Delanoy and Morris situated their DMILS work within a lineage in which living systems had served as psi targets for decades, noting that the research is historically linked with notions of psychic healing and that its growth in popularity followed the successful outcomes reported by William Braud’s progressive research programme. Braud further developed existing methodologies for working with living systems and labelled the basic procedure DMILS; by the late 1990s, DMILS protocols were increasingly being used to address process-oriented questions rather than solely proof-oriented ones.¹ Braud’s role was also direct and practical in Delanoy’s own programme: he visited the Edinburgh laboratory in the summer of 1993 to assist with training in conducting DMILS studies, and the experimenter for the Delanoy–Sah study was trained in a chain descending from that visit.²

Sixty-four Edinburgh University students (41 female, 23 male; ages 19–30, median 21; sole selection criterion willingness to participate in a parapsychology study, an opportunity sample with no psi pre-screening) took part in 32 sessions, each involving an agent and a percipient who were close friends, with each participant contributing only one session. The experimenter for all sessions was the second author, a medical student with no prior experimental experience, trained by Delanoy on the basis of training received from William Braud. The study recorded both EDA and conscious-response psi measures; conscious responses were elicited for only one pseudo-randomly assigned half of each session, on the rationale that recording the physiological measure was unlikely to interact with making a conscious response. The positive/neutral comparison was chosen over a negative/neutral or negative/positive design for four stated reasons: earlier findings indicating negative emotions are less likely than neutral or positive ones to produce psi-hitting on conscious-response measures; agents’ likely reluctance to experience a negative emotion or transmit it to a close friend, who might be less open to receiving it; the possibility that positive and negative emotions, being equally arousing, would not produce differential EDA responses; and findings from Braud and colleagues interpretable as suggesting differential EDA responses between an emotionally arousing and an emotionally neutral condition. Significant psi-hitting was predicted for both psi measures, with exploratory analyses planned to examine interactions between them.²

The study comprised 36 sessions. Experimenters were drawn from an experimenter-training course at the IGPP; each of six trainees conducted six sessions as experimenter. To gain experience with all aspects of the DMILS environment, the trainees took turns acting as agent and receiver for the first half of the study; in the second half they worked with friends and colleagues as participants, a small, non-pre-screened convenience sample in both halves. Overall, EDA was non-significantly greater during activate than calm periods (Stouffer Z = 0.94; effect size r = .16). In the 18 sessions among trainee experimenters acting as agent and/or receiver, results showed slightly greater EDA during calming periods (Stouffer Z = −0.082; r = −.02). In the 18 friends/colleagues sessions, greater EDA during activate periods approached significance (Stouffer Z = 1.417, p = 0.07, one-tailed; r = .33). A significant release-of-effort effect was found for the non-experimenter population (Stouffer Z = 1.826, p = 0.03, one-tailed; r = .43), and in both populations release-of-effort effects exceeded the primary effects, leading the authors to suggest the possible utility of longer interaction periods and to advise against shorter rest periods.¹

The study explored local sidereal time (LST) effects for the first time in a DMILS context, as an explicitly preliminary analysis with very small cell sizes. Sessions falling within ±2 hours of LST 13.5 showed an approximately 400% increase in mean session z relative to the overall mean (N = 3, mean z = 0.629, against an overall N = 36 and mean session z = 0.157), while sessions within ±2 hours of LST 18.5 showed lower z-scores than the overall mean (N = 4, mean z = 0.076), a pattern the authors noted as consistent with Spottiswoode’s anomalous-cognition findings, but resting on three and four sessions respectively and carrying no confirmatory weight.¹

The Wackermann, Delanoy and Morris paper, published in the Journal of Parapsychology, framed DMILS detection in electrodermal records as a signal-engineering problem.³ This methodological strand sat alongside the group’s primary experiments: the agent–receiver pairing study with Morris, Brady and Roe examined a process variable within the same EDA protocol family⁷, and the IGPP study’s design recommendations, longer interaction periods, avoidance of short rest periods, derived from the observed release-of-effort pattern, fed scoring-window considerations back into protocol design.¹

The IGPP study’s internal contrast, trainee experimenters at chance when serving as participants, friends/colleagues sessions approaching significance (Stouffer Z = 1.417, p = 0.07, r = .33, vs Z = −0.082, r = −.02)¹, connects to Delanoy’s broader documentation of experimenter and setting variables. Her 1997 report on visits to six U.S. parapsychology laboratories, conducted in 1989 to inform the design of the Koestler Chair’s new facilities, organized largely tacit laboratory lore into four categories: laboratory design; orientation towards participants; participant/experimenter interactions; and experimenter orientation and preparation. Emergent principles included tailoring the laboratory environment to be comfortable and welcoming for the typical participant, spending time putting participants at ease and providing success-oriented expectations, selecting stable and psi-curious participants, and choosing experimenters with good conversational and social skills, offered explicitly not as hard facts but as shared perspectives intended to prompt future experimental examination.⁸