## Interpretation of Results for Organized Meditations |

We have had many requests for interpretations of the results in the formal tests of the hypothesis that we will find deviations from random expectation in GCP data corresponding to global events. The question is particularly pertinent to large scale organized meditations where many participants are interested in the GCP's attempt to capture some indication of effects of the meditations. As background we recommend reading the GCP website to learn more about this work. It is not intended and should not be seen as a test of any particular event such as the 20 May meditations. The proper test of meditation, as teachers will inevitably say, is your experience. The GCP is a long term project asking a very general question about the presence of consciousness in the world. Each event provides an increment of insight, but no event stands alone. We know that the effect size in our statistical measures is too small for reliable interpretation of single event outcomes, but we can explain how to regard the results. The May 20 2007 meditation and prayer event involved hundreds of thousands of people around the world, and I will try to suggest what the results may imply in the context of related analyses. Complete information about the event and the formal specification of the statistical hypothesis test is linked at Global Peace Day. A brief description of the graphs we use and what they represent is here. This event is number 240 in the table of Primary Results. The formal hypothesis test specified a concatenation of data from three periods of three hours on May 20 during which large numbers of people participated in synchronized meditations. The first period corresponds to time in the middle of the day in East Asia / Australia. The second is mid-day in Europe / Africa / Middle East and Central Asia, and the third is mid-day for North, Central and South America and The Pacific. The results are interesting, with the cumulative deviation of the scores from their chance expectation showing a strong and persistent slope over the concatenation of nine hours of large-scale organized meditation. (Expectation for a cumulative deviation is a level, horizontal trend). The composite result is significant, with odds against chance greater than 20 to 1. For the sake of interpretation it is noteworthy that this result shares common features with other spiritually focused events. One useful comparison is with Transcendental Meditation research addressing the possibility that when large numbers of meditators gather and meditate together, there may be a "calming" effect on the environment. The GCP looked at a data corresponding to a series of TM collective consciousness events, examining two periods of one hour of meditation on seven Saturdays in 2006: July 29, August 5, 12, 19, 26, Sept 2, 9. These were concatenated, giving a total of 14 hours of meditation. Again the result is statistically significant, with odds of more than 100 to 1 against chance. A similar comparison can be made with an assessment in October, 2000, of a Group Mind Meditation. Though this was probably a smaller group, and the time was only 15 minutes, the analysis also showed a significant trend, with odds against chance of 50 to 1. A very similar event we explored was the Avebury Global Meditation, which has replicatins in 2002 and 2004, and both local (FieldREG) and global (GCP) analyses. The nature of random data is such that these results could be chance fluctuation, but the likelihood of a chance interpretation seems small. Graphic displays of the historical data accumulation indicate a small but very consistent bias correlated with the meditation periods. The timing of the strong secular trends is strikingly precise. Surrounding data displays expected random variation, but does not show consistent trends such as we find during the periods when great numbers of people are sitting in meditation and prayer with a common focus. The correspondence of the results in these selected examples is striking, and lends credence to a meaningful interpretation. The departures from expectation are statistically significant in these examples, but there are other formal tests with weaker or ambiguous results. The GCP database as a whole shows that we are dealing with a low signal to noise ratio, with the implication that statistical noise may either obscure a real result or masquerade as one. We will acquire more confidence in the general implication that meditation and prayer have a creative presence in the world as we gather more samples of what happens in scientific research. Of course the real measure of the effect of meditation is on the people involved. The Buddhist wisdom that we must "start at home" has application and importance well beyond our statistics.
Because the public and even most scientists do not have a
sophisticated understanding of statistical analysis and the
implications of signal to noise ratios, it is worth repeating
that a discussion like this does not answer all the questions,
though it addresses most of the relevant issues. This text
was drafted in response to questions Ervin Laszlo asked
pursuant to writing a press release. To the extent such a
public statement refers to the GCP analysis, I want to avoid
misleading the reader, and in correspondence with him added
more cautionary statements: "Some of the thoughts and questions you
present are not ones for which there is a straightforward
picture. In particular, the GCP data simply do not have the
character (viz effect size) you imagine, that is, we are not
dealing with a strong causal relationship. Thus, the
comparisons of magnitude you envision are not viable because
the statistical noise prevents differentiation. You will
note, probably, that of the three examples I give, the odds
ratio is least impressive for May 20. However, the
differences are |