Childhood Trauma and the Emergence of Precognitive Abilities: A Correlational Study

Cameron, K., & Wade, J. (2024). Childhood Trauma and the Emergence of Precognitive Abilities: A Correlational Study. International Journal of Transpersonal Studies, 43(1–2), 60–93. https://doi.org/10.24972/ijts.2024.43.1-2.60

AI Assessment

A correlational study testing the idea that childhood trauma might foster precognition, on the theory that anticipating danger would be adaptive for a child in an unpredictable, threatening environment. With 227 adults completing a standard childhood-trauma questionnaire and four online ESP tasks, the severely traumatized group outscored the least-traumatized group on just one of the four tasks, long remote viewing (p = .039, a small effect, d = 0.36), while the other three tasks showed no differences. The finding is genuinely weak and mixed: a single marginally significant result out of four tasks that would not survive correction for multiple comparisons, a self-selected sample recruited from trauma-survivor and psi-interested communities, and post-hoc sub-analyses that were incoherent across abuse types (neglect associated positively, sexual abuse reversed). To the authors’ credit, the null tasks and the reversed pattern are reported plainly. This audit reports what the study found and how; it takes no position on whether precognition is real.

Provenance

DOI. 10.24972/ijts.2024.43.1-2.60 · International Journal of Transpersonal Studies 2024, 43(1–2), 60 to 93. Open access under CC BY-NC-ND.

Study type. A cross-sectional correlational study relating self-reported childhood trauma to performance on ESP tasks. It is observational, not experimental.

Authors. Kirsten Cameron (private practice) and Jenny Wade (California Institute of Integral Studies).

Data availability. Not stated in the article.

Source basis. Every figure below is taken from the article’s own Abstract and Results.

What the paper reports

Survivors of childhood abuse and neglect report higher rates of paranormal belief, which is usually attributed to cognitive error. This study asks whether that higher belief might instead partly reflect real precognitive experience, hypothesizing that dissociative tendencies and a need to anticipate danger could make precognition an adaptive by-product of a traumatic childhood.¹ It therefore correlates self-reported childhood trauma with performance on four ESP tasks, comparing the most-traumatized participants against the least-traumatized. The result is a single significant difference on one task, which the authors present as tentative support requiring further study.

The severely abused/neglected group performed significantly better on one precognitive task using a protocol for remote viewing.

How it was run

• Sample. 227 adults recruited via social media and online newsletters aimed at trauma-survivor and psi-interested communities, a self-selected sample.
• Trauma measure. The Childhood Trauma Questionnaire, with subscales for emotional, physical, and sexual abuse and physical and emotional neglect, summed into an overall trauma score; participants were also split into severely-traumatized (ST, n = 112) and least-traumatized (LT, n = 51) groups.
• ESP tasks (IONS Discovery Lab). Four online tasks: a jar task (intuitive guessing), a location task (spatial precognition, combined by the Stouffer method), a short remote-viewing task (forced choice, 1 of 5, 20 trials, chance 20%), and a long remote-viewing task (free-response descriptors, 2 trials, scored 0 to 100 against a 218-item target pool).
• Analysis. Pearson correlations between trauma score and task performance for the full sample, t-tests comparing the ST and LT groups on each task, and post-hoc analyses by trauma subscale and by high-versus-low long-remote-viewing scorers.

Results, as reported

Values are reproduced from the article’s Results. The evidence for a trauma-precognition link rests on one of four tasks, at a p just under .05 with a small effect, and the trauma subtypes do not point consistently in the same direction.

Eleven-dimension audit

Pre-registration

Not preregistered. With four tasks, both correlational and group analyses, and several post-hoc breakdowns, the absence of a registered plan makes it hard to know how the single significant result should be weighted against the many tested comparisons.

Randomization

Not applicable in the assignment sense: trauma is measured, not manipulated, and the sample is self-selected. The ESP tasks themselves use randomized targets.

Sensory leakage

The ESP tasks are standardized online protocols with forced-choice and free-response formats against defined target pools, which controls the psi task itself. The residual concern is the unsupervised online setting and, more importantly, the self-report and self-selection issues below.

Blinding

The tasks are automated and the long-remote-viewing responses are scored against a target pool, limiting scorer bias on the ESP side. Trauma is entirely self-reported via questionnaire, and the participants were drawn from communities primed to both topics.

Optional stopping

Not described, and without preregistration not documented; the analytic latitude lies mainly in the number of tasks and post-hoc analyses examined.

Outcome measure

Four distinct ESP outcomes plus a validated trauma instrument. Using multiple tasks is reasonable, but it also means the headline is one selected task (long remote viewing) among four, and the free-response scoring of that task is the least objective of the set.

Effect size

Small: the one significant group difference is d = 0.36 at p = .039, and the post-hoc correlations are small (r = .12 to .16). These are modest associations, and three of the four tasks show nothing.

Multiple comparisons

This is the central concern. Four tasks were tested by both correlation and group comparison, followed by subscale and high-versus-low post-hoc analyses. A single result at p = .039 would not survive a correction for the four primary tasks, so it should be read as suggestive at best rather than as an established effect.

Internal replication

The study is framed as a partial replication of earlier trauma-psi work (Rabeyron & Watt, 2010), and the four tasks provide an internal consistency check that the effect largely failed: only one task showed a difference, and the trauma subtypes disagreed.

External replication

Partial replication of prior work, but on a self-selected sample and with a single significant task; independent, preregistered replication with a general-population sample would be needed to take the association seriously.

Transparency

Reasonable in reporting: the three null tasks are reported, and so is the awkward finding that sexual abuse ran opposite to the neglect subscales. The limits are the lack of preregistration, an unstated data-availability provision, and the self-selected recruitment.

The adversarial record

• One task out of four, barely significant. The entire positive claim rests on a single task at p = .039 with a small effect, while the other three ESP tasks showed nothing. Correcting for the four primary comparisons would very likely erase the result, so it is weak evidence on its own terms.
• Self-selected, doubly primed sample. Recruiting from trauma-survivor and psi-interested online communities produces a sample motivated toward both the exposure and the outcome, which raises the possibility of demand and reporting effects rather than a genuine trauma-precognition link.
• Incoherent trauma subtypes. The post-hoc pattern is not consistent: physical and emotional neglect were positively associated with the significant task, but sexual abuse ran in the opposite direction. A real causal mechanism would not be expected to reverse across abuse types.
• Correlational, but framed causally. The theory (trauma fosters precognition as an adaptation) is causal, while the design is cross-sectional and correlational and cannot separate ability from belief, experience, or self-selection.
• What is genuinely creditable. The study uses a validated trauma instrument and standardized ESP tasks, tests multiple outcomes rather than cherry-picking one, replicates the structure of prior work, and reports its null results and its self-contradicting subscale pattern honestly. As a correlational probe it is transparently reported; its own numbers make clear that the trauma-precognition association, if any, is small, task-specific, and unconfirmed.