Testing The Torah Code Hypothesis: The Experimental Protocol

This is the second part of a tutorial discussing the experimental protocol issues in testing the Torah code hypothesis. The principal concept is the test statistic which is used to do the actual hypothesis testing of the Null hypothesis against a simple alternative or against a complex of alternatives. We illustrate the methodology using the data sets from the WRR (Witztum et al., 1994) experiment. We use the WRR key word sets of list 1 and 2 combined. The experiment produces a p-value of less than 1/100,000 in the Genesis text. We performed another experiment pairing rule based transliterations for the spellings of the names of the American presidents into Hebrew with the Hebrew word for president. Taking into account Bonferroni, the resulting p-value of the 100,000 trial experiment was less than 1/66,667