Overview

Beginning in the 1990s, Sheldrake undertook a sustained empirical investigation into whether humans can detect when they are being stared at by an unseen observer. This research departed from anecdotal reports and folklore to establish controlled experimental protocols that could be replicated in schools, laboratories, and other settings. The core question was deceptively simple: when a person cannot see another person looking at them, can they nonetheless sense that observation?

Sheldrake’s approach was methodologically rigorous. He designed experiments in which one participant (the “starer”) would either look at or look away from another participant (the “subject”) sitting with their back turned or eyes closed, with the assignment of starer or non-starer determined randomly. The subject would then indicate whether they believed they were being stared at. If the sense of being stared at were purely illusory, performance should be at chance level (50%). If genuine, accuracy should exceed chance significantly.

The implications of such a finding would be substantial: it would suggest that human perception extends beyond the known sensory channels (vision, hearing, touch, taste, smell) and operates through mechanisms that conventional neuroscience has not yet identified. Sheldrake situated this research within his broader theoretical framework of the extended mind, the proposition that consciousness and perception are not confined to the brain but extend into the environment through what he termed morphic resonance.

Experimental Design and Methods

Sheldrake’s experimental protocols evolved over time, becoming progressively more stringent to eliminate potential artifacts and confounds. In early school-based studies, he employed a simple two-choice paradigm: the subject would sit with eyes closed or back turned, while a starer either looked at them or looked away, with the choice randomized. After each trial, the subject would guess whether they were being stared at.¹

A critical concern in such experiments is whether subjects might detect staring through subtle sensory cues (air movement, peripheral vision, heat, or sound) rather than through any anomalous mechanism. Sheldrake addressed this systematically. In one series of experiments, he demonstrated that the sense of being stared at does not depend on known sensory clues, such as the starer’s breathing, body heat, or any detectable physical signal.² He also examined whether possible artifacts (such as experimenter bias, sensory leakage, or statistical flukes) could account for positive results.³

The experiments were conducted in diverse settings: schools, where large numbers of participants could be tested; specialized research facilities; and other institutional environments. This breadth of venue helped establish whether effects were robust across different contexts and populations. Sheldrake also published detailed methodological critiques of his own work, identifying potential sources of error and describing how subsequent experiments were designed to eliminate them.

Key Findings

• Across multiple experiments, subjects performed significantly above chance when asked to detect whether they were being stared at, with effect sizes typically in the range of 55–60% accuracy when 50% would be expected by random guessing.⁴
• The effect was robust across different age groups, from schoolchildren to adults, and across different cultural contexts.¹
• Elimination of known sensory cues, through the use of opaque barriers, distance, and controlled environmental conditions, did not eliminate the staring-detection effect, suggesting that conventional sensory mechanisms cannot fully explain the phenomenon.²
• The effect appeared to be directional: subjects were more accurate at detecting stares from the front than from the back, and more accurate when the starer was closer rather than farther away, patterns consistent with a genuine perceptual mechanism rather than random guessing or artifact.⁵
• Sheldrake’s analysis of the sense of being stared at led him to propose implications for theories of vision itself, suggesting that vision may not be a purely passive reception of light but may involve an active, directional component that can be detected by others.⁶

Extended Mind Framework

Sheldrake situated the staring-detection phenomenon within his broader theoretical framework of the extended mind. Rather than viewing consciousness and perception as confined to the brain, he proposed that mental activity extends into the environment through morphic resonance, a process by which self-similar systems influence each other across space and time.⁷

In this framework, the act of staring (of directing attention toward another person) creates a morphic field that extends from the starer to the subject. This field is not a conventional physical force but a non-local influence that connects the two individuals. The subject’s ability to sense this field would constitute a form of extended perception: perception that reaches beyond the individual brain and body to encompass the relational space between observer and observed.

Sheldrake argued that this extended-mind perspective resolves a conceptual puzzle posed by the staring-detection effect. If the effect is real but cannot be explained by known sensory mechanisms, then either the effect is illusory (contradicting the experimental evidence) or perception must operate through mechanisms beyond conventional neuroscience. The extended-mind hypothesis offers a third possibility: that perception is fundamentally relational and non-local, extending through fields that connect organisms to their environment and to each other.⁸

This theoretical framework also connects staring detection to Sheldrake’s other research areas, including animal telepathy and telephone telepathy. In each case, the phenomenon involves apparent communication or perception across distance without known sensory mediation, and in each case, morphic resonance offers a unifying explanatory principle.

Skeptical Critiques

The staring-detection research has faced substantial skepticism from mainstream psychology and neuroscience. Critics have raised several classes of objection.

First, some skeptics argue that even small effect sizes above chance (such as 55% accuracy) can accumulate across many trials to produce statistically significant results, but that such effects may still reflect subtle sensory leakage, experimenter bias, or other methodological artifacts rather than a genuine anomalous phenomenon. The concern is that despite Sheldrake’s efforts to control for known sensory cues, unmeasured or unrecognized sources of information leakage might still be present.

Second, critics point to the file-drawer problem: studies with null results may be less likely to be published or publicized, creating a publication bias that inflates the apparent effect size in the literature. If many experiments have been conducted with negative results but remain unpublished, the published record would overestimate the true effect.

Third, some skeptics question whether the effect, even if real, requires the exotic theoretical apparatus of morphic resonance or extended mind. They suggest that conventional psychological mechanisms, such as subtle body-language reading, unconscious inference, or heightened vigilance in certain contexts, might account for above-chance performance without invoking non-local fields or anomalous perception.

Responses and Evaluation

Sheldrake addressed the artifact concern directly through progressive refinement of experimental methods. His 2001 paper on the elimination of possible artifacts documented how subsequent experiments were designed to close potential loopholes identified in earlier work.³ By increasing the distance between starer and subject, using opaque barriers, and employing randomization procedures that prevented experimenter knowledge of trial outcomes, he demonstrated that the effect persisted even under conditions that should eliminate subtle sensory leakage.

Regarding the file-drawer problem, Sheldrake published detailed accounts of his experimental protocols and results, including null findings and negative cases, making his work more transparent than typical in parapsychology. He also conducted meta-analyses of his own data across multiple studies, examining whether the overall effect size remained consistent or varied in ways that would suggest publication bias.

On the theoretical question of whether morphic resonance is necessary to explain staring detection, Sheldrake argued that conventional psychological mechanisms cannot account for the directional and distance-dependent nature of the effect. If the phenomenon were purely a matter of subtle body-language reading or unconscious inference, one would expect performance to be relatively independent of the starer’s distance or direction relative to the subject. The fact that subjects are more accurate when the starer is closer and facing them suggests a mechanism that is sensitive to spatial relationships in a way that conventional sensory psychology does not predict.⁵

Furthermore, Sheldrake’s theoretical framework makes testable predictions. If staring detection operates through morphic resonance, then the effect should be stronger between individuals with prior relational history (such as family members or close friends) than between strangers, because morphic resonance is stronger between systems with greater morphic similarity. Some of his experiments examined this prediction, though results have been mixed.

The broader evaluation of Sheldrake’s staring-detection research remains contested. Within parapsychology and consciousness studies, the work is widely cited as evidence for anomalous perception and for the extended-mind hypothesis. In mainstream psychology and neuroscience, the research is largely ignored or dismissed, with critics maintaining that the effect size is too small, the theoretical implications too radical, or the methodological concerns too substantial to warrant acceptance. The question of whether the sense of being stared at represents a genuine anomalous phenomenon or an artifact of experimental design and statistical analysis remains unresolved in the broader scientific community.⁹