Abstract
This paper reviews and develops techniques of estimating the proportion of marriages that end in divorce which are based on disruption rates observed during a particular period of time. A source of bias in commonly used indirect estimation procedures is identified and methods of adjusting for the bias are developed. Application of the new techniques to American data of 1973 suggests that 44% of marriages would end in divorce under disruption rates of that year.

This publication has 1 reference indexed in Scilit: