Data-driven versus traditional definitions of household membership and household composition in demographic studies: does latent class analysis produce meaningful groupings?

Estelle McLean, London School of Hygiene and Tropical Medicine (LSHTM)
Alison Price, London School of Hygiene and Tropical Medicine
Luigi Palla, University of Rome La Sapienza
Emma Slaymaker, London School of Hygiene and Tropical Medicine (LSHTM)
Abena Amoah, London School of Hygiene and Tropical Medicine
Amelia Crampin, Karonga Prevention Study
Albert Dube, Malawi Epidemiology and Intervention Research Unit
Fredrick Kalobekamo, Malawi Epidemiology and Intervention Research Unit
Rebecca Sear, London School of Hygiene and Tropical Medicine (LSHTM)

Determining the impact of household composition on human wellbeing is challenging, particularly in settings where use of traditional, western-centric definitions may not capture important, context-specific differences in household composition. We used data on adolescents from rural northern Malawi from 2004-2016 to create context-relevant household composition variables using latent class analysis (LCA) with two household membership definitions, ‘immediate’ and ‘expanded’ which included relatives living close by. LCA consistently identified household compositions with greater complexity than those represented in western-centric definitions, with few individuals living in ‘nuclear’ families. We created manual ‘LCA-guided’ variables, with categories similar to the latent classes, to produce household composition definitions which were suitable for use as predictor variables. Compared to western-centric definitions, LCA-guided household composition definitions provided greater detail about the contribution of household composition to variation in associations with socio-demographic factors and educational outcome. The ‘expanded’ household definition showed consistent tendencies for individuals to live close to particular relatives, however, it did not significantly change analytical conclusions. LCA identified substantive variation in household composition and provided a useful guide for generating meaningful, context-relevant definitions. We recommend that other researchers consider such an approach rather than applying Western-centric household definitions cross-culturally.

Keywords: Latent class analysis, Methodology, Family demography, Longitudinal studies

See extended abstract.

  Presented in Session 101. Living Arrangements in a Global Perspective