Radin’s anomalous cognition program spans multiple paradigms: (1) mediumship, blinded accuracy testing and electrocortical recording during claimed communication with deceased individuals ¹²; (2) channeling phenomenology, qualitative and quantitative characterization of full-trance channelers ⁴; (3) telephone telepathy, forced-choice identification of callers ⁵; (4) intuitive mortality assessment from facial photographs ¹⁰; and (5) large-scale online precognition experiments ⁷. The breadth reflects a strategy of triangulating across independent paradigms rather than relying on any single experimental design. Most studies are explicitly labeled exploratory by the authors, reflecting the early-stage nature of the research program.

Delorme, Beischel, Michel, Boccuzzi, Radin, and Mills recorded electrocortical activity in mediums and controls during a structured image-classification task designed to probe claimed communication with deceased individuals. The study examined whether mediums’ performance accuracy differed from controls and whether EEG patterns during claimed communication differed from baseline. Results indicated accuracy differences between groups and distinctive electrocortical patterns in mediums during the task condition, though the authors characterized the findings as exploratory and noted that sample sizes were small. The specific artifact addressed was sensory leakage: images were presented in a blinded, computer-controlled sequence, and raters were unaware of participant group membership during scoring. The strength of ruling-out for sensory leakage was mitigated but not eliminated, as experimenter-participant interaction during setup remained a partially unresolved confound.¹²

Wahbeh, Carpenter, and Radin invited five full-trance channelers to a structured pilot study. Prior to the event, participants completed measures of demographics, spirituality, religious background, medication use, personality, and dissociation. During channeling sessions, physiological and behavioral data were recorded. The study was explicitly phenomenological and exploratory in design, with N=5 precluding inferential generalization. The primary artifact addressed was demand characteristics: participants were not told what outcomes were expected, and measures were standardized. The strength of ruling-out for demand characteristics was partially addressed. The researchers noted that full-trance channeling had received little systematic scientific attention and that this study was intended to establish a conceptual framework for future work.⁴

Wahbeh and Radin compared dissociation symptom scores (using a standardized dissociation inventory) between individuals who self-reported mediumship experiences and those who did not. The mediumship group scored significantly higher on dissociation measures, but mean scores remained below the clinical threshold for pathological dissociation disorder. The study was published as a research note in F1000Research and underwent open peer review, with two reviewers approving and one not approving. The primary artifact addressed was sampling bias: the study used self-selected volunteers, and the authors acknowledged that this limits generalizability. The proposed mechanism, that mediumship involves a non-pathological dissociative capacity, remains the researchers’ preferred interpretation; alternative interpretations include that dissociation-prone individuals are more likely to interpret ordinary experiences as mediumistic.³¹¹

Wahbeh, Cannard, Radin, and Delorme conducted a telephone telepathy study in which participants attempted to identify which of several known callers was on the line before answering. The primary artifact addressed was sensory leakage: caller identity was randomized and concealed from both participant and experimenter during the trial, eliminating direct cueing. Response bias, the tendency to guess certain callers more frequently regardless of who was actually calling, was assessed statistically. The companion dataset was made publicly available via Figshare to allow independent reanalysis, partially addressing the file-drawer artifact by enabling scrutiny of the raw data.⁵¹³ The proposed mechanism, that participants accessed information about the caller’s identity through a non-sensory channel, remains the researchers’ preferred interpretation; the alternative interpretation is that above-chance accuracy, if present, reflects subtle sensory cues not fully controlled by the blinding procedure.

Radin’s “Tricking the Trickster” analysis of two online psi experiments (combined N ≈ 200,000 participants, 114 million trials, five-target forced-choice protocol) found that the overall hit rate was null, consistent with chance expectation. However, a planned secondary analysis examining sequential structure in the data found that the probability of a correct response following a correct response (p1) deviated from the chance-expected rate (p0 = 0.32) by a small but statistically unambiguous margin: combined observed p1 = 0.320502 ± 0.000044, z = 11.28, p = 1.7 × 10⁻²⁹. Control tests found no evidence that this sequential deviation arose from artifacts in the random number generator or from optional stopping. The primary artifact addressed was multiple comparisons: the sequential analysis was described as planned rather than post-hoc, though the absence of a public preregistration record means this claim cannot be independently verified, leaving the analysis provenance partially unresolved. The proposed mechanism, that psi effects manifest as sequential dependencies rather than overall hit-rate elevations, is the researcher’s preferred interpretation; the alternative is that the sequential pattern reflects an undetected artifact in the online data-collection pipeline.⁷

Delorme, Pierce, Michel, and Radin invited 12 individuals who claimed to intuitively gauge mortality from photographs to participate in a structured test. All photographs were converted to uniform grayscale and counterbalanced across eight visual categories to eliminate confounding cues that might allow ordinary inference (e.g., apparent age, health markers visible in color images). Behavioral accuracy, electrocortical activity (EEG), and machine learning classification of neural patterns were analyzed. The study was explicitly exploratory with N=12, precluding strong inferential conclusions. The primary artifact addressed was judging contamination: the grayscale transformation and counterbalancing were designed to eliminate ordinary visual cues that could support above-chance accuracy without anomalous cognition. The strength of ruling-out was mitigated but not eliminated, as subtle residual cues in standardized photographs cannot be fully excluded.¹⁰ Note: a related Frontiers in Human Neuroscience paper by the same group was subsequently retracted ¹⁴; the retracted paper’s evidence yields no conclusion in either direction because of the documented retraction. The EXPLORE study ¹⁰ is a separate publication and is not affected by that retraction.

Wahbeh, Delorme, and Radin adapted a method used in medical evidence synthesis to rate the persuasiveness of different categories of survival-relevant evidence. Raters evaluated controlled, prospective experiments as well as case studies across categories including near-death experiences with veridical out-of-body perceptions, mediumship studies, reincarnation research, and apparition cases. The most frequently selected controlled, prospective experiment rated as potentially persuasive was veridical out-of-body perception during a near-death experience. Mediumship and reincarnation studies were also rated as potentially persuasive by a subset of raters. The exercise was explicitly not a meta-analysis and did not produce quantitative effect-size estimates; it was a structured expert-opinion exercise. The primary limitation acknowledged by the authors was that rater selection and rating criteria introduce subjectivity that limits generalizability of the conclusions.⁶ A related essay submitted to the Bigelow Institute for Consciousness Studies contest reviewed the broader evidence landscape for survival, concluding that some evidence categories are stronger than skeptics typically acknowledge while stopping short of claiming proof.¹⁵

Braude, Barušs, Delorme, Radin, and Wahbeh published a commentary responding to Keith Augustine’s critical evaluation of the Bigelow Institute for Consciousness Studies essay contest on survival evidence. The response argued that Augustine’s critique rehearsed familiar objections without adequately engaging with the strongest evidence categories, and that methodological problems in survival-type studies do not uniformly undermine all evidence. The commentary did not present new empirical data; it was a theoretical and methodological argument. The primary claim, that some survival evidence is stronger than skeptics conclude, is the researchers’ preferred interpretation and remains contested in the literature.¹⁶