The first thing to keep in mind is that some reports published weeks or days earlier about the intent of the population to vote have long been part of the campaign of certain political groups so that they "motivate" their electorate to go to vote , that is to say, sometimes it does not reflect so much the reality of "the street" but the image that one wants to offer of victorious party, for example.
On other occasions it is, on the contrary, showing result statistics without a clear winner what makes the undecided can change their vote with the thought that they reinforce a certain party or prevent between another.
And then there is the "real" statistic that reflects what respondents respond down the street to the interviewers who, from weeks before, are asking everyone who crosses. These surveys, if they are well-made, must pick up a number large enough to be meaningful, and thus representative of the population, that is, you cannot ask ten and pretend with it that the population of a country thinks that way.
Assuming that the survey has been successful, with a sufficient number of interviewees to be representative and maintaining the percentages corresponding to the demographic characteristics of the population, assuming that all of the above is Meets, then comes the day of the election and… Surprise, the results have nothing to do with the polls and their predictions, how is that possible?
— Politics on the Couch (@PoliticaDivan) April 4, 2017
This is what he has tried to answer with a research conducted from the Institute of Behavioral Sciences, Radboud University (Netherlands) whose results have been published in the scientific journal Frontiers in Psychology.
Three studies were conducted to validate a new measuring instrument called Self-Approach-Avoidance Task (Self-AAT) to determine the implicit attitude and compare whether it offers the same results as the explicit attitude.
In other words, it is intended to see if it is possible to know the person's "true" intentions, beyond what he or she can declare in a survey.
This was compared to the results obtained with standardized instruments such as the Rosenberg Self-Esteem Scale (SSOLR) for the evaluation of the explicit measures and the Implicit Association Task (IAT) for the evaluation of the implicit measures.
Among the advantages of this new instrument is the one that is faster to apply and does not require so many items.
The results show how to respond to family stimuli before strangers.
It also validates the results of the comparison between this new instrument and the standardized ones.
Despite the foregoing, it continues to find significant differences between the data offered by the participants in terms of explicit and implicit information.
Therefore, in spite of the novelty in the design of this test, capable of detecting the "true" intentions, it still cannot be used for the scope of the opinion study or for a poll of intention of vote, because it requires that each voter feels during a Det Erminado time on the computer to perform the task.
Therefore, until the procedure is improved, "relying" will have to be followed from the results of the surveys carried out on the street, taking into account the limitations that it entails.