Evaluation of the Nigeria Fertility Survey 1981-2.

Type Working Paper
Title Evaluation of the Nigeria Fertility Survey 1981-2.
Author(s)
Publication (Day/Month/Year) 1985
URL http://www.popline.org/node/420690
Abstract
This document evaluates the quality of the data collected in the 1981-82 Nigeria Fertility Survey. It aims to discover the types of errors found in the data, the sources of these errors, and their effects on estimates of age, nupitality, fertility, and infant and child mortality. The assessment was based largely on internal consistency checks and comparisons of data obtained from different sections of the same survey. Reporting of age was very poor in both the household and individual survey. There was substantial heaping at ages ending in 0 and 5 (especially at age 30) and advoidance of almost all other digits. Age misreporting was especially pronounced among males and rural resents. There was also substantial heaping in all nuptiality data involving dates: age at 1st marriage, year of 1st marriage, and years since 1st marriage. Heaping tended to occur in rounded years or in rounded durations from date of interview. Age at 1st marriage showed a greater concentration around the age of 15 years than would be expected on the basis of cultural patterns. Older women were more likely to report higher ages at 1st marriage and a longer interval between 1st marriage and 1st birth than younger respondents. However, it was observed that there has been a recent, genuine decline in the proportion of Nigerian women married, especially among those under age 25 years. Similarly, knowledge of the exact dates of births of children ever born was low. There were indications of omission of births, particularly female births, among the older cohorts. The exact data of birth was reported for 1/3 of the last live births and 1/4 of 1st live births. There may have been some omission of infant and child deaths, especially deaths to females and 1st births, in periods 10 years or more before the survey. However, the decline in infant and child mortality for the last decade is considered genuine and universal. It is concluded that whereas dating of vital events was poor in this survey, coverage of events was relatively adequate. Thus, most estimates of fertility, nupitiality, and mortality are reliable to the extent that they are presented for either broad age groups or broad durations. - See more at: http://www.popline.org/node/420690#sthash.ckOu7qYQ.dpuf

Related studies

»