Human random capacities through repeated numeric sampling
Thomas, Aileen Elizabeth
MetadataShow full item record
Randomness in the macroworld of human actions would contradict the theory of causal determinism, that all events occur as a direct result from antecedent factors, and be a pivotal contribution to philosophical debates on free will. The aim of this research is to understand human capacities and motivations of randomness. Can human beings act truly randomly? Participants completed a short survey and entered 100 “random” digits from 1 to 10 into a grid. The numeric sequences generated were statistically analyzed through tests described by Donald Knuth in The Art of Computer Programming, Vol. 2 (1981) to determine their degree of randomness. These sequences were compared against sequences generated from different methods of randomization consisting of dice rolls, decimal digits of pi, and deterministic formulas used by Texas Instruments programmable calculators and Java. Hypotheses of uniformity were tested using Chi-Squared analysis in frequency, serial, gap, and poker tests. Half the sequences generated from dice did not adequately pass the frequency test. Java, a Texas Instruments calculator, the decimal digits of pi and a majority of human participants produced sequences that did adequately pass the frequency test. Due to the theoretical foundation of the research question, it is impossible to produce concrete conclusions. However, higher proportions of survey sequences failed the various tests as they progressed in rigor and a similar pattern was present only in sequences produced from deterministic formulas in Java.