When a series of experiments are done on a topic, it is important to document all the experiments that have been done so that a reasonable evaluation can be made of their significance taken as a whole. This is what we do here. No other experiments involving other key words were done other than what is documented here.
After having done the experiments using as a required key word set ורשא גטו, Warsaw Ghetto, against other key words 9th of Av, Deportation, (5)702 for the year 1942, Holocaust, 14th of Nisan Rebellion, is Burned, and is Destroyed, we thought to check using these two key words as a key word phrase either as ורשאגטו or as גטוורשא . We found that the key word phrases ורשאגטו and גטוורשא bring some additional support to the encoding. When we looked at גטוורשא we realized that the last letter of גטו and the first letter of ורשא are identical so we thought also to try the phrase גטורשא where the ו overlaps.
The table below shows a summary of the supplementary experiments we did and the resulting p-values taken individually and the resulting p-value when all the seven experiments with the required key word/phrase are taken together. This combined p-value is also produced by a Monte Carlo experiment using our standard protocol.
|טאב||גרוש||תשב||השואה||יד ניסן מרד||נשרפים||נשמדים||Expected
|ורשא גטו||34/1,000||*772.5/1,000||26.5/1,000||22/10,000||111.5/1,000||116/1,000||17.5/10,000||40||< .5/10,000|
+ The observed ELSs had non-resonant skips and no table was possible.
Now we show the most statistically significant table from each of the three supplementary experiments.
We took the six key word lists which produced statistically significant tables with the expected number of ELSs set to 40 and made a combined experiment in accordance with our standard protocol. Each table had the key word Warsaw Ghetto. The probability that a text from the ELS random placement text population would produce tables as compact as those produced by the Torah text is 1.5/100,000. The experiment pairing Warsaw Ghetto with Deportation with the expected number of ELSs set to 500 produced a p-value of less than 1/100,000.