Birth order and family size: bias caused by changes in birth rate.

The article examines the birth order and sibship size in a series of over 20,000 psychiatric patients. Results showed that the distribution of birth ranks deviated very significantly from random, there being an over-representation of early ranks in sibships of two and three, and of late ranks in sibships of five and larger. The extent of deviation varied with the year of birth of the patients. Moreover, the type of deviation was in conformity with theoretical prediction and its fluctuations were shown to follow changes in the number of marriages and of births in the population of England and Wales. The practical and theoretical implications of the findings are discussed.