When interviewing: how many is enough?

Researchers need to know what is an appropriate sample size for interview work, but how does one decide upon an acceptable number of people to interview'? This question is not relevant to case study work where one would typically interview every member of a case, or in situations where it is both desirable and feasible to interview all target population members. However, in much of qualitative and mixed-methods research and evaluation, the researcher can only reasonably interview a subset of the target population. How big or small should that subset be? This paper provides a brief explanation of why the concept of generalization is inappropriate with respect to the findings from qualitative interviewing, what wording to use in place of generalization, and how one should decide on sample size for interviews.