Data Access:

Display Descriptions:

General Description

For convenience, we refer to each random source (REG or RNG), together with its computer and internet link, as an egg. The fundamental data from the egg are binary samples from the REG, an electronic source of "true" random fluctuations. Each sample is registered as a bit with equal probability for the values 1 or 0. These sample bits are accumulated in units (sometimes called "trials,") of 200 bits each, taken at one-second intervals. The bits are summed, and the sums, indexed for location and time, are archived for analysis. The expected value for such a sampling unit is 100, with a variance of 50 (standard deviation 7.071). Thus, the mean score during any arbitrarily selected time-period for any of the eggs, or for the composite across all eggs, is expected to be about 100, with chance fluctuations in a range that is predictable from theory and confirmed in calibrations. Simple statistical procedures are applied to determine whether, in any properly specified subset, the data show evidence of deviation from this expectation.

During the first few days of network up-time, three eggs were running, and later these were joined by a fourth; more will be added as host sites are developed. The current number of eggs and some summary statistics are available in the network status display.

All the data from the Global Consciousness Project are presented as summaries for each 24-hour day (in Coordinated Universal Time or UTC, also called Greenwich Mean Time). To facilitate selection of the data of interest, the available dates are provided in calendar form. All dates and times in the tables are in UTC, to provide an unambiguous and consistent reference for the data. At the time of this writing in early September, the clocks for the various eggs are synchronized by hand, and will typically differ by a few seconds. As of 98-09-21, all of the online eggs use the NTP program to synchronize their clocks, and hence the data, precisely.

1. Z-Score Tables

For each egg and the composite of all eggs, the first eggsummary table presents the number of "ones" in a given number of samples, and an unsigned Z-score (normalized deviation) for each 15-minute segment. The terminology of the column labels is slightly different from that described above. In these tables, "trials" is used instead of "samples" to refer to the total number of bits accumulated in the 15-minute interval.

Significant Z-scores (that is, Z is 2 or larger) are highlighted in blue.

Differing numbers of trials occur because, during this pilot phase, the eggs are not always connected and gathering data, and because higher-priority computing tasks sometimes prevent the intended sample from being taken within its specified 1-second interval.

Please note that while the composite Z in the right-most column is a correct representation of the behavior of the data averaged across all eggs, this is not the measure that is defined for evaluating the global consciousness hypothesis. That measure is based on the independent deviations of the individual eggs within the periods specified in hypotheses, and accumulated as a Chisquare-distributed sum of Z�.

2. Raw Score Extrema

The maximum and minimum raw scores (sums of 200 bits) and the corresponding Z-Scores for these extrema within 15-minute segments are shown for each egg. Cell background highlighting in red indicates a significant minimum value, green a significant maximum value, and yellow indicates significant minimum and maximum values in this interval.

3. Graphs of the Data

3a.
The mean deviation for each egg, cumulated within the 15-minute segments, is plotted as a series of linked points, allowing a visual inspection of tendencies toward lower or higher raw score levels during the course of 24-hour periods. In one version of the display in which many of the earlier days' data are presented, a plot for each egg was created. The current version shows all eggs overlaid in one graph. Although different colors are used, the impression is mainly of the range and general variability of the data.

3b.
A separate plot presents the composite mean value across all eggs in each 15-minute period across the 24 hours.

3c.
A third graph shows the same data converted to the Chisquare format (Squared Z-scores), and cumulated in 15 minute blocks over the 24-hour period for each egg separately (colors). The composite Z-score across eggs (last column in the Z-score table) is treated the same way and plotted in black with points. In both cases, the single degree of freedom associated with each Z-squared is subtracted ((Z*Z)-1), so that the expectation for the cumulative deviation is a line of zero slope.

Note: defective data were sent by one egg (1005) following the San Francisco blackout, on 1998-12-09 to 16. These data will be removed, but for the time being they make the corresponding Cumulative graphs unreadable. If you download data from this period, these data must be excluded from any analysis.

The Chisquare for each of the data lines can be estimated by reading the y-axis value of the termination, and adding the expectation, 96. For example, if the line terminates at +12, the Chisquare is 108, with 96 degrees of freedom. You can obtain the probability that this value is a chance fluctuation using a Chisquare probability calculator. Since Chisquare is additive, you can sum independent Chisquares and their associated degrees of freedom to obtain a composite estimate.

3.1. Colors

The colors assigned to the individual eggs in the Cumulative Z-Squared graph are in the order of the ID numbers. As the list of reporting eggs changes, the color assignments will change accordingly, and with a larger number of eggs, the assignments will become redundant. As an example, on Jan 1, 1999 the assignments were:

4. Other Data Presentations

The analytical displays described above can be accessed by clicking on Go to the data. The main page shows in calendar form the days for which summaries are currently available. Click on the day of interest and scroll down to see the available tables and graphs. The tables are large, and will take a proportionately long time to download. Created by John Walker.

The Real-Time Display is a Java-based presentation of the most recent data. It includes both sound and graphics to show the deviations registered at each egg-site, a cumulative deviation of the squared deviations over the past minute and the past hour, and a "coherence" measure that sums across all eggs. For more information go to the page. Created by Dick Bierman and Jeroen Ruwaard.

An example of the capability for comprehensive assessment of the influence of global-scale conditions on the egg data is provided in the Long-term correlation page. Here you will also find links to access the data, providing the opportunity to do your own analyses. Please let us know of interesting findings. Created by John Walker.

A view of the EGG network located on a map of the earth, and responding to the data fluctuations visually and musically is available for recent days. These are 10 Megabyte files, and will take a very long time (perhaps 30 minutes or more) to download at modem speeds. The format is an Apple QuickTime movie which can be viewed by a QuickTime plug-in on Windows 95/98 or MacOS 7+ or Silicon Graphics Media Tools in IRIX 6.5. First Cut version. Created by John Walker. The daily movies older than one month are are offline, but can be obtained by request. For a few special cases such as New Years, the daily movies remain online.

Several people have done analyses on specified data subsets corresponding to events for which a prediction had been made. Examples of these selected results provide an introduction to some of the types of analysis that can be done. More examples will be added as they become available.