Exploratory data analysis of the FAO/DCI dataset on "improving livelihoods of HIV/AIDS affected households in Northern Province, Zambia"

Alessandra Ms Garbero, Università di Roma La Sapienza

The aim of this paper was to apply exploratory techniques in order to discover underlying patterns within the FAO/DCI quantitative baseline survey dataset. The aim of the FAO/DCI Project was to conduct a household livelihood research in order to gain a better understanding of the dynamics affecting assets and livelihood strategies that are induced by the presence of HIV/AIDS in communities and households in Northern Province, Zambia. In this paper, we decided to adopt an inductive approach, free from model based-assumptions and a priori categorisations. The complementary use of two techniques, multiple correspondence analysis and hybrid cluster analysis allowed us to derive a categorisation a posteriori which confirms and adds more information to the a priori taxonomies hypothesised in the research design of previous qualitative and quantitative studies (FAO, 2004). Key words: exploratory data analysis, multiple correspondence analysis, cluster analysis, HIV/AIDS

  See paper

Presented in Poster Session 3