Towards a Forensic Parapsychology in the OT Paradigm

In parapsychology, unlike psychology, there is no known way to determine whether any measuredeffect comes from the (so-called) subjects - S, or the (nominal) experimenter - E. In each generation parapsychology has been dominated by a mere handful of experimenters who report regular extra-chance results, while the great mass of experimenters encounter, at best, only sporadic success. This paper pursues the view that parapsychology's elite are themselves particularly endowed with psi ability: they attribute scoring to subjects, in line with the psychological tradition, while it actually comes from their own psi. The Geiger counter of the radioactive man ticks everywhere! It is suggested that the experiment using subjects is a ritual tailored to maximize experimenter-psi. Today the experiment with subjects seems to replace the Juju of former times. The E-centric view explains why parapsychological experiments can rarely be repeated by different experimenters. Skeptics maintain that many parapsychology results are due to the experimenter affecting his results in an "improper" way: the E-centric model agrees, but suggests that he does so with his mind rather than with his hands. What has been lacking until now are systematic methods to measure how much of a psi effect is due to subjects, and how much to the experimenter. The physics-based Observational Theories break this impasse. An outline is given of the author's take on these theories as Minimalistic Observational Theory (MOT). The parapsychological experiment is physically no more than a source of (informational) entropy which is acted on by psi. If the subject is the source of psi then he affects the result at the level of the trial. On the other hand, if the experimenter is the source of psi he (typically) affects the experiment as a single global unit. The consequences of this elementary dichotomy are remarkably wide-reaching and testable. Only two examples are noted here: for E-psi a small experiment is as good as a big one and the usual measure of effect size (based on the trial) is inappropriate; another is that the "error" mean squares in ANOVA will be depressed by E-psi. The effects predicted by the MOT allow partitioning of the effect into S and E components. Typically Psi effects are small and differences between models are smaller still; so a great deal of data is required. Recourse must be taken to meta-analysis as well as the modern ultra-large experiment.