnall led a 2026 study with Denovan, Murphy-Morgan, Drinkwater, Powell, and Neave examining how cognitive processing styles — specifically evidence-based vs. experiential thinking — differentiate individuals who endorse paranormal belief from those who hold higher belief in science. The study used latent profile analysis to identify subgroups within a sample.

Paranormal belief, conspiracy endorsement, and wellbeing — network analysis ^[3]: With Drinkwater, Denovan, and Escolà-Gascón, Dagnall ran a network analytic study mapping the structural relationships among paranormal belief (PB), conspiracy theory endorsement (CT), and positive wellbeing indicators. A companion study ^[5] extended this to a three-wave longitudinal design examining adaptive function and wellbeing outcomes differentially for PB vs. CT.

Paranormal belief, conspiracy endorsement, and meaning in life ^[6]: Dagnall, Denovan, and Drinkwater conducted a mediation study testing whether creativity and self-esteem sequentially mediate the relationship between paranormal belief / conspiracy endorsement and meaning in life.

Haunted People Syndrome and anomalous experience case study ^[2]: Dagnall co-authored (with O’Keeffe, Massullo, Laythe, Drinkwater, and Houran) a 2025 study providing concurrent validity evidence for the Haunted People Syndrome (HP-S) model via independent case-study analysis, applying the AECKO framework to a reported haunting case from the existing literature.

Illusory Health Beliefs Scale validation ^[4]: As a co-author with Denovan and Drinkwater, Dagnall contributed to the development and psychometric validation of a new scale measuring illusory health beliefs, using exploratory structural equation modeling (ESEM) and multidimensional Rasch analysis.

Methodology

Across these studies, several consistent methodological features appear.

Latent profile analysis (LPA): The 2026 cognitive styles study ^[1] used LPA to identify discrete subgroups rather than treating belief as a continuous variable. Model selection compared 1-, 2-, and 3-profile solutions using AIC, BIC, ssaBIC, and entropy criteria, with LMR-A-LRT significance testing for incremental fit. The 2-profile solution was retained on the basis of superior entropy and a non-significant improvement from 2 to 3 profiles.

Network analysis: The 2025 paranormal belief / conspiracy / wellbeing study ^[3] used network analytic methods — an approach Dagnall’s group has applied in multiple papers (a 2022 predecessor network analysis paper is cited in ^[3]) — to map the pattern and strength of associations among constructs without assuming a single latent variable structure. This allows identification of which variables act as central “bridge” nodes and which are peripheral.

Longitudinal structural equation modeling: The three-wave study ^[5] used SEM with bootstrapped indirect effects (1,000 resamples, 95% bias-corrected CIs) and applied the Satorra-Bentler chi-square test for nested model comparison. Data collection was online via Qualtrics.

Mediation modeling: The meaning-in-life study ^[6] specified and compared saturated and constrained mediation models, testing creativity and self-esteem as sequential mediators between belief constructs and meaning-in-life outcomes. Model fit criteria used were CFI > 0.95, SRMR < 0.05, and RMSEA < 0.05.

Psychometric validation: The IHBS validation study ^[4] combined ESEM and multidimensional Rasch analysis, with Harman single-factor test used to assess common method bias (single factor accounted for 20.98% of variance, below the 50% threshold). The studies consistently use established, psychometrically attested scales — the Revised Paranormal Belief Scale (RPBS), the GCB-5 for conspiracy endorsement, the SPQ-B for schizotypy, and wellbeing instruments including the MLQ, RES, and SWLS.

Case-study application ^[2]: The HP-S study applied an existing theoretical model (Haunted People Syndrome) to an independently published case study, using correlational analysis of narrative frequency distributions and Rasch-scaled scoring to assess convergent validity.

Data

Because no structured evidence block was provided for this question, the figures below come directly from the retrieved primary sources and are presented in table form where the surrounding text is sufficiently intact to make attribution unambiguous.

Cognitive styles / latent profile analysis ^[1]:

The 2-profile solution separated a “Higher Evidence-based Thinking” group (approximately 55%, n = 165) from a “Lower Evidence-based Thinking” group. The MANOVA comparing the two profiles on cognitive and belief variables yielded a large effect size (η² = 0.37). Significant between-profile differences were found for Reality Testing, Systematizing, Empathizing, Rational, and Experiential processing. The LMR-A-LRT for the 3- vs. 2-profile comparison was non-significant (p = 0.178), supporting retention of the 2-profile model.

Paranormal belief, conspiracy endorsement, and wellbeing ^{[3] [5] [6]}:

The network analysis ^[3] found that paranormal belief and conspiracy endorsement, while positively correlated, showed opposing patterns of association with positive wellbeing. Paranormal belief was more strongly associated with affirming ideations (hope, universal meaning, connectedness), while conspiracy endorsement was characterized by negative, antagonistic cognitions. Paranormal belief and conspiracy theory endorsement shared approximately 40% variance ^[6], indicating meaningful overlap but substantial independence. The RPBS used across these studies reported ω = 0.94 ^[3].

In the three-wave longitudinal study ^[5], bootstrapped mediation paths (1,000 resamples) found: conspiracy endorsement predicted avoidant coping (indirect effect coefficient 0.01, p < 0.05); paranormal belief showed divergence from conspiracy endorsement on positive attributes including Presence of Meaning, Active Coping, Optimism, and Social Identity. Final model fit: χ²(3, N = 1,158) = 2.84, p = 0.416, CFI = 1.0, SRMR = 0.01, RMSEA = 0.01.

In the meaning-in-life mediation study ^[6], the constrained model (creativity and self-esteem as sequential mediators) fit well: χ²(2, N = 647) = 3.94, p = 0.139, CFI = 0.99, SRMR = 0.01, RMSEA = 0.03.

IHBS validation ^[4]: The six-factor ESEM model showed good fit: χ²(940) = 4725.02, p < 0.001, CFI = 0.98, RMSEA = 0.04 (90% CI: 0.04–0.05), SRMR = 0.02. A refined solution also showed good fit: χ²(897) = 4387.73, p < 0.001, CFI = 0.98, RMSEA = 0.04, SRMR = 0.02. High positive correlations between Precognitive, Superstition, and Religious/Spiritual subscales confirmed convergent validity with the RPBS.

Haunted People Syndrome case study ^[2]: Correlational analysis of narrative frequency distributions found that the case most closely resembled a “Primed” haunt context (r = .43, p < .01); other narrative types produced non-significant associations. A previously reported finding cited in ^[2] recorded a significant decline in daily disturbances over time in the source case (r = −0.397, p = .001, two-tailed).

Skeptical critiques

The sources retrieved for this question do not include published criticism directed specifically at Dagnall’s methods or findings — no named critic, reanalysis, or failed independent replication appears in the retrieved material. One methodological limitation the authors themselves register across several studies ^[3][5][6] is the use of cross-sectional or online self-report designs, which constrain causal inference; the three-wave design ^[5] was partly a response to this. The HP-S study ^[2] is explicitly framed as a concurrent validity test against a single independently published case, and the authors note that most non-significant correlational associations in the haunt narrative analysis (four of five narrative-type comparisons) did not reach significance, limiting conclusions about narrative classification.