Nonresponse and generalizability in academic research

Generalizability of results is an important issue in academic research. We have attempted to identify some of the major issues related to this topic and to propose consequent guidelines for research practices. Any working researcher knows all too well how difficult it can be to gain access to data and get good samples. Fortunately, academic research can make good use of imperfect samples, but it is incumbent on us to be aware of this issue and make our best efforts.