Users and non-users of contraception: Tests of stationarity applied to members of a family planning programme

Regression analysis is used in examining the cumulative nature of contraceptive use among clients of a family planning programme. Markov chain analyses are performed on the data. χ 2 tests are used to test the stationarity of the transition probabilities from the state of user to non-user as well as from non-user to user over the period of operation of the programme. A good predictive equation is developed relating the cumulated number of users and the years of operation of the programme. The transition probabilities are not constant over the entire period of twelve years. The data are not homogeneous. When the twelve year-period is divided into three four-year periods, the transition probabilities are constant within each period. Each of these three periods coincides with a distinctive period in the development of the programme.