Annex 9: Example of Triangulation: Comparing Estimates of Household Income and Poverty from Different Sources

This is one of the most frequently used mixed-method designs. Two or more methods are used to confirm, cross-validate, or corroborate findings within a single study. Either the QUANT or QUAL design may be dominant or the two may be fully integrated and given equal weight. Figure 1 illustrates how this design could be used to obtain and triangulate different estimates of poverty and household income. The QUANT study uses household income as the primary indicator of poverty. A household questionnaire is administered to a sample of households, and a number of questions are included on household income and expenditures. Statistical analysis is conducted to calculate means, frequency distributions, and standard deviations of household income. Internal consistency checks are also used in which income is compared with expenditures. At the same time, a number of QUAL data collection techniques, including interviews with household members and key informants and observation, are used.

While the QUAL researchers look at household income and expenditures, they also try to understand how people in the community think about poverty, well-being, and security. Detailed notes are taken on the interviews and observation, and some interviews are taped. Photographs are also taken to illustrate the economic conditions of individual households and the community in general. A few of the households included in the QUANT household sample survey are also covered in the QUAL survey to permit direct comparisons. The notes are analyzed using the constant-comparative method, and some preliminary conclusions are drawn on how families think about well-being and economic security as well as describing their income and expenditure patterns.

The QUANT and QUAL teams then meet to present and compare their findings. For example, the QUAL team reports that although it may be true (from the QUANT data coming in) that many families have very low incomes and very modest consumption, there are some families (hard to know how many), particularly female-headed households, that are equally concerned about vulnerability and how well families will be able to survive crises such as unemployment, illness, or death. Crime and physical security are also major concerns expressed by those whom the QUAL team interviewed. The two teams also compare information obtained on the families that were included in both the QUANT household survey and interviewed and observed by the QUAL team. If specific inconsistencies arise, such as different estimates of income or expenditures for particular families, the two teams compare their notes to try to explain the differences. In some cases, resources and time permitting, one or more researchers might return to the field to check on some of these issues. The final report will integrate findings from all the different sources of QUAL and QUANT data collection and analysis.