The allocation of resources across generations and the consequences of these allocations represent a research agenda with significant policy implications. At the same time, their empirical investigation imposes immense data requirements, and therefore data collection challenges. In this paper, we describe how we met these challenges, in the Resource Flows Among Three Generations in Guatemala Study, or IGT, carried out in 2006–07. In doing so, we provide a guide for using and interpreting the data collected as part of IGT, as well as an example for others interested in implementing research projects on similar themes elsewhere. Complex research topics, across generations and across a range of possible measures of well-being, led to a relatively complicated sample selection process and survey design, with component modules that were applicable to different “types” of sample members, depending on their generational status and age, and who often lived in different locations. It also led to a wide set of survey domains, ranging from economic, educational, and psychological surveys to clinical medical exams for both the young and the elderly. Survey coverage was above 85% of the targeted sample for most categories of respondents and most modules, and a number of safeguards were in place to ensure high quality data. Biases due to attrition, measured against the original 1970s rounds of survey work upon which IGT built, while present, should not reduce substantially the validity of research findings to come from this rich sample. The extent to which this is true, though, may vary depending on the topic under consideration and the controls included in the analyses.
