My Library

This problem of sampling can be solved only by resorting to the theory of probability and higher mathematics. There are at least four stress factors that statistical analysis can seldom escape altogether. First, no matter how sophisticated the mathematics that can be applied to statistical methods, the final yield only indicates a probability, not a certainty, that the assertions behind the mathematics are true. Second, the probability arrived at may be impossible to demonstrate. Third, there has been no shortage of new mathematical formulas developed within various disciplines, and so something depends of which formula statisticians utilize to crunch their numbers -- while submitting the same samples to different statistical analyses often yields significantly different probabilities. Fourth, the contexts of samples considered consistent enough can suddenly undergo change by new discoveries regarding them, and so statistical analysis of a given sample is relative to what the sample is thought to represent at any given time. It is somewhat amusing to learn that "science" of statistics was invented by a gambler as a way to help wager bits in line with statistically indicated probabilities. Thus, the USE of statistical analysis has always been a risk-taking affair: i.e., this or that statistical number crunching gives this or that probability. After the numerical probability is achieved, all that remains is how much is one willing to bet that the probability is true. With regard to statistical parapsychology, its researchers confirmed time and again that microPSI is statistically present in the many different kinds of samples with which they have experimented. Indeed, there can be no scientific or any other kind of doubt about this.