Dean Radin’s Statistics: The “Significant” Correlation Matrix (PA 1993)

In 1993 Dean Radin served as president of the Parapsychological Association (PA), the professional association of parapsychologists. By that time he had been working in the field for over a decade. He had held positions doing psi research at Bell Laboratories, Princeton University, SRI International, and the University of Edinburgh. This impressive set of credentials may have led some to feel confident in his experimental methods and statistical analyses.

At the 1993 PA convention held in Toronto, Radin presented a paper titled: “Environmental Modulation and Statistical Equilibrium in Mind-Matter Interaction.” It reported a study of a human’s mental influence on a Geiger counter.

Method

Four Geiger counters were connected to a computer that recorded the radiation counts. The human subject attempted to influence one of the Geiger counters. The 65 test sessions each had two conditions (real-time and pre-recorded). For each condition, the computer randomly designated 25 influence periods and 75 control periods. Results from influence and control periods were compared for each Geiger counter, for each condition. Two measures were computed (effect size and F-score). Four Geiger counters, two conditions with two measures each, resulted in 4 × 2 × 2 = 16 outcome measures for each session.

In addition, 33 environmental variables were recorded for each test session (e.g., humidity, barometric pressure, time of day, local magnetic field, precipitation, sunspot number, background xray).

Statistics

Correlation coefficients were calculated for each of the 16 outcome measures with each of the 33 environmental variables, which resulted in a total of 16 × 33 = 528 correlations.

Radin reported that the 16 × 33 matrix produced 44 correlations that were associated with p < .05. He then used the binomial probability distribution to compute the probability of obtaining that many, or (presumably) more, correlations associated with p < .05.

He reported a value of p = .0004.

I have been unable to reproduce this number; three online binomial calculators have all given me p = .000779. I did check the general accuracy of these online calculators by comparing their results with the Tables of the Cumulative Binomial Probability Distribution (1955) for similar values with N = 550 and p = .05. In any event, Radin’s reported result is statistically significant.

However, the binomial distribution assumes independence for each of the measurements. But the correlations were clearly not independent. For instance, the environmental variables included background X-ray flux and log of background X-ray flux, humidity and precipitation, sunspot number and sunspot number for the day before.

Implications

The program chair for the 1993 conference was Marilyn Schlitz, who is now Vice President for Research and Education at the Institute of Noetic Sciences, where Radin is Senior Scientist. Schlitz was responsible for seeing that the papers were adequately reviewed. She has been involved in parapsychology since 1979.

Radin has been doing statistically based parapsychology research since 1981.

Readers who have some familiarity with statistics may wish to ponder the implications.


References

Radin, Dean I. (1993). Environmental modulation and statistical equilibrium in mind-matter interaction. In Marilyn J. Schlitz (program chair), The Parapsychological Association 36th Annual Convention: Proceedings of Presented Papers (pages 157-176). The Parapsychological Association.

Staff of the Computation Laboratory. (1955). Tables of the Cumulative Binomial Distribution. Cambridge, MA: Harvard University Press.

Dean Radin’s Statistics: Entangled Minds, page 120

Dean Radin discusses a meta-analysis of 88 ganzfeld experiments in his book Entangled Minds: Extrasensory Experiences in a Quantum Reality (Paraview, 2006). Figure 6-6 on page 120 has a graph of the results.

The three-sentence caption is odd. The first sentence seems direct and plausible: “Cumulative average hit rate in the ganzfeld experiments, from 1974 through 2004, with one standard error bar.”

But the third sentence reads: “Each dot represents an experiment and the dates on the x-axis indicate the average year of study publication.” Can these two sentences be reconciled? It seems unlikely that each dot represents one experiment (the variance of the hit rates would be remarkably small). Why would the average year of publication be plotted on the x-axis? What range of years were used for each average?

I perused the book but could find no citation that might lead me to more description.

Before one relies upon Dr. Radin’s claims here, one may wish to seek further details of the meta-analysis.