Send message to Author

Paper long abstract:

In development projects and interventions, survey work assumes that good data follows a particular track. Enumerators should be trained, should follow the survey wording and should operate under a series of protocols that meet externally set standards. Deviations from this are a sign of poor performance or pathologized as “cooking data” and undermining the overall quality of work. Drawing on a body of ethnographic work studying the lives of enumerators in Kenya and Uganda, we found a number of ways of working that did not conform to “best practice”, but which were adopted by enumerators because they were felt to elicit more truthful, honest answers from survey participants. Softening questions, where the enumerator knows a more reliable answer can be achieved through rephrasing; deciding to offer refreshments at a different point during the interview cycle to the one agreed; probing patiently for responses in a conversational manner, that might mean the enumerator does not achieve their assigned quota. These were all ways in which enumerators showed a reflexive relationship to their work, and a degree of craft. Such practices, informed as they were by inter-personal ethics, self-evaluation, and an interest in getting the right answer, are not captured in framings which assume divergences as incompetence at best, or “cooking data” at worst. Such practices, we found, were also shaped by the desire of enumerators to not “feel like a robot” as they moved from home to home.