Observations on the time-dependence of GCP EGG behaviour

William C. Treurniet, February, 2008

John Walker presented an investigation of the behaviour of GCP EGGs as a function of time in a report on the GCP web site entitled, Egg Database Time Dependence Analysis. The report shows that the mean EGG scores fluctuate as expected around a value of 100 over a 24-hour period. Further, the appearance of the graph depends on whether the chosen representation of time is Universal Time, Local Mean Time, Greenwich Mean Sidereal Time, or Local Mean Sidereal Time. The graphs can be compared visually, but no quantitative comparisons were made. Spectrum analyses of the four time series are presented here to study further whether the EGGs might be influenced by external variables correlated with the different time scales. Analysis of the 96 points of each time series was done using a 128 point FFT with zero padding.

Figure 1. Power spectrum of time-dependence data
Time Domain  Frequency Domain 

It is clear from the spectrograms that the strongest cyclic behaviour occurred in the mean EGG data plotted against Local Mean Sidereal Time (LMST) in the last graph of Figure 1. Here the power is greatest at a periodicity of 3.75 cycles/day. This translates to a cycle length of 6.4 hr. It is interesting that Spottiswoode (1997) found LMST to be related to human ESP performance across a large number of experiments. Spottiswoode showed that the mean effect size in these experiments increased by about 350-450 percent in the hour spanning 13.5 hr local sidereal time. He concluded that "this is evidence of a causal connection between performance and the orientation of the receiver (i.e., a term for subject or participant), the earth and the fixed stars". Since both EGG behaviour and ESP performance appear to be subject to changes in human consciousness, it should not be too surprising if they share a correlation with another variable.

Figure 2. Power spectrum of Relative Effect Size data
Time Domain  Frequency Domain 

We can analyze the sequence of effect sizes as a function of local sidereal time in the same way. The data plotted in Spottiswoode's Figure 3 was digitized manually at the same resolution as Walker's EGG plots, and the left side of Figure 2 displays the result. The accompanying spectrogram shows only one significant periodicity at 2.25 cycles/day which corresponds to a cycle length of 10.67 hr. If the effect size data does contain a periodicity near 3.75 cycles/day, it is swamped by the much stronger 2.25 cycles/day. Given the different periodicities in the two data sets, we cannot yet say that EGG behaviour and ESP performance may be influenced by a common source.

It is worth noting that the peak effect size at 13.5 hr corresponds to a minimum in the mean EGG values. Since the lower EGG values were not a specific target in a psychokinesis experiment, this deviation from chance cannot be construed as an effect size. If the periodicities were similar, we could say only that the lower mean EGG values result from an influence that also facilitates a large effect size in ESP experiments, and that this influence varies with local sidereal time.

The relationship between local sidereal time and EGG behaviour should be examined more closely by extracting multiple independent samples from the EGG database. If the relationship is real, it will be present in more samples than would be expected by chance.