Analyzing fertility histories: Do restrictions bias results?

The practice of collecting information on only the last closed and the open birth intervals results in a biased set of intervals. Given this bias, can analysis of these intervals provide unbiased estimates of the effects of socioeconomic and proximate variables on birth interval dynamics? Using an unrestricted data set, we examine the determinants of birth interval dynamics on both a set of unbiased intervals and a set of intervals that only include last closed and open intervals. The surprising result is that the biased set of intervals gives unbiased results regarding the structure of the effects of socioeconomic and proximate variables.