Discovery Lab Answers

The PEAR laboratory

Princeton Engineering Anomalies Research (PEAR) operated in the basement of Princeton University’s School of Engineering and Applied Science from 1979 until its closure in 2007 — a 28-year span that produced more than 60 publications ^[3][4]. The lab was founded by Robert G. Jahn, then Dean of Princeton’s School of Engineering, and built on a collaborative foundation with Brenda J. Dunne, its founding laboratory manager and primary experimental designer ^[3][5]. Roger D. Nelson worked at PEAR for roughly 22 years, coordinating the systematic research program ^[3]. You can read the full ESP-Nexus treatments at PEAR Lab: Operator Intention and REG/RNG Anomalies and the two Dunne pages linked at the end.

What the lab studied

PEAR pursued two main experimental lines:

• REG/RNG experiments (mind-matter interaction). Unselected human operators — volunteers with no claimed psychic ability — sat in front of electronic random event generators (REGs) and held a mental intention for the device to produce more high bits (HI condition), more low bits (LO condition), or to run without intention (baseline, BL). Operators worked at their own pace, over months or years. The pre-stated intention direction and use of unselected operators was central to the claim that any observed effect was not an artifact of experimenter selection of gifted subjects. Dunne designed the protocols that controlled for artifacts such as electromagnetic interference.
• Precognitive Remote Perception (PRP / remote viewing). Operators attempted to describe a distant target location; a committee not otherwise involved read and scored transcripts against descriptors. Dunne developed the randomization, blinded judging, and statistical-analysis refinements for these studies ^[1].

Dobyns notes a specific evolution in the remote-perception work: the scoring methods became more refined over time, but the perception transcripts themselves grew briefer and less informative — by the program’s later days operators produced much shorter descriptions than in the earliest experiments ^[1].

The data

A caution before the figures: the sources retrieved for this question are largely tributes, obituaries, and a review essay, not the primary experimental reports themselves. The quantitative figures below come mostly from a single secondary paper (Wilkinson 2023) that is in turn citing PEAR’s own summary publications. Treat them as PEAR’s reported benchmark figures, not as independently re-derived values, and note the library’s share of the primary PEAR literature on this question has not been measured — this is not a complete accounting of the program’s data.

The core empirical signature of the REG program was thus an extremely small deviation — on the order of one part in ten thousand — accumulated across a very large number of trials. This is the key methodological vulnerability of the research area: an effect that small requires enormous statistical power to detect and is correspondingly sensitive to any low-level bias in the apparatus or analysis (a Type-II / small-effect-size concern the PEAR page discusses directly).

The unsettled part

The evidence is not one-directional, and PEAR’s own program surfaced the tension:

• The MegaREG null. A high-volume replication designed to either confirm the benchmark effect with high confidence or rule it out returned a null result — no significant deviation in either intention condition. The authors reported and analyzed this themselves, weighing both that the effect might be real but suppressed by high-volume protocols and that the original results might be statistical artifacts.
• Failed external replications. Attempts to replicate PEAR’s results at other laboratories failed, a concern discussed in PEAR’s own literature (Dunne et al., 2005) and flagged by Wilkinson ^[2]. Wilkinson quotes the difficulty of not being able to replicate results on demand as “a major concern” ^[2].

ESP-Nexus takes no position on whether these anomalies reflect a genuine mind-matter interaction; the record shows a reported, very small positive benchmark effect alongside a self-reported null replication and unreplicated results elsewhere.

For deeper dives, see PEAR Lab: Operator Intention and REG/RNG Anomalies, Dunne: PEAR laboratory management and experimental design, and Dunne: PEAR laboratory management and institutional legacy.