Ah, I see the disconnect now. I take you to be saying that the result can be used for some purpose, and with that I agree (clearly it has been useful in misleading certain uncritical readers to advance a political ideology).
What I (and, correct me if I’m wrong, Phyllo) mean is that the truth about the world for which this result provides support does not have important consequences; it is through misuse and misunderstanding that it is being put to use, but that misuse or misunderstanding requires substituting false premises (e.g. “this survey tells us about people’s actual willingness to go to war for their country”) or making bad deductions (e.g. “therefore people must be sick of liberalism, feminism, and multiculturalism”) from the premises for which this provides empirical evidence. The well-supported premises and the deductions that can be made using them are boring.
I recommend Kahneman’s Thinking Fast and Slow for discussion of why and how much this difference matters.
Autsider’s views and reasons for using the survey map do not make the map hold water as useful information. The survey map is incomplete, lacking details as to its nature and whatever the OP implies that the survey means is not presented on the map itself.
That’s a criticism against surveys in general, Carleas. Of course some people might claim they would go to war but then decide not to, and others might claim they wouldn’t go to war and still go.
I didn’t make this deduction, if that’s what you mean. I think there is a variety of reasons that people wouldn’t fight for their country, some of them more justified than others.
Yes, it is. Surveys as a method of getting information about a population are always subject to this kind of criticism. But not all questions have the same problem, or to the same degree. If you ask people for their height, you’ll get a lower rate of error than if you ask them how likely they are to vote in the next election. Responses to certain types of questions are less reliable than responses to others, because people have more or less reason or ability to lie to pollsters, and they be more or less well informed about the answer to the question.
To the latter point, a person’s predictions about what they will do in the future are pretty unreliable, especially when they are predicting what they would do in an unlikely hypothetical that they have never experienced before. On questions like those, people just aren’t that well informed. As Kahneman notes in his book, in response to a question like the one presented in this survey, people are likely to (unconsciously) substitute a much easier question, something more like “do you like war?” or “is the army a good career?”.
And, as Wendy and others have pointed out, the framing of the question matters. The way a question is worded, and questions that lead into it, can affect the outcome of the answer substantially. And again, that effect is larger for less definite questions like this one, where people are very likely to mentally substitute a more intuitive question.
So you’re right, it’s a general criticism of surveys, but this specific question is one for which the criticism should carry greater weight.