In the early days of psychical research, evidence for psi
was gleaned from large effects and clear-cut demonstrations: proof was tantamount to
observing a psychic phenomenon in the raw, as it were. Nowadays, the exploration of psi is
based on a much subtler form of evidence: statistical inference. Psi is inferred from a
multiplicity of experimental trials when the average success rate deviates significantly
from chance baseline. (Actually Sir Francis Bacon sensed this way back in the 17th
century!)
Suppose we toss a perfectly balanced coin 100 times. It is intuitively obvious that the
most probable end result is 50 heads (or tails). This is the chance baseline rate -- the
outcome which, on the average, is most likely. Obviously, over the course of many sets of
coin-tosses, the actual results will fluctuate quite a bit. Outcomes of 52, 47 or 55 are
hardly unusual. On the other hand, results which deviate greatly from the expected average
- say, 65 or 30 - while possible, are improbable. The more improbable the result, the more
we are justified in considering it as evidence of psi rather than mere chance (assuming,
once again, that we verified that this is a "good" coin, with no flaws). The
research approach here is the same whether we are using coins, dice, cards, or other
random systems with a fixed set of possibilities: we ask a person to try for a particular
outcome, and compare the results with what would be expected by chance alone.
Statistical analysis of random events, whatever the field of scientific study, is based on
the "normal" curve. Its form is the famous bell-shape which represents the
distribution expected for a large sampling of random events. A basic concept here is the
"standard deviation" ( sigma or sd ), which is a standardized measure of the
distance between a given result and the expected average (chance baseline). In principle,
the majority of random events - 64% - fall within 1 sd of the expected mean; 95% of random
events should fall within 2 sd. |
