Discovery Lab Answers

Forced-Choice Clairvoyance vs. Precognition

This is one of the more carefully examined comparisons in experimental parapsychology, with a dedicated meta-analysis and several direct head-to-head experimental series addressing it.

The core methodological distinction

In a forced-choice clairvoyance trial, the target already exists at the time the participant responds — it has been selected (often by a random process) but is hidden from the participant. In a precognition trial, the target does not yet exist when the participant responds; it will be determined by a future random process, and the participant’s task is to anticipate which option that process will eventually produce.

This design difference matters because it creates genuinely different theoretical implications. If the two conditions yield comparable results, it is harder to argue that the effect depends on a classical “signal” from a hidden-but-present object. If precognition yields systematically more or fewer hits, that itself is informative about what mechanism (if any) might be at work.

What the direct meta-analytic comparison found

The most focused evidence comes from Steinkamp, Milton, and Morris (1998), a meta-analysis built specifically to compare the two conditions using studies that had tested both under similar experimental conditions ^[2]. Key findings from that analysis:

• Both precognition and clairvoyance studies produced a statistically significant cumulated overall effect.
• The mean effect sizes for the two types were virtually indistinguishable: 0.010 for precognition and 0.009 for clairvoyance (expressed as z/√N) across 22 study pairs ^[2].
• There was no statistical evidence that clairvoyance outperformed precognition, or vice versa ^[2].

That null difference is itself a substantive result: the temporal structure of the task — whether the target exists yet or not — did not appear to drive differential performance in this dataset.

The broader forced-choice meta-analytic picture

Storm and Tressoldi (2023), an update of the forced-choice literature covering 1987–2022, examined a combined database that included telepathy, clairvoyance, and precognition studies ^[1]. Their analysis found:

• Forced-choice experiments show a statistically significant but small overall effect.
• Within that combined dataset, telepathy, clairvoyance, and precognition studies were not significantly different from each other ^[1].
• A homogeneous subset of 23 studies yielded a notably weaker mean effect size ^[1].

This converges with the Steinkamp et al. finding: across the broader literature, the type of forced-choice ESP task does not reliably predict differential effect size.

Experimental attempts to probe true precognition

Steinkamp (2000, 2001) ran a more theoretically refined series designed to test whether true precognition — targets determined by genuinely future, unknowable-at-time-of-response processes — differs from a condition that superficially resembles precognition but could, in principle, be explained by real-time clairvoyance (i.e., a target selected by a pseudorandom algorithm that already existed, just unknown to the participant) ^[3][4].

The design embedded both conditions within the same experiment:

• Clairvoyance condition: target selected by pseudorandom number generator before the participant responded, but unknown to anyone.
• True precognition condition: target determined by a future algorithm (e.g., stock market figures) that genuinely did not yet exist at response time ^[3][4].

Across multiple series — postal, laboratory, and web-based replications — neither condition produced a significant advantage over the other, and the experiments did not provide strong evidence that true precognition outperforms the real-time clairvoyance control ^[3][4]. The series also highlighted that participant framing matters: one earlier experiment (Freeman, 1962, cited in ^[4]) found significantly better results when participants were explicitly told precognition was easier, suggesting expectation and belief state may modulate performance regardless of the actual temporal structure.

Unsettled signals and what remains open

Intellectual honesty requires surfacing what is not resolved here:

• No theoretical consensus exists on why clairvoyance and precognition should — or should not — produce equivalent effects. The equivalence finding is empirically interesting but theoretically underdetermined.
• The Steinkamp et al. (1998) meta-analysis ^[2] was built from studies that paired the two conditions; studies not designed for direct comparison could produce different patterns.
• Charles Tart’s work (retrieved for this question) identified a consistent performance difference between present-time ESP and precognition in forced-choice tasks, with the two producing markedly different patterns of information acquisition — a finding that runs in tension with the simple “no difference” conclusion and suggests the picture may be more complex than aggregate effect sizes convey. See the Tart: Information Acquisition Rates in Forced-Choice ESP Experiments page for detail.
• Mario Varvoglis has developed the Selfield protocol, designed specifically to optimize precognition research and distinguish it more cleanly from GESP, partly in response to the difficulty of settling this question with classical designs. Details are on the Forced-Choice Precognition and GESP Frameworks page.
• Heterogeneity across studies is present in this literature, and small effects are difficult to interpret without larger, pre-registered replications.

Summary

The overall picture from the sources retrieved for this question is one of functional equivalence across the three main forced-choice ESP categories, with a small but statistically detectable aggregate effect that does not clearly favor one temporal framing over another — while leaving open significant theoretical and methodological questions about what that equivalence (or any divergence at finer grain) actually means.

For broader context on the forced-choice paradigm itself, see Forced-Choice Psi Tasks.