Serendipity and Paradox
One of the joys of conducting basic research is encountering the unexpected. Occasional serendipities are more than delightful, they are actually essential to the lifeblood of science. Without surprises, scientific theories risk congealing into dogma. A discovery I recently ran into (or perhaps it ran into me) provides unanticipated support for the effects we previously observed in the Global Consciousness Project's (GCP) worldwide network of electronic random number generators (RNG) on September 11, 2001. I described those results to IONS members in an article entitled "For Whom the Bell Tolls" (IONS Review #63). The GCP runs dozens of RNGs 24 hours a day, in many major cities worldwide. The GCP is a type of mindmatter interaction monitor for events of mass interest. Headed by Dr Roger Nelson of Princeton University, the project postulates that during moments of high mental coherence among groups of people, the state of physical randomness will fluctuate, and these changes are detectable in the worldwide RNG network. So far, the evidence supports a genuine mass mind-matter interaction, including some particularly striking effects on 9/11. There are several interpretations of how these connections may arise and what they might mean; you can read more about them at the GCP's website at http://noosphere.princeton.edu.
The background of the serendipity is as follows: In September 2000, when I was working at the Boundary Institute (www.boundary.org ), I developed a series of simple web-based tests to allow people to easily test their psi abilities and to compare their performance to others. Those games can be found at www.gotpsi.org . When I joined the IONS Research Department in 2001, I designed a new set of online psi games located at www.psiarcade.com , and they too continue to attract hundreds of people every day.
`This propitious observation suggests a new way of studying collective intuition.'
Together, these two suites of online psi tests have collected more than 60 million individual trials from more than 200,000 people. These datasets represent one of the largest systematic collections of psi data ever amassed, and they provide new ways of studying these abilities. The specific test I was examining was a simple ESP card test on the gotpsi.org site. The webpage for this test presents an image with the backs of five playing cards. The user's task is to guess which card the computer will select immediately after the user makes his or her decision. This is thus a precognition test with immediate feedback. One trial involves the user making a decision and the computer randomly selecting a card and displaying it. One run in this test consists of 25 trials. Users can perform as many trials or runs as they wish. If they produce at least one run in a given day, their results appear in a daily "Hall of Fame," ranked by their performance.
The data for gotpsi.org's ESP card test from August 29, 2000 through May 31, 2004, shows that an average of 12,400 individual trials were contributed daily by dozens to hundreds of people. The data of interest were the total number of trials and hits contributed per day. Over the nearly 4-year period I was examining, there were just over 17 million trials.
Daily combined z-scores from 30-day "sliding windows," reflecting performance in the www.gotpsi.org ESP card test. The vertical dotted line corresponds to September 11, 2001.
To visualize each day's performance on this test, I calculated a measure that determines the degree to which the daily hit rate deviated from chance expectation (called a "z-score").These scores typically range between ±2; rarely do they drift beyond ±3. Positive z values correspond to users correctly selecting the ESP card targets more often than expected by chance, and negative z values correspond to systematically missing the targets.
To make it easier to visualize slower moving trends, I determined z-scores representing the overall performance observed for days 1 - 30, 2 - 31, and so on. This 30-day "sliding window" graph is shown at right. Notice that most of the curve resides between z-scores of ±2, as expected. But at one point the curve drops almost to -4, and it stays there for about two weeks. Statistically speaking the depth of this drop is most unexpected. But even more surprising was that the dotted vertical line in the graph corresponds to September 11, 2001.
The time-correspondence is visually striking. To see if this might have been a coincidence, I determined the probability of finding a curve as low as was observed within two weeks of a specified date. The answer was p = 0.0002; in other words, only one time in 5,000 occurrences of similarly constructed databases would we expect to see a drop of a similar magnitude within 14 days of 9/11. To put this into perspective, recall that this database represents about four years of data, thus we'd potentially have to wait four years times 5,000 repetitions, or nearly 20,000 cumulative years, to obtain a curve like this purely by chance.
What might this result mean? One speculation is that since the curve becomes strongly negative prior to 9/11, it may represent a collective repression of psychic ability. Repression allows us to hide from ourselves facts or feelings that are too painful to acknowledge. The hundreds of users taking this test may have been unconsciously avoiding their intuitive impressions of the events about to unfold, and as a result they also actively avoided hitting the target. While such an explanation is pure speculation at this point, this propitious observation suggests a new way of studying collective intuition.
DEAN RADIN, PhD, IONS Senior Scientist, earned a PhD in psychology and an MS in electrical engineering from the University of Illinois. He is author of The Conscious Universe and nearly 200 scientific and popular articles. 
