32 subjects attempted to generate 100 random digits. Their responses were measured using eight tests, the results of which were intercorrelated and factor analyzed. 32 simulated subjects were also generated from a table of random numbers and subjected to the same analysis. Three factors emerged from the human data: repetition, seriation, and cycling. These results support the 1970 analysis of Rabinowitz. It was suggested that these factors should be considered in any attempt to relate random production to a concept of randomness.

The intrauterine position relative to members of the same or opposite sex that a rodent fetus occupies affects both its morphology and behaviour when adult. Female fetuses that mature between males are androgenized by testosterone crossing fetal membranes, and their phenotypes as adults differ significantly from those of sisters that received less intrauterine exposure to exogenous testosterone. We report here that adult female Mongolian gerbils that gestated between male fetuses produce litters containing a significantly greater proportion of sons than the litters produced by those that gestated between female fetuses. Consequently, daughters delivered by dams that gestated between male fetuses are more likely to have gestated between male fetuses and be androgenized in utero than are daughters of dams that gestated between female fetuses. Female gerbils thus tend to inherit the phenotype (either androgenized or not androgenized) of their respective mothers.

A method was described for simulating human performance on a random generation task. It was illustrated by comparing protocols thus simulated with results produced by humans themselves. Using tests to represent scores on repetition, seriation, and cycling, simulated results were similar to those of actual subjects but differed significantly from scores taken directly from truly random numbers.

No abstract