Discovery Lab Answers

William G. Braud, PhD was an experimental parapsychologist whose career moved from conventional animal learning and neurochemistry research in the early 1970s through a sustained program in anomalous mental influence that he developed primarily at the Mind Science Foundation in San Antonio. His most consequential contribution was the direct mental interactions with living systems (DMILS) paradigm, which used continuous physiological recording — especially electrodermal activity (EDA) — as an objective, moment-to-moment measure of whether a distant agent’s intention could influence a shielded receiver. His ESP-Nexus profile is at William G. Braud, PhD.

Experiments

Animal learning and neurochemical transfer (early 1970s). Before moving to parapsychology, Braud ran controlled laboratory experiments on memory transfer via brain extracts in goldfish. Working with Porter V. Laird at the University of Houston, he prepared RNA-protein brain extracts from donor fish trained to high or low performance levels and injected them into naive recipient fish, testing whether behavioral tendencies transferred biochemically ^[6]. With Ronald B. Hoffman, he extended this to both response facilitation and response inhibition — injecting recipients with extracts from inhibition-trained donors to test whether suppressed behavioral patterns could be chemically communicated ^[5]. These experiments established his capacity for tightly controlled biological assay design, a methodological foundation that carried directly into his later physiological psi work.

Relaxation as a psi-conducive state (mid-1970s). Working at the University of Houston and then the Mind Science Foundation, Braud ran a series of nine experiments investigating whether muscular and mental relaxation facilitated free-response GESP (general extrasensory perception) performance ^[4]. He compared relaxed versus tense participants and correlated EMG-defined physiological relaxation with psi scoring. A related ganzfeld experiment, conceived and designed by Braud together with Robert Wood, tested free-response GESP following ganzfeld stimulation versus induced relaxation, with verbalized versus nonverbalized mentation — and reported a failure to replicate ^[3].

DMILS — remote influence on electrodermal activity. The core program Braud became most identified with tested whether a distant agent, given epochs randomly assigned as “activate” or “calm,” could shift the EDA of a shielded receiver in the intended direction. He ran this paradigm across multiple experimental series at the Mind Science Foundation, often collaborating with Marilyn J. Schlitz, who spent approximately a decade there co-developing the protocol ^[1]. Schlitz’s DMILS page documents how the Braud–Schlitz collaboration produced a series of controlled experiments in which EDA during activate epochs differed from EDA during calm epochs.

Remote observation / remote staring detection. A separate but related line examined whether a distant observer’s act of looking at a closed-circuit video image of a participant produced detectable EDA changes in that participant — eleven successive experimental series, summarized in ^[1].

Hemolysis studies. Braud also investigated whether focused intention could influence the rate of hemolysis of human red blood cells held in saline solution in a distant room — extending DMILS beyond the autonomic nervous system to a cellular biological substrate ^[1].

Pilot DMILS replication (1995). Braud co-authored a pilot replication of the remote mental influence on EDA paradigm with Dean I. Radin and Robin K. Taylor, conducted at the University of Edinburgh ^[2].

Methodology

DMILS protocol design. The core design physically separated agent and receiver in shielded rooms, with a random-epoch schedule (activate/calm) generated independently of both parties. The receiver’s EDA was recorded continuously, allowing epoch-by-epoch comparison without the receiver knowing which type of epoch was in progress. This eliminated sensory cueing and kept the receiver blind to condition assignment throughout.

Relaxation studies. Braud’s early psi-conducive-state experiments used both EMG (electromyography) to provide an objective, physiological measure of relaxation depth and self-report ratings, correlating both with psi-task outcomes ^[4]. The ganzfeld experiment reported in ^[3] followed a free-response design in which receivers attempted to identify a target from a pool after either ganzfeld stimulation or progressive relaxation induction, with mentation either verbalized or kept silent — a two-factor design intended to distinguish the contribution of sensory attenuation from that of verbal processing.

Animal neurochemistry experiments. These used blind injection conditions, pretest performance matching, and Mann-Whitney nonparametric tests across multiple post-injection test periods to track the duration as well as the magnitude of any transfer effect ^[5][6].

Statistical approach. For the DMILS and remote observation series, Braud used single-mean t-tests per experiment and, across series, Stouffer’s method for combining independent z scores into an overall meta-analytic result ^[1]. Effect sizes were reported as r values derived from t and degrees of freedom, permitting cross-study comparison.

Pilot replication (1995). The Edinburgh replication with Radin and Taylor used a “percentage score index” (PSI) statistic for direct comparability with the Braud–Schlitz Mind Science Foundation series, applied a 30-second epoch window selected before analysis, and examined geomagnetic Ap indices as a post-hoc correlate ^[2].

Data

The retrieved sources do not include a structured evidence block, so statistics here are drawn only from what the source text contains, and are reported with their surrounding context intact.

Relaxation experiments ^[4]. Across nine experiments, participants following relaxation instructions outperformed those given tension instructions on the psi task. Relaxation participants scored 9 hits and 1 miss; tension participants scored 6 hits and 4 misses. A correlation between EMG-defined relaxation and psi performance yielded ρ = +.49 (p < .05); self-rated physical relaxation correlated at ρ = +.53 (p < .05); self-rated mental relaxation at ρ = +.49 (p < .05). The overall probability across the nine experiments was reported as p = .000007.

Ganzfeld free-response experiment ^[3]. The Wood, Kirk, and Braud (1977) study was explicitly reported as a failure to replicate — neither ganzfeld stimulation nor induced relaxation produced significant free-response GESP performance, and the study’s own title names this outcome.

DMILS / EDA series — eleven remote-observation experiments ^[1]. The eleven successive experimental series on electrodermal detection of remote observation included results ranging from an effect size of r = –.57 to r = +.50, with a mean of r = +.25. Seven of the eleven experiments reached independent statistical significance (p ≤ .05), against a chance expectation of 5%. The combined Stouffer z across the eleven series was 3.87 (p = .000054).

DMILS / EDA — nineteen combined experiments ^[1]. Across the broader set of nineteen DMILS-type experiments summarized in ^[1], the combined Stouffer z was 4.82, with an associated p of .0000007.

Pilot replication — Radin, Taylor, and Braud (1995) ^[2]. This Edinburgh study found that overall there was less EDA in calm periods compared to activate periods, producing an effect size of r = .27 based on individual sessions as the unit of measurement. A paired activate/calm analysis showed a significant tendency for EDA to be higher during activate periods than calm periods (p = .03, one-tailed). A post-hoc analysis found that remote attention alone, independent of assigned direction, tended to raise autonomic activity over baseline (p = .001). A correlation between geomagnetic three-hourly Ap indices and absolute session-level effect magnitude was r = .727 (p = .001) — a striking finding the authors themselves treated as exploratory. The r = .27 effect size compared favorably to the prior meta-analytic estimate of r = .25 from the Mind Science Foundation series, and to r = .33 from a broader meta-analysis of 655 sessions cited in ^[2].

Goldfish neurochemistry ^[5][6]. Brain-extract recipients from inhibition-trained donors showed statistically significant inhibited striking at 24 hours (U = 14.5, p < .01) and 48 hours (U = 10, p = .001), with effects decaying at 72 and 96 hours ^[5]. Recipients of high-performance-level donor extract outperformed low-level extract recipients on Days 2 and 3 of testing ^[6].

The direction of results is not uniform. The 1977 ganzfeld experiment was a null result, and the eleven remote-observation experiments contained one negative effect size (r = –.57). The pilot replication’s geomagnetic correlation was post-hoc. These unsettled elements are part of the record.

Skeptical critiques

What critics argue. Schlitz and Braud (1997) ^[1] note within the paper itself that the experimenter-effects literature poses a direct challenge to the DMILS body of work: the same paradigm, run with an experimenter who is skeptically oriented, has produced different outcomes than when run by a believer-experimenter, raising the possibility that experimenter expectation — rather than the agent’s intention — drives observed EDA differences. The paper documents that conceptual replications at the University of Edinburgh produced outcomes that were consistent in effect size but not independently significant, which Schlitz and Braud acknowledge as a limitation. The decision augmentation model — which proposes that any apparent influence effect could be explained by the influencer selecting, through precognition, epochs that would naturally produce higher EDA — was explicitly tested in one experiment and, while the results were not supportive of that model, the authors note the test was inconclusive rather than definitive ^[1].

What the experimental data show. The eleven remote-observation experiments summarized in ^[1] included effect sizes ranging from r = –.57 to r = +.50, which means that not all series produced above-chance results. The Edinburgh replication series (Experiments 14 and 15 in ^[1]) produced effect sizes consistent with the Mind Science Foundation baseline (r = .27 in the Radin, Taylor, and Braud pilot ^[2]) but did not reach independent significance on their own. The 1977 ganzfeld study ^[3] was a direct failure to replicate an earlier positive finding, and its authors named that outcome explicitly.

Analysis. The University of Edinburgh replication attempts — including the Radin, Taylor, and Braud (1995) pilot ^[2] and the series documented in ^[1] — were conducted by researchers outside the original Mind Science Foundation group, which is relevant to the experimenter-effects concern. Those replications produced effect sizes numerically similar to the originals but did not reach independent significance. The experimenter-expectancy question — whether belief orientation of the experimenter systematically modulates DMILS outcomes — is the sharpest open methodological issue the sources name, and it has not been resolved within the literature retrieved here. Schlitz’s own three-study collaboration with a skeptically oriented experimenter, which bears directly on this question, is documented on her ESP-Nexus profile.