It is important that a native English-speaking person conduct the survey.  Often times different cultures view the same question differently.  If an interpreter is used it is suggested that he or she be asked to repeat the questions exactly as asked.  If the villager asks any clarification questions, the questions should be repeated back to the native English-speaker who can then respond back through the interpreter.  This will prevent side discussions from taking place between the interpreter and the villager.  It will also and most importantly give the person conducting the survey an insight into what the villager is thinking. Try to conduct the survey with the woman who is head of the house as she is the one who mostly uses the filter.

During the Post-Filter Survey when the filter recipient is demonstrating how to backwash the filter do not assist in any way.  Do not help hold or position the filter or offer any hints.  It is only good data if we know for sure the recipient DOES OR DOES NOT know the proper method to backwash.  After the survey one should feel free to give a backwash refresher demonstration if needed.   

Obtaining good data takes time.  One thing Americans like to do is ask multiple choice questions.  Asking open-ended questions is better but takes more time.  For example: “Where do you get your water?” vs “Do you get your water from the stream or the well?” may yield more useful data.  After asking a question then pause and wait silently for a response.  Continue to ask the same question exactly as asked the first time and wait patiently.  Most times you will get a response that is very accurate and possibly something you did not expect.  Do not let the interpreter interrupt and take charge of the interview.  This happens more times than not and produces skewed data.