Polls and opinion are reliable?

Author: personal Andre Augusto Vallado Baptist


With the approaching elections for President, Governor, Senator, MPs, opinion polls have become part and every time more news, conversations and discussions daily. Today, according to data from Tse (Supreme Electoral Tribunal.) , There are about 146 million voters across the country. The polls will go beyond voting intentions. The views and research trends. It is precisely at this time that many people, including some of the candidates they disagree with or place in doubt the validity of this research. But of one thing we are certain: the search results of its influence. – conduct of candidates and their teams, as well as voters.

What happens is that usually the research get it right the results. This fact have already been studied at the University of Houston in United States of America and there was a indicator of accuracy of about 90%. In Brazil, where the research methods are quite similar to Americans, the results should be similar. Now the question that always arises is: how can you trust these research usually I haven’t even been interviewed and I don’t even know the man? To clarify this doubt, a simple and educational way, it is necessary to know how to search of this size.

In the first place in the institutes that perform research to select a sample of people is representative. This model aims to simulate the total voters in their characteristics. For example, in recognition of the socio-economic levels of the sex and age of the surveyed persons, so that these data are diverse and similar to those of the real population. It is clear that any Management Institute search only within the Union of metal workers within the Afro Brazil Museum (Federation of industries of the state of São Paulo). Will identify small groups of people in different regions of the country, and respect for diversity required. Some institutes also choose to interview a variety of people with respect for standards development in places of large flow such as the Centers of large cities, for example, a Paulista in the city of São Paulo.

After completion of data collection, these are collected and processed statistically. According to the institutes depends Datafolha, the Ipsos and Ibope, interviewed about 2 thousand people. This is a sample only of 2 thousand people, a far cry from the 146 million voters, but it can be very large, though not quite accurate in the result, that there is a margin of error which can vary from 2% to 4%, “more or less”. This explains why you probably haven’t attended the search. If all these cares are taken seriously by the research institutes will be reliable. We can’t ever forget that the statistics show a picture of that moment which they were made, if the new reality happen, for example, in the context of the tract of data, certainly the result may not show the truth. In the political scenario in which new facts related can occur in every moment, this risk will always be present.

Another important detail: the poll is not Research. The polls from a site on the web, for example, do not use any scientific methodology. In fact, the managers of the ballot boxes did not have any type of control on the public common. Therefore, the opinion polls do not have the same accuracy and credibility of research.

High school students from Santa Maria College already as a tradition in the course of the year, workshops and studies to enhance their knowledge with a view to the elasticity of the exam. In the workshop of the mathematics they have the opportunity to review their knowledge of Statistics examples situations such as those described in this issue, was the purpose of the study and a brief discussion, he revealed the enormous importance of mathematics in the formation of a citizen’s most important activity.

Seja o primeiro a comentar

Faça um comentário

Seu e-mail não será publicado.