Perhaps the person collecting the data is a student learning about how to design measurement instruments, or piloting a measurement instrument. It's hard to learn how to do this, or to test it, without doing it, and going to the effort and expense of getting a random sample to find that the questionnaire was flawed would be a waste of time and resources. Might as well try it out on a group that's easy to get hold of, and iron out the flaws that way. I encourage my students to use their Facebook friends as samples, because you can tell them that designing surveys is hard until you're blue in the face - they never really get that it's hard until they do it, and then try to analyze their results.
Jeremy