This is a comment on the whole series, not just the last one. At the
beginning of my career, I spent about 5 years in public opinion
research, some in a small business of my own, most of it working in the
Division of Program Surveys (BAE) as a study director. Since then, I
have done occasional jobs, including one collaboration with Daniel
Yankelovich. Meanwhile, there has been a good deal of technical advance
in the field, so that today I would not presume that I am competent to
supervise a doctoral dissertation like this one. I feel pretty
confident in saying, however, that the poor student is putting a good
deal of effort into a futile endeavor, from which he will not have
any valid basis to draw conclusions or answer any questions with which
he began.

It is a bit sad to see how earnest and probably quite intelligent
people are still plunging into a rather technical field assuming that
they can conduct a valid survey simply by using common sense. I don't
fault the graduate student in this instance as much as the supervising
faculty member, who ought to have known that s/he was not competent to
guide a student in a dissertation of this kind.

The comments have hit on quite a few of the many flaws in this survey.
I could add several others but am not sure it would accomplish
anything. The main point is: don't assume that just because survey
research is social science (which a lot of engineers sneer at as
lacking in any rigor), anyone is qualified to do a survey, or even to
evaluate it.