Attrition in Longitudinal Household Survey Data : Some Tests for Three Developing-Country Samples

For capturing dynamic demographic relationships, longitudinal household data can have considerable advantages over more widely used cross-sectional data. But because the collection of longitudinal data may be difficult and expensive, analysts must assess the magnitudes of the problems, specific to l...

Full description

Bibliographic Details
Main Authors: Alderman, Harold, Behrman, Jere R., Kohler, Hans-Peter, Maluccio, John A., Cotts Watkins, Susan
Format: Publications & Research
Language:en_US
Published: World Bank, Washington, DC 2015
Subjects:
Online Access:http://hdl.handle.net/10986/21336
id okr-10986-21336
recordtype oai_dc
spelling okr-10986-213362021-04-23T14:04:01Z Attrition in Longitudinal Household Survey Data : Some Tests for Three Developing-Country Samples Alderman, Harold Behrman, Jere R. Kohler, Hans-Peter Maluccio, John A. Cotts Watkins, Susan aged biases Child Development data set data sets dependent variable developed countries developing countries Development Research employment equations family planning fathers Food Policy Research household characteristics household data household income household level household survey household surveys individual level informal networks intervention interviewer interviews labor force labor force participation labor market Living Standards Living Standards Measurement migration Mortality multivariate analyses multivariate analysis multivariate regression Nutrition nutritional status outcome variables parents policy research population size Population Studies precision preschool children probabilities probability radio random sample Research Institute Research Working Papers researchers rural areas siblings significance level social networks social scientists Specification tests survey data urban areas urban communities weighting Household surveys Longitudinal method Demographic indicators Longitudinal data Sample surveys Household size Reduction in force Migrations Case studies Family background Variables (mathematics) Multivariate analysis Behavioral responses For capturing dynamic demographic relationships, longitudinal household data can have considerable advantages over more widely used cross-sectional data. But because the collection of longitudinal data may be difficult and expensive, analysts must assess the magnitudes of the problems, specific to longitudinal, but not to cross-sectional data. One problem that concerns many analysts is that sample attrition may make the interpretation of estimates problematic. Such attrition may be especially severe where there is considerable migration between rural, and urban areas. And attrition is likely to be selective on such characteristics as schooling, so high attrition is likely to bias estimates. The authors consider the extent, and implications of attrition for three longitudinal household surveys from Bolivia, Kenya, and South Africa that report very high annual attrition rates between survey rounds. Their estimates indicate that: 1) the means for a number of critical outcome, and family background variables differ significantly between those who are lost to follow-up, and those who are re-interviewed. 2) A number of family background variables are significant predictors of attrition. 3) Nevertheless, the coefficient estimates for standard family background variables in regressions, and probit equations for the majority of outcome variables in all three data sets, are not significantly affected by attrition. So attrition is apparently not a general problem for obtaining consistent estimates of the coefficients of interest for most of these outcomes. These results, which are very similar to those for industrial countries, suggest that multivariate estimates of behavioral relations may not be biased because of attrition. This would support the collection of longitudinal data. 2015-01-22T21:13:39Z 2015-01-22T21:13:39Z 2000-09 http://hdl.handle.net/10986/21336 en_US Policy Research Working Paper;No. 2447 CC BY 3.0 IGO http://creativecommons.org/licenses/by/3.0/igo World Bank, Washington, DC Publications & Research Publications & Research :: Policy Research Working Paper Africa Latin America & Caribbean Bolivia Kenya South Africa
repository_type Digital Repository
institution_category Foreign Institution
institution Digital Repositories
building World Bank Open Knowledge Repository
collection World Bank
language en_US
topic aged
biases
Child Development
data set
data sets
dependent variable
developed countries
developing countries
Development Research
employment
equations
family planning
fathers
Food Policy Research
household characteristics
household data
household income
household level
household survey
household surveys
individual level
informal networks
intervention
interviewer
interviews
labor force
labor force participation
labor market
Living Standards
Living Standards Measurement
migration
Mortality
multivariate analyses
multivariate analysis
multivariate regression
Nutrition
nutritional status
outcome variables
parents
policy research
population size
Population Studies
precision
preschool children
probabilities
probability
radio
random sample
Research Institute
Research Working Papers
researchers
rural areas
siblings
significance level
social networks
social scientists
Specification tests
survey data
urban areas
urban communities
weighting
Household surveys
Longitudinal method
Demographic indicators
Longitudinal data
Sample surveys
Household size
Reduction in force
Migrations
Case studies
Family background
Variables (mathematics)
Multivariate analysis
Behavioral responses
spellingShingle aged
biases
Child Development
data set
data sets
dependent variable
developed countries
developing countries
Development Research
employment
equations
family planning
fathers
Food Policy Research
household characteristics
household data
household income
household level
household survey
household surveys
individual level
informal networks
intervention
interviewer
interviews
labor force
labor force participation
labor market
Living Standards
Living Standards Measurement
migration
Mortality
multivariate analyses
multivariate analysis
multivariate regression
Nutrition
nutritional status
outcome variables
parents
policy research
population size
Population Studies
precision
preschool children
probabilities
probability
radio
random sample
Research Institute
Research Working Papers
researchers
rural areas
siblings
significance level
social networks
social scientists
Specification tests
survey data
urban areas
urban communities
weighting
Household surveys
Longitudinal method
Demographic indicators
Longitudinal data
Sample surveys
Household size
Reduction in force
Migrations
Case studies
Family background
Variables (mathematics)
Multivariate analysis
Behavioral responses
Alderman, Harold
Behrman, Jere R.
Kohler, Hans-Peter
Maluccio, John A.
Cotts Watkins, Susan
Attrition in Longitudinal Household Survey Data : Some Tests for Three Developing-Country Samples
geographic_facet Africa
Latin America & Caribbean
Bolivia
Kenya
South Africa
relation Policy Research Working Paper;No. 2447
description For capturing dynamic demographic relationships, longitudinal household data can have considerable advantages over more widely used cross-sectional data. But because the collection of longitudinal data may be difficult and expensive, analysts must assess the magnitudes of the problems, specific to longitudinal, but not to cross-sectional data. One problem that concerns many analysts is that sample attrition may make the interpretation of estimates problematic. Such attrition may be especially severe where there is considerable migration between rural, and urban areas. And attrition is likely to be selective on such characteristics as schooling, so high attrition is likely to bias estimates. The authors consider the extent, and implications of attrition for three longitudinal household surveys from Bolivia, Kenya, and South Africa that report very high annual attrition rates between survey rounds. Their estimates indicate that: 1) the means for a number of critical outcome, and family background variables differ significantly between those who are lost to follow-up, and those who are re-interviewed. 2) A number of family background variables are significant predictors of attrition. 3) Nevertheless, the coefficient estimates for standard family background variables in regressions, and probit equations for the majority of outcome variables in all three data sets, are not significantly affected by attrition. So attrition is apparently not a general problem for obtaining consistent estimates of the coefficients of interest for most of these outcomes. These results, which are very similar to those for industrial countries, suggest that multivariate estimates of behavioral relations may not be biased because of attrition. This would support the collection of longitudinal data.
format Publications & Research
author Alderman, Harold
Behrman, Jere R.
Kohler, Hans-Peter
Maluccio, John A.
Cotts Watkins, Susan
author_facet Alderman, Harold
Behrman, Jere R.
Kohler, Hans-Peter
Maluccio, John A.
Cotts Watkins, Susan
author_sort Alderman, Harold
title Attrition in Longitudinal Household Survey Data : Some Tests for Three Developing-Country Samples
title_short Attrition in Longitudinal Household Survey Data : Some Tests for Three Developing-Country Samples
title_full Attrition in Longitudinal Household Survey Data : Some Tests for Three Developing-Country Samples
title_fullStr Attrition in Longitudinal Household Survey Data : Some Tests for Three Developing-Country Samples
title_full_unstemmed Attrition in Longitudinal Household Survey Data : Some Tests for Three Developing-Country Samples
title_sort attrition in longitudinal household survey data : some tests for three developing-country samples
publisher World Bank, Washington, DC
publishDate 2015
url http://hdl.handle.net/10986/21336
_version_ 1764447973507334144