It all started in Chilesoc. Jose Manuel Ferreiro sending an email to the list noting that over The Counter had published a column Sabrobsky to the purpose of the surveys-the column in this link-, and that ‘on the other hand, I think that all those who, directly or indirectly, we have worked with surveys (such as survey or its design and/or analysis) we have had several of the questions that here arise (and probably have activated mechanisms to set aside and continue to work)’. I wrote a reply, and for that I do not forget, I write also on this blog, and there is-such which was sent to Chilesoc – below:
I actually disagree with the basic assumption of the column: it Is said that the surveys do not represent the real opinion of the people, and that this is a problem. But it is evident that the surveys may not represent the real opinion of the people (i.and the real opinion about a topic always goes beyond the four or five alternatives that are presented to him, it is clear that it is not always clear about what they believe in etc). Polls work because the society does not work on the real opinion of the people (for example, many times the opinion desirable to have more implications in real society than the supposedly real), and the opinion limited that respond on a survey is often the opinion is limited to that actually used in social life.
In fact, in my opinion, most of the problems with opinion polls is a problem of interpretation rather than a question. There is a famous case of the effects of context in surveys. To the gringos in the years 50 were asked if they believed if a reporter soviet had the right to go to EE.USA and to return to his country to say the things that I would like. The majority was opposed. Now, if you asked then ask them the same question but inverted (on a reporter gringo in the Soviet Union) the numbers were changing. The question was repeated in the ’70’s, with the result that the effect was maintained but much more diluted. Now, all of that is informative and tells us things about the american society. Imagine the question is more biased than you possible think, and even those answers will be informative of something.
If one remembers how limited is the answer to the questions of a survey (selecting an alternative entity which I presented, in that specific context of the questionnaire; there are giving me the advice on the topic), then they do not cease to be informative and useful; treat it as another thing causes a series of problems
The surveys generated reality and not only the described is well known. But all the actions in the society construct of society. It is not something particular of the surveys.
BTW, the initial paradox is false. The following statement might fall under a paradox: ‘I, who am chilean, I affirm that no government in any occasion says things in front of’. Now the following statement is not equivalent to the previous: ‘I, who am a chilean, I can say that it is typical of the chileans not to say things in front’. The subject may well think that doesn’t do the typical things of the chileans, or at least that it does not always (including the answer to the question). I have already said that the main problem of the survey was the interpretation and there he is again