Birth order as a quantitative expression of date of birth.

Publications on birth order mostly differ in their findings. In 1874 Galton1 stated that firstborn dominated among eminent men. Thirty years later, Pearson2 claimed that society was corrupted through the defects of firstborn. Surveys3-covering hundreds of books and articles give a remarkably inconsistent picture of the influence of sibling position on physical and psychological development. In 1913 Weinberg10 showed that methodological slips are an important source of confusing results. One such slip is the idea that the number of older siblings (earlyborn) and the number of younger siblings (latebom) should balance in the population. If a total sibship is considered as a unit, beginning at 0 with the oldest sibling and ending at 1 with the youngest one, it is obvious that 0-5 is the exact mean position of its members. The same statement is true for any sample of total sibships but not for a sample consisting of members of different sibships. In fact, Cobb,"1 as early as 1914 pointed to circumstances that might fundamentally distort birth order balance in the general population, but his warnings were not taken seriously. The consequences have been devastating. Those who have studied the influence of sibling position on disease, character, performance, or fate have mostly used methods based on the Greenwood-Yule model,12 with its erroneous assumption of a balanced birth order distribution in the general population. If a researcher found that a group of trait-carriers contained more lateborn than earlyborn, or the reverse, he asked whether the ratio deviated significantly from a 50: 50 distribution. If so he felt entitled to state that the trait and the skewed distribution ("overrepresentation") had something to do with each other. Empirical models Some investigators have taken the precaution of comparing their propositi with trait-carriers of other types13 or with "social twins"" in order to keep as many variables as possible under control. These authors, however, did not look for the reasons why the birth order distribution of their comparison group tended to deviate in a similar manner as that of their propositus group. This remarkable fact cannot be explained by random deviations but indicates regular influences causing a characteristic skewness of birth order distribution in both groups. Balanced models rejected A survey of more than 400 studies15 discovered only four authors who have tackled the question why lateborn and earlyborn surge in powerful and seemingly spontaneous waves in cohorts of the general population. Half a …