Overview

Edwin May’s work on random number generators spans from the mid-1980s through the present, establishing him as a central figure in the quantitative study of anomalous mental phenomena. His research trajectory reflects a deliberate move away from force-based explanations toward an informational and decision-theoretic framework. This shift was not arbitrary but emerged from careful examination of experimental data that consistently showed small, replicable effects that resisted interpretation as direct physical influence.

May’s approach to RNG research combined rigorous experimental design with theoretical innovation. Rather than accepting the null hypothesis or invoking ad hoc mechanisms, he developed a formal model that could account for the observed patterns while remaining consistent with quantum mechanics and information theory. Decision Augmentation Theory became the intellectual centerpiece of this effort, offering a parsimonious explanation for why consciousness appeared to influence random processes in laboratory settings.

RNG research foundations

The study of psychokinetic effects on random number generators began in earnest during the 1980s, with May contributing to early collaborative efforts that established the empirical baseline for the field. In 1985, May and colleagues published an informational model of psi that reframed RNG-PK experiments in terms of information transfer and decision processes.¹ This model departed from earlier conceptualizations that treated psi as a force analogous to gravity or electromagnetism.

The informational approach recognized that RNG outputs constitute a probability distribution, and that anomalous effects might operate by biasing selection within that distribution rather than by altering the physical properties of the generator itself. This distinction proved crucial for subsequent theoretical development. May and Radin collaborated on methodological refinements, proposing systematic approaches to testing intuitive data sorting models with pseudorandom number generators.⁶

Meta-analytic work conducted in the mid-1980s synthesized results across multiple RNG-PK studies, revealing effect sizes that were statistically significant but small in magnitude.³ These findings posed an interpretive challenge: the effects were real enough to warrant theoretical attention but too modest to support claims of direct physical causation. This empirical pattern became the primary motivation for developing Decision Augmentation Theory.

Decision Augmentation Theory

Decision Augmentation Theory emerged as May’s response to the need for a coherent theoretical framework that could explain RNG-PK effects without invoking implausible mechanisms. Published in its foundational form in 1995, DAT proposed that anomalous mental phenomena operate through goal-directed selection of outcomes from probability distributions.² The theory treats psi not as a force but as a form of information access that allows the mind to select preferentially among quantum-level outcomes.

The core insight of DAT is that consciousness does not need to “push” on physical systems; instead, it can “choose” among the outcomes that quantum mechanics permits. When a random number generator produces a sequence of bits, each bit represents a quantum event with inherent indeterminacy. DAT proposes that psi operates by biasing which of the quantum-mechanically permitted outcomes actually manifests. This reframing preserves the integrity of physical law while allowing for anomalous mental influence.

May formalized this intuition mathematically, developing equations that describe how decision augmentation would appear in experimental data.⁴ The theory makes specific predictions about effect sizes, their relationship to task difficulty, and how they should scale with information content. These predictions proved testable and, in many cases, consistent with empirical observations from RNG-PK studies.

A key feature of DAT is its applicability beyond RNG research. The theory extends naturally to precognition and presentiment, treating these phenomena as instances of goal-directed selection operating across time. Rather than requiring separate mechanisms for different psi phenomena, DAT provides a unified framework in which all anomalous mental effects reflect the same underlying process: selection of outcomes that serve the organism’s goals or interests.

Theoretical implications

Decision Augmentation Theory carries profound implications for how parapsychologists conceptualize the relationship between mind and matter. By rejecting force-based models, DAT aligns psi research with contemporary physics, which emphasizes information and probability rather than deterministic causation. This alignment strengthens the theoretical credibility of parapsychology within the broader scientific community, even as it challenges conventional assumptions about the limits of mental influence.

The theory also reframes the interpretation of RNG-PK effect sizes. Rather than viewing small effects as evidence of weak or marginal phenomena, DAT suggests that the magnitude of anomalous effects reflects the information-theoretic constraints on decision augmentation. A small but consistent bias in outcome selection is exactly what the theory predicts when consciousness operates within the quantum-mechanical framework. This perspective transforms apparent limitations into theoretically expected features.

May’s work demonstrates how theoretical innovation can resolve apparent contradictions in empirical data. The consistent observation of small, replicable RNG-PK effects had puzzled researchers for decades. Force-based models predicted larger effects or none at all; decision-based models, by contrast, naturally accommodate modest but statistically significant anomalies. This explanatory power represents a significant advance in parapsychological theory.

The integration of precognition and presentiment into DAT also addresses a longstanding puzzle in psi research: the apparent temporal asymmetry of anomalous phenomena. If consciousness can influence future random events (precognition), why does it not influence past events with equal facility? DAT resolves this by treating all psi effects as goal-directed selection, with temporal direction determined by the organism’s intentional state rather than by fundamental asymmetries in the psi mechanism itself.

Skeptical critiques

Despite its theoretical sophistication, Decision Augmentation Theory has faced skeptical scrutiny on multiple grounds. Critics argue that the theory, while mathematically elegant, remains unfalsifiable in practice. Because DAT predicts small effects and allows for considerable variation in effect magnitude depending on task parameters and individual differences, skeptics contend that the theory can accommodate almost any empirical outcome. This flexibility, they suggest, renders DAT immune to refutation and therefore unscientific in the Popperian sense.

A second line of criticism questions whether DAT truly explains RNG-PK effects or merely relabels them. Skeptics argue that calling anomalous outcomes “goal-directed selection” rather than “psychokinetic influence” does not constitute a genuine explanation; it simply substitutes one mysterious process for another. Without a mechanistic account of how consciousness accesses quantum-level information or biases outcome selection, the theory remains, in this view, descriptive rather than explanatory.

Additionally, some critics point out that the empirical evidence for RNG-PK effects themselves remains contested. If the underlying phenomenon is questionable, then theoretical refinements to explain it become moot. Skeptics note that effect sizes in RNG-PK research are small, that replication rates vary considerably across laboratories, and that publication bias may inflate the apparent consistency of findings. From this perspective, DAT addresses a phenomenon that may not be robustly established in the first place.

Responses and evaluation

May and his collaborators have addressed the charge of unfalsifiability by emphasizing that DAT generates specific, quantitative predictions about effect magnitudes, their relationship to information content, and their scaling properties.² These predictions are testable and, in principle, falsifiable. If RNG-PK effects departed systematically from the patterns predicted by DAT, for instance, if they scaled differently with information content than the theory predicts, the theory would be refuted. The charge of unfalsifiability, May argues, conflates theoretical flexibility with empirical untestability.

Regarding the criticism that DAT merely relabels rather than explains, May’s response emphasizes that the theory provides a mathematical framework connecting psi effects to quantum mechanics and information theory.⁴ By grounding psi in established physics rather than invoking unknown forces, DAT offers genuine explanatory progress. The theory specifies how consciousness would need to interact with quantum systems to produce the observed effects, making explicit the assumptions underlying anomalous mental phenomena.

The empirical foundation for RNG-PK effects, while modest, has proven more robust than skeptics sometimes acknowledge. Meta-analytic reviews have consistently identified small but statistically significant effects across multiple independent laboratories and experimental designs.³ While publication bias and replication challenges remain legitimate concerns, the cumulative evidence supports the reality of anomalous RNG effects at some level. DAT’s contribution lies in providing a theoretically coherent interpretation of these effects.

May’s development of Decision Augmentation Theory represents a mature response to the empirical challenges posed by RNG-PK research. Rather than retreating from the data or inflating claims beyond what the evidence supports, he constructed a theoretical framework that takes the small, consistent effects seriously while remaining consistent with contemporary physics. This approach exemplifies how parapsychology can advance through rigorous theoretical work grounded in careful empirical observation.