Consistency of Reporting Fertility Planning Status

Abstract

2361 respondents who had 1st been interviewed in the 1970 National Fertility Study were interviewed again toward the end of 1975 in order to analyze the discrepancies between reports about the same pregnancies that were given by the same respondents. 37 of the respondents reported different numbers of births; dates of birth discrepant by more than 12 months were reported for 56 of the 4950 births to the remaining respondents. The study is based on the characteristics of the remaining 4894 matched births. In order to measure individual consistency of response a new index that is independent of the dimension of the distribution of the coded variable was developed. For the components of the code consistency is 58% for use or nonuse of contraception 65% for success or failure with use and 35% for success or failure with nonuse. The classification of failures in timing or quantity showed a consistency of 64% for users and 75% for nonusers. Differentials in consistency of response have been examined for birth order length of recall interval religion and education. While consistency is lower for events occurring a longer time before interview for Catholics and for those with less education the main finding is that these relationships are weak.