SOME COMMENTS ON SCHÜTZENBERGER'S ANALYSIS OF DATA ON THE HUMAN SEX RATIO

Abstract

Summary Schützenberger (1949) has found that his data on the distribution of the sexes in families cannot be explained on a simple binomial hypothesis. In order to find the cause of the anomalies he considers the hypothesis that p , the probability of a birth being male, varies between families, which is supported by Edwards (1958). However, he rejects this owing to a faulty calculation and resorts to the hypothesis that the sexes of successive children in a family are correlated. The present author analyses the data using the rejected hypothesis and finds that in reality all the anomalies that Schützenberger notes may be explained with its aid. The mechanism by which Schützenberger postulates that the correlation operates is shown to be unacceptable. The two theories are discussed in the light of some results of Turpin & Schützenberger (1948), which support Schützenberger's. It is argued that the acceptance of either theory in preference to the other would be premature.