“The first lesson you learn as a pollster is that people are stupid,” said Tom Jensen
This quote from a Democratic survey researcher is not surprising. It is simply a sign of a stupid person with no experience in the field.
I virtually never use the word 'stupid' in this blog, because it is a fantasy description of an unknown person very similar to the psychoanalytic descriptions I so often hear of Republican presidents by Democrat psychologists who never met them. I think psycho-analytic descriptions of people you have not met is absurd and rarely used on anyone other than a Republican office holder. 'Stupid' is the same thing.
I call this survey researcher stupid in this case because I have done survey research for nearly 50 years. In my earliest days of doing survey research I conducted many hour-long interviews by myself of a randomly selected population. Since then I have often done personal interviews in order to carry out focus group work or to refine the wording of a questionnaire.
The single most striking quality of randomly selected citizens when interviewed, is the intelligence and coherence of their worldview in virtually every single case. People are not 'stupid.' Of course you have to be honest to say this.
No person I have ever met who did this type of qualitative in-depth interviewing has ever come to a different conclusion about random citizens.
It is with that firm and convincing experience that I call the man referred to in the first paragraph 'stupid'.
What this man is himself mistaking for stupidity is the fact that all questions are probes into the respondents' mental world. The probe can be ridiculous, not understandable or can have many other confusing elements. That is why a good survey researcher pretests his/her questionnaires to make sure that the questions are not totally ambiguous nor will result in contradictory responses. Tom Jensen does bad survey research work.
Over several years, using the same questionnaire for tens of thousands of people, I went back to the raw data to see the range of interpretations of the questions' wording that were possible. In a large sample virtually every interpretation of every word in the questionaire turned out to be on record. Contradictions did occur even with a first-rate questionnaire when used over a large enough sample.
Recent Comments