STUMBLING ON PSI

A psi experiment that doesn’t look like one.

A new study reporting “Abnormal Returns From the Common Stock Investments of Members of the U.S. House of Representatives” has been making the rounds on the web (following the same authors’ similar 2004 analysis of Senators’ investments):

We measure abnormal returns for more than 16,000 common stock transactions made by approximately 300 House delegates from 1985 to 2001. Consistent with the study of Senatorial trading activity, we find stocks purchased by Representatives also earn significant positive abnormal returns (albeit considerably smaller returns). A portfolio that mimics the purchases of House Members beats the market by 55 basis points per month (approximately 6% annually).

To the authors, these superior returns point to an “information advantage” in trading that Congressmen have because of their legislation- and oversight-related industry ties.

Perhaps so.  But what really struck me about this study was that it was an analysis designed to reveal non-random influences on everyday-type decision-making, and could be applied in parapsychology — even disguised as a non-psi experiment, which should make it easier to fund.

I’m aware of an experiment like this in the parapsychology literature; it’s described in a cult-fave parapsychology book, Executive ESP, published in 1974.  Apparently (I haven’t read it) two guys at the Newark College of Engineering got a sample of business executives to sit for precognition tests, and found that those with above-average precog scores tended to be those whose company profits had recently increased the most.

To me, such an experiment doesn’t sound convincing, because of subject recruitment bias issues and the numerous known and unknown variables involved in determining a corporate profit increase.  Also, it requires an overt precognition test, for which, these days, recruitment of appropriate subjects could be difficult.

I think that a better group of subjects would include traders at banks and hedge funds; and a better test-outcome would be imaginary earnings on a “trading test.”  Ostensibly, the test would include old ticker data from some obscure market that none of the subjects would have seen before.  Each trader would have an initial stake and would have to make trades based on his ability to read and anticipate the ticker data.  If the ticker data is real data for the price of a real security, then this ability would not necessarily be psi-related.  But the experimenters could interleave intervals of “random-walk” data  within the real data.  In a grant proposal, they could explain this as a “control” test, for comparing trading prowess in a real market to trading in a random environment.  But from a psi perspective, the trading outcomes during the random-walk segments would be the outcomes of primary interest.  The subjects would be given psych tests and told simply that they were being studied for correlates of trading ability.  Neither they nor the funder would be told that they were participating in a psi experiment.  Any psi-supporting data could be reported, but as a “serendipitous” finding.

In addition to avoiding the stigma of “psychic research” and making funding easier, this style of experiment should steer clear of the sheep-goat effect, bias in recruiting (i.e., a bias towards psi believers) and the general “giggle factor.”

Traders normally work in an environment that seems almost the Darwinian ideal for the promotion of psi abilities, because even a subtle influence of such abilities on their decisions should bring very large and relatively uncomplicated rewards.  Traders also are plentiful in NYC and some other big cities; they seem to be highly competitive and love contests; and they are used to being studied for the physiological correlates of their trading prowess (e.g., finger-length ratios that indicate their fetal testosterone exposures).  So I’m a bit surprised that this sort of experiment hasn’t been done (or seems not to have been done) before.

STUMBLING ON PSI