Mark Rosenzweig and I have just written the preface for a special issue of the Journal of Development Economics focused on measurement and survey design. Rather than just summarize the papers, we tried to draw some lessons/themes of what the 13 papers in the special issue suggest. You can find the preface here.
Here are a couple of the points – read the preface for the full list of lessons:
Measurement choices can substantively affect our understanding of the prevalence of certain phenomena. For example, Bet Caeyers and co-authors note that the incidence of malnutrition in their standard paper-based survey in Tanzania would be 21 percent, compared to 8 percent using a computer-assisted questionnaire which employed consistency checks and pictures to better measure consumption units.
Survey design and methods of obtaining information can have substantial effects on inferences about key relationships. This is most striking in the paper by Jishnu Das and co-authors, who show that differential recall across income groups actually reverses the sign of the gradient between doctor visits and per-capita expenditures when using weekly rather than monthly recall surveys
Overall I think this is an interesting special issue and well worth reading for those of you involved in designing surveys or experiments. I note finally that efforts to improve measurement can themselves give new insights into the lives of the poor. This is evident in some of the papers in the special issue, as well as in the intensive financial diaries-based approach used in the book Portfolios of the Poor, which uses much more intensive measurement than is commonly the case to offer new insights into how the poor manage their money.