Real trade-offs
- The more personalized the picture, the less amenable to overview or generalization.
Open-ended text replies cannot be scored to identify priorities.
- The thicker (more personalized) the survey, the more time-intensive the
data-collection. Anthropologists go for dense, qualitative accounts, but they
typically spend about 2-person-years per village
- Either invest at the outset (in terms of questionnaire design) or you will
have to spend more time interpreting the collected information.
- Detailed surveys can discourage all but chronic procrastinators from responding.
Don't make everyone answer every question. Instead, develop multiple separate
instruments, and then aggregate for a full picture.
- Naturalistic settings reduce control over variables. Combine real world data with more tightly controlled research that may lack full correspondence.