we improved reliability by having a local vezo team undertake all surveys, with standard explanations to start the interviews and a comment period at the conclusion. Locals with extensive training and experience in conducting socio-economic surveys conducted all surveys in Vezo. The survey team leader was fluent in both Vezo and English, and the team's efforts were supervised and results were quality controlled daily by the PI and/or a research assistant. Additional training and piloting of both instruments in two villages occurred prior to roll out. The HHS' pilot included 42 households. while the pilot interviewed 24 fishers: both surveys changed based on the pilot. The pilot results were also the inputs for the power analysis that set the required sample sizes. Quality control constituted double entry in Excel with inconsistency procedures checks.