Summary
- Parity progression ratios, an, measure the proportion of women with n child who go on to have n+1 or more children.
- PPRs may be calculated for age cohorts of women who were born in the same period, for parity cohorts who have their nth birth in the same period, or as period measures referring to women who have their n+1th birth in the same period of time.
- PPRs can be calculated from birth history data using standard life table methods. These methods can be adapted to produce unbiased estimates of cohort PPRs from data that are incomplete because they are censored at the date of interview.
- The crucial issue to handle correctly in the calculation of parity cohort PPRs from data of women of childbearing age is censoring by interview of open-ended intervals.
- The crucial issue to handle correctly in the calculation of period PPRs is the calculation of the time women spent exposed to the risk of having an n+1th birth in each interval segment within each period of interest.
- Calculation of a life table measuring fertility by interval duration also allows one to estimate a range of indices of the average length of birth intervals including their median, mean and life table expected durations. Conditional indices of the length of birth intervals are often calculated referring to women who have another birth before some specified cut-off duration.
- One extension of the methods described in this session is to use regression modelling of the fertility rates to control for age so as to adjust for truncation bias and to examine the impact of other covariates on parity progression.
- Another extension of the methods is to calculate projected PPRs that estimate the final values of the measures for women of childbearing age from their progression to date and current fertility.