Responds to "Birth Order and Intelligence: Further Tests of the Confluence Model" by Robert D. Retherford and William H. Sewell. Reviews their arguments and suggests that their analyses of the Wisconsin Longitudinal Study data support rather than contradict the model. (CJS)

Extensive data on twins, nontwin siblings, siblings separated by adoption, and parent-child similarity indicate that the heritability of intelligence is closer to .50 than to .70. Differences could be due to environmental or genetic changes in the population, or to methodology. (Author/CP)

Confluence theory was developed to explain the negative effects of birth order on intelligence. Using aggregate, between-family, within-family, and paired-sibling data from the Wisconsin Longitudinal Study, tests the mathematical form of confluence theory and finds no support for it. Suggests that statistical methods used to fit the model to the…

Changing ideas on intelligence testing and the heritability of intelligence are followed through a fifty-year period. Common criticisms of intelligence tests are examined, but it is concluded that intellectual tests will continue to be of value in diagnosing strengths and weaknesses, particularly of exceptional children. (Editor)