RESULTS

During the public ceremonies for Princess Diana, the data taken in 12 independent recordings at various locations in Europe and the United States compounded to a significant result indicating an anomalous global effect which would occur by chance only about once in 100 repetitions of an experiment of this nature. For one of the 12 datasets (recorded by the first author), concurrent notes and index entries allowed a more detailed analysis of the time from 10:38 to 12:59, including all of the service at Westminster Abbey and some of the cortege and recessional. The result was independently significant, with Chisquare = 28.196, 16 degrees of freedom, and a probability of 0.030. The most striking deviation occurred during the data segment corresponding to the invocation and recitation of the Lord's Prayer.

Table I shows the results for each of the 12 separate databases, indicating the location, the persons responsible, and the type of random event generator, followed by the number of trials, the Chisquare, its degrees of freedom, and the associated probability. The last line of the table shows the composite chisquare and probability for all 12 datasets.

Table I: Princess Diana, 12 Datasets
LocationOperatorSource*N-TrialsChisqdfp-value
Princeton RN MicroREG 1 28438 7.417 4 0.115
Princeton YD MicroREG 2 28362 3.948 4 0.413
Princeton AL MicroREG 3 28462 12.900 4 0.012
Princeton Cont. PortREG 6 26330 6.698 4 0.153
Freiburg HB, EB PortREG 24 27976 5.386 4 0.250
Freiburg HB, EB PortREG 26 27976 7.694 4 0.104
Florida GS, LR PortREG 34 27978 0.993 4 0.911
Chicago JW PortREG 36 27980 6.387 4 0.172
Giessen JH PortREG 20 77 3.527 4 0.474
Giessen JH PortREG 21 77 3.537 4 0.472
Giessen JH FREMM 775 6.369 4 0.173
Las Vegas DR FREMM 3829 7.612 4 0.107

Total


72.468 48 0.013

* Random sources include two independent designs for random event generators from the PEAR lab, the PortREG and MicroREG, and a third independent design (FREMM) by Dick Bierman and Joop Houtkooper of Amsterdam.

Figure 1 shows the accumulating chisquare over the 12 independent recordings during the ceremonies for Princess Diana, compared with its expectation and with a curve describing the locus of a significant deviation (p = 0.05) as the database grows. Because most of the individual Chisquare increments are larger than the expected value defined by the degrees of freedom, the cumulative curve takes on a definite trend culminating in a probability of 0.013.

Figure 1: Cumulative squared deviation, or Chisquare, of the 12 independent recordings of data during the funeral ceremonies for Princess Diana. Each Chisquare has four degrees of freedom, and the cumulative trend is compared against expectation and against a 95% confidence envelope. See text for more detail.

The near coincidence of the funeral of Mother Teresa with that of Diana led to another FieldREG recording for this occasion, with essentially the same procedures. All but one of the contributors were able to provide data, and 11 independent records were obtained. Information available for specifying the active segments was less detailed, and in this case only two segments were defined, although the total time was similar for both funerals. As detailed in Table II, the results in this case show little indication of an anomalous effect, and the composite outcome is indistinguishable from chance.

Table II: Mother Teresa, 10 Datasets
LocationOperatorSource*N-TrialsChisqdfp-value
Princeton RN MicroREG1 26808 0.074 2 0.964
Princeton YD MicroREG2 26701 1.919 2 0.383
Princeton AL MicroREG3 26824 0.015 2 0.992
Princeton Cont. PortREG6 24814 2.630 2 0.268
Freiburg HB, EB PortREG24 26364 3.314 2 0.191
Freiburg HB, EB PortREG26 26365 0.120 2 0.942
Tucson GS, LR PortREG34 26367 2.003 2 0.367
Giessen JH PortREG20 72 1.461 2 0.482
Giessen JH PortREG21 72 4.010 2 0.135
Giessen JH FREMM 720 2.696 2 0.260
Las Vegas DR FREMM 3609 0.743 2 0.690

Total


19.004 22 0.645

Figure 2 shows the accumulating chisquare over the 11 independent recordings at the ceremonies for Mother Teresa, using the same format as in Figure 1. In this case, the accumulating Chisquare values do not take on a trend relative to expectation.

Figure 2: Cumulative squared deviation, or Chisquare, of the 11 independent recordings of data during the funeral ceremonies for Mother Teresa. Each Chisquare has two degrees of freedom, and the cumulative trend is compared against expectation and against a 95% confidence envelope. See text for more detail.

Graphs of the individual random walks described by the data from the 12 recordings during Diana's funeral show strong, highly variable excursions, some of which coincide with the defined analytical segments, resulting in large contributions to the Chisquare. If all the traces are averaged, the resulting graph shows a consistent negative trend for the first four hours, including a relatively strong portion during the ceremonies at Westminster Abbey. About an hour later, the averaged data take on a positive trend that brings the cumulative deviation back to expectation. In contrast, the random walks described by the 11 datasets recorded during the ceremonies for Mother Teresa generally exhibit smaller deviations. An average across the traces is striking for its close adherence to the straight line of theoretical expectation.