| Although true random samples are necessary for the successful completion of some research projects, studies using non-random samples can also be valuable. Research using non-random samples can be improved, however, and our recommendation is for a more careful description of them, to appear in methodology sections of published research. This practice has three purposes: (1) It will allow other researchers to choose their samples with more precision. (2) Careful description allows others to combine various sets of data into a functional relation. (3) It allows others to evaluate the plausible rival hypotheses that may threaten a study's validity. Suggestions for the content of a sample description are given. |