Exact Analysis of a Paired Sibling Study

A data set on categories of congenital heart malformations for sibling pairs (with different malformations) of Fraser and Hunter (1975) is analyzed exactly for quasi-independence with Monte Carlo methods. Exact p-values are computed for a test of parameter significance and a test of goodness-of-fit which contradict the model of quasi-independence and confirm an earlier analysis of MacGibbon (1983).