Periodic effect of group meditation on GCP EGG behaviour

William C. Treurniet, July, 2009

Summary. Analysis of two comparable series of group meditation sessions indicated that the effect on GCP EGG behaviour was cyclic with a periodicity of nine days. The cumulative negative deviation from chance expectation appeared larger during four consecutive days of the cycle than during the remaining five days. This difference was marginally significant (p= .065). Confirmation of such a cycle would mean that there is a periodic enabling of the effect of consciousness on EGG behaviour. Taking this cycle into account might improve the accuracy of formal GCP predictions.

The Global Consciousness Project (GCP) records and stores the behavior of a network of globally distributed random event generators called EGGs (ElectroGaiaGrams). Each EGG creates a stream of random bits every second by sampling "quantum-indeterminate electronic noise". The resulting bit stream is then sent to a centralized database. The project evaluates the likelihood that the EGGs respond to events that are meaningful to humans. The evidence shows that the cumulative behaviour of the EGGs does respond to such events with odds against chance greater than a million to one.

The events that appear to affect the EGGs usually have a clear emotional component. Destructive earthquakes or terrorist attacks that evoke a fight or flight reaction are often associated with a positive deviation from chance across EGGs. There are other events, however, that are associated with a negative deviation, and these events appear to evoke compassion or empathy. For example, the airliner crash on January 15, 2009, dubbed the "Miracle on the Hudson" was associated with a negative cumulative deviation from chance. There was also a negative deviation associated with the "Live 8 Concert" on July 2, 2005, a compassionate and celebratory event. It was seen as well during New Years' Eve and other periods of assembly for the purpose of promoting compassion and peace. The negative deviation also seems more likely when a large group of people participate in a period of synchronized meditation in order to promote positive change. However, it also occurred following the "Terror Attacks in Madrid" on March 11, 2004, so the negative deviation can happened after negative events as well.

A group meditation by the Transcendental Meditation (TM) organization in 2006 is somewhat unique in that meditations happened during a number of times separated by a constant interval. Analysis of this event is presented in the GCP formal results section under "TM Resonance Aggregation". Members were asked to meditate for one-hour periods in the morning and evening of seven successive Saturdays and at the same Universal Time. The objective was to "help create a harmonious collective consciousness on a broad scale". The results showed a significant overall negative cumulative deviation from chance in EGG behaviour. The TM data were analyzed again in order to compare the size of the effect on the EGGs over successive days, and the results are presented below.

A series of synchronized meditations, called the "Nexus 2012 meditations", were also examined. People everywhere were encouraged to meditate every Saturday evening at a specified time (UTC) on the safety of the planet. As of this writing, the series consisted of thirteen sessions, and the EGG data for those times were analyzed. The Nexus group also meditated on the health of one of their members every Wednesday evening at the same time, and twelve of these sessions were included in the analysis as well.

The TM group meditations

An analysis of the TM group meditations was already presented on the the GCP Results page. Figure 1, taken from that page, shows the cumulative deviation of the Stouffer Z score from chance expectation (CumDev) for the data from two sessions per day over seven Saturdays.

Figure 1. Deviation from chance for all sessions combined. 

The data were re-analyzed to show more precisely how the effect varied over time by comparing the CumDev score per day over successive days. The two sessions for each day were summed to obtain one CumDev score per day. Figure 2 shows that the cumulative deviation from chance varied widely across days. The points on the abscissa represent successive Saturdays. Clearly, the deviation from chance reached its lowest values on the fourth and fifth Saturday.

Figure 2. Deviation from chance for successive weekly meditations. 

The extreme variability of the CumDev score over the seven successive Saturdays may be due to a number of factors including who participated in the meditations and how many meditators there were in individual sessions. Another intriguing possibility is that the variation in effectiveness is periodic for reasons not related to the meditators themselves. That is, there may be independent periodic constraints on whatever mechanism is responsible for affecting EGG variance.

The Nexus meditations

The Nexus 2012 meditations offered another opportunity to study the effect of meditation on the EGGs, and to look for a possible periodicity as suggested in the EGG data corresponding to the TM group meditations. The Nexus meditations happened every Saturday from 20:00-21:00 UTC starting on May 9, 2009, and at the time of this analysis, the series consisted of thirteen sessions. The CumDev score over all the sessions was -667.97 (df=46800), which has a Chi-square probability of 0.985. This means that there is a .015 probability that the negative cumulative deviation occurred by chance. This significant overall negative deviation is consistent with the result of the TM group meditation analysis.

Inspection of Figure 3 shows that cumulative deviations over the successive Saturdays appear to vary periodically. Sessions 3 and 8 were the most negative, and these were separated by 36 days. Sessions 6 and 11 had the most positive deviations, and the time between these sessions is also 36 days. These identical intervals suggest a possible cyclic pattern with a period of 36 days.

Figure 3. Cumulative deviations for the Nexus Saturday meditations. 

The Nexus group also meditated every Wednesday at 20:00-21:00 UTC to promote the health of an ailing member of the group. At the time of this analysis, this second series consisted of 12 individual sessions. Figure 4 shows the variation in CumDev for this series of meditations. The CumDev score over all of these sessions was 13.88 which is not statistically significant.

Figure 4. Cumulative deviations for the Nexus Wednesday meditations. 

Figure 5 shows the merged Saturday and Wednesday sessions for the Nexus series. The odd numbered tics on the abscissa refer to consecutive Saturday sessions while the even numbered tics refer to the intervening Wednesday sessions. The periodicity is no longer as clear as in Figure 3 . The samples in Figure 5 are not quite equally spaced since the time from Saturday to Wednesday is not the same as the converse. Nevertheless, the possible periodicity in the series was evaluated with a 32-point FFT using a sampling rate of 2/7 samples/day. The resulting power spectrum is shown in Figure 6.

Figure 5. Cumulative deviations for the merged Wednesday and Saturday series. 
Figure 6. Power spectrum calculated from the CumDev data shown in Figure 5. 

The maximum values in the power spectrum are at FFT bins 13 and 14 which correspond to 9.3 and 8.6 days/cycle, respectively. Much smaller peaks at bins 7 and 5 correspond to periods of approximately 18 and 37 days/cycle. It appears, therefore, that the group meditation affected EGG variance with a periodicity of about 9 days/cycle.

Further analysis of the 9-day cycle

We can assign a number from 1 to 9 to any given day of the 9-day cycle using any arbitrary starting point. A convenient calculator for this purpose is based on an interpretation of the Mayan calendar. The calendar has a repeating 9-day sequence of days called the Lords of the Night. Although not relevant to the present study, the calculator suits our needs very well. For any Gregorian date, the calculator returns the corresponding day number in the repeating Lords of the Night sequence. We will call this number the LN number.

The LN number was determined for each day of the Nexus meditations, and the associated CumDev data were grouped according to the LN number. Then the CumDev values for each LN day were averaged. Figure 7 shows the mean Cumdev per LN day for the combined Wednesday and Saturday series of the Nexus meditations. The figure shows relatively large negative deviations from days 4 to 7 inclusively, and smaller negative or positive deviations on the other LN days. This reflects the periodicity already measured with the time series analysis shown in Figure 6.

Figure 7. Mean CumDev as a function of LN day. 

The possible presence of this periodicity in the TM meditation data was also examined. Figure 8 shows the CumDev results of the morning and afternoon TM meditation sessions plotted against LN number. Note that LN days 5 and 7 were not found in the session sequence, so placeholders in the graph for these days were obtained by duplicating the value for the immediately preceding day. We can see that the patterns for the morning and afternoon sessions are quite similar to the Nexus results in Figure 7. The only significant discrepancy is the CumDev value for the morning session on LN Day 1.

Figure 8. CumDev scores for AM and PM TM sessions as a function of LN day. 

The CumDev values for the morning and afternoon TM sessions were added to the Nexus data set. Figure 9 shows the mean for each LN day after the TM values were included. Comparison of Figures 7 and 9 shows that adding the TM values resulted in relatively small changes to the shape of the function. The error bars show plus and minus the standard error of the mean, and these suggest that the distributions for LN days 4, 5, 6, and 7 may be different from those for the other days. An analysis of variance with LN day as a between factor and sessions the random factor, showed a marginally significant effect of LN day (F(8,31)= 2.109, p= .065).

Figure 9. Mean CumDev for Nexus and TM meditations as a function of LN day. 

The sum of the session CumDevs on LN days 1, 2, 3, 8, and 9 is 21.712 (df= 90,000), while the sum for LN days 4, 5, 6, and 7 is -1386.31 (df= 54000). The former deviation from chance is not statistically significant, but the latter indicates a highly significant overall reduction in EGG variance with p= 1.04 x 10-5. This statistical test is post hoc and needs to be confirmed with new data. Nevertheless, it supports the observation from each of two separate meditation series that the meditation effect might be expected primarily during LN days 4, 5, 6, and 7, and gives us some confidence that the 9-day cycle is not due to chance.

Discussion and Conclusion

The analysis of the cumulative deviations of the Nexus series of meditations suggests that the deviations varied with a period of nine days. This newfound periodicity helps to explain why a significant overall reduction in deviation from chance was obtained for the Saturday series of meditations but not for the Wednesday series. The Wednesday series may not have shown a significant overall effect because most of the sessions occurred on inauspicious days of the cycle.

The existence of this 9-day cycle may explain why some compassionate and celebratory events in the formal GCP prediction list did not reach significance while others did. As noted earlier, the airliner crash on January 15, 2009, the "Miracle on the Hudson", was associated with a negative cumulative deviation from chance. That date fell on LN Day 4. Also, the negative deviation associated with the "Live 8 Concert" on July 2, 2005, fell on LN Day 7. These responses were consistent with Figure 9. On the other hand, the "Fire the Grid II" meditation session on July 28, 2009, (LN Day 9) and the "Live H2O Celebration" on June 21, 2009, (LN Day 8) were not significant. These results might also have been predicted from Figure 9.

Not all events thought to elicit compassion have been associated with a negative deviation. A significant positive deviation from chance occurred during the "Global Peace Demonstrations" event on February 15, 2003. The date of the event corresponds to LN Day 3, a day associated with no significant deviation according to Figure 9. Perhaps the nature of the event was more aggressive than peaceful, thus accounting for the significant positive deviation.

An unexpected significant negative deviation from chance occurred after the "Madrid terrorist bombing" on March 11, 2004. Perhaps this event elicited compassionate human responses similar to events like the "Miracle on the Hudson". The date of the Madrid bombing corresponds to LN Day 6, a part of the cycle when a compassionate response would be expected to yield a negative deviation according to Figure 9.

The latter two somewhat unexpected results suggest the need for a more complete characterization of major events with regard to the effects on observers. Specifically, it would help to have an estimate of the degree of compassion and empathy elicited. People may respond fearfully to a violent event, or they may empathize with the victims of the disaster. Or, fear and/or aggression might sometimes be the predominant emotion in events that were motivated by compassion such as a public demonstration to promote peace.

In conclusion, the results suggest the hypothesis that a cumulative deviation from chance expectation due to group meditation is modulated by a cyclic influence that has a period of nine days. Perhaps the accuracy of GCP formal predictions concerning group meditations would improve if this cycle were taken into account.