Finding the right words: Qualitative study to design a survey module for measuring a natural resource management package in Vietnam

It is well known that careful qualitative research can help inform survey design, especially when the issues are complex or new. An important challenge in agricultural policy is to find ways of improving yields and profits while also delivering environmental benefits. Natural resource management (NRM) packages which are designed to deliver on these twin goals have consequently become a policy priority.  The jury is still out, however, on whether there is evidence to support the idea that farmers have successfully adopted climate-resilient practices when combined in such packages.

One of the challenges in assessing the evidence is measurement. NRM packages have multiple components and include a complex set of practices making it difficult to design survey questions that accurately measure adoption rates. To address the challenge, we conducted a qualitative study of a package of agronomic practices in Vietnam called “1Must Do, 5 Reductions” (1M5R), to design a good survey module to integrate in the forthcoming VLHSS (Vietnam Household Living Standards Survey).  Under 1M5R the one “must do” is to use certified seeds, and the five prescribed reductions are on seed rates, pesticide use, fertilizer inputs, water use, and postharvest losses. The motivation here is that we felt that simply asking farmers if they adopted 1M5R would result in noisy (very possibly biased) data without telling us anything about partial adoption of the components of the package.

The qualitative study was conducted in two phases. In Phase 1, two experienced interviewers conducted 45 open-ended semi-structured interviews with farmers in the upstream area of the Mekong Delta. The respondents included both male and female farmers, as well as farmers who had received training on 1M5R and those who had not. All interviews were recorded (after seeking consent), transcribed into Vietnamese and translated into English. The transcripts were the basis of detailed coding of data to inform the development of close-ended survey questions.