Video interview 12 days to presidential: a well -known sociologist challenges the myth of manipulation of the electorate through polls. And then why are the results so different?

On Friday, May 2, the vote in the diaspora for presidential elections begins. Sociologist Cătălin Stoica, former director for eight years of one of the most important polling institutes in the 2000s-course, explains, in an interview for the Hotnews public why certain polls are not confirmed, from where the mistakes can occur and who influences the incorrect data.

The sample, the interview method, the period in which it was carried out, the questions and reputation of the research institute are elements according to which the Romanians could distinguish a misleading one that reflects the voting intention, says the sociologist Cătălin Stoica, in an interview given hotnews.ro.

However, it explains that the errors in the polls can occur from several causes, including the wrong design of the sample, the formulation of the questions, the unprofessionalism of the polling operator or even the misleading answers.

Even when they are wrong, the figures do not have a significant effect for people, says Cătălin Stoica. “If we talk about an attempt to handle surveys, it is extremely limited to the world of potential party sponsors,” he says. However, Cătălin Stoica argues that the polls will show more precisely the voting intention of the Romanians as the voting date of the elections are approaching.

Cătălin Stoica also talks about the most recent volume published at Humanitas Publishing House, “the second round that was”, in which he analyzes the political events in Romania after the annulment of the presidential elections.

Cătălin Stoica is a university professor at the National School of Political and Administrative Studies, is a doctor in sociology of Stanford University and has two master's degree in Stanford and the Central-European University. Between 2007 and 2015 he was general director of the Center for Urban and Regional Sociology (course).

https://www.youtube.com/watch?v=bsa7Oecqu2k

“Interested politicians appear that modify the data to mobilize their party asset, for internal struggles, to impress various sponsors ”

HotNews.ro: Many Romanians say we should not trust the polls in this campaign, because these researches have tricked us last year by ignoring Călin Georgescu. How justice are they?
Cătălin Stoica: There is, without a doubt, a distrust in the polls that I can deny. We have, on the one hand, in the background, some expectations of the majority of the population who wants us to measure the comma and make absolutely accurate predictions. Or, first of all, polls are a form of scientific research, which operates within probabilities. We have no certainties. Therefore, at a survey with a volume of 1,067 people there is a margin of 3%error, at a level of 95%. This means that if we discover that the X party or the X candidate has 30% in the 30% survey, that would mean that in 95% of all the polls we could do on different samples, the score of the party would be between 27% and 33%. However, we admit that in 5 out of 100 surveys the real score will be beyond. So it's a probabilistic knowledge. We can be wrong. Secondly, we work with people, we do not measure with the strap. This is why, especially in the polls on the outskirts, there were differences between exit polls and the official results in the country. Therefore, the “atoms” that we study, sociologists, have conscience, have self-representation, can change their behavior, can tell you that they do something other than in reality. We are neither the police nor the prosecutor, nor the DNA. We need to record what the respondent tells us.

– With these limitations that the survey operator has, you claim that the final result is as close as possible. But sometimes it's not. Why isn't it?
– The sources of errors can be multiple. On the one hand, some people are not sincere, but not because they are Machiavelici, but because they do not want to answer. But there may be errors that are committed by the researchers. Someone who is a person who has no experience in designing a sample, for example. You can have effects induced by the interview operator itself. The surveys can also be mistaken for the stupidity of the researcher who knew how to ask a correct methodological question.

– Last year, there were polls that placed Nicolae Ciuca on the 2nd place. How do these estimates appear so far from reality?
– As a rule, it is assumed that, as we approach the date of elections, the polls should get closer to reality. Mircea Geoană, for example, when she announced her candidacy, was on the 2nd place or 3rd place. After the PSD set her candidate, Marcel Ciolacu, the party aligned and Mircea Geoană decreased in polls.

There are undoubtedly surveys that circulate “on sources”. If a journalist receives such a survey “on sources”, he should get his hands on the phone, find out the data of the public opinion polling company, check if, indeed, the survey exists. Because in the political parties there are people interested in vehicles various figures who benefit one or the other of the camps. For example, I was hit by polls attributed, for credibility, to the research institute we were working on, which were circulating “on sources”, but which were not made by us.

And there is something else: let's not forget that the political parties have certain sponsors, that is, businessmen who are going to ask for financial support, that money is needed in the campaigns. Well, politicians can use such surveys-which may not exist, but be some graphs made in excelled by a student-to show the sponsors. “Here is our candidate! Do you want to contribute?” So politicians appear who are interested, change their data in their favor, either to mobilize their party asset, for internal struggles, or to impress various sponsors. If we talk about an attempt to handle surveys, it is extremely limited to the world of potential sponsors.

“An influence exists ”can a survey overthrow the real ranking?

– Can the electorate also handle the surveys and, in particular, such figures obtained “on sources”?
– I do not think it is possible to manipulate the electorate by publishing the results of some polls. Let's talk about influence. No one says that an influence does not exist. But there is the question of the extent to which this influence is one of a nature to radically overturn a situation. In order for this to happen, it should – hypothetically – have a situation as it was in the time of Ceausescu, in which there was only a television station broadcast, in which the newspapers were controlled, so that a published survey would reach a significant number of individuals.

Let's say that the USR voter carries on certain news sites and looks at certain television stations. The probability that a USR voter will read about the results of an opinion poll that would give the PSD candidate for an astronomical score. That we use different information channels, on the principle of biased confirmation. That is, we no longer have that situation in which we have a single communication channel and everyone finds only about a certain survey that has certain figures and acts accordingly. So we are not talking about significant overturns, I cannot have a 35% candidate and one of 20% and, following the repeated publication of some polls, the ranking will change.

– Are there, however, effects for the public of opinion polls?
– There are two types of effects. Following the publication of the results of some opinion polls there is an influence. Some people, perhaps undeclared, informing the results of opinion polls, can raly around the winners. This is called “Bandwagon Effect”. Other people can be attracted to the one who is in second place. This is “The Underdog Effect” and is raised. There are such effects. But this situation has been studied and those who are raised around the winner are canceled with those who go to the loser. A significant influence involves tens, hundreds of thousands of voters who would vote because they saw in the survey.

How can “a sample of 30,000 respondents” mistake

– Does it matter that the sample is as big as possible?
– At the level of the common sense, it seems a cheap trick: can you say how millions of people will vote if you interview only 1,100 people? Yes, you can say. There is a statistical theory. Not the more people you get closer to reality, but depending on how those people are chosen. There is a methodology of opinion polls: choosing the sample, interviewing method. It is more expensive to probe face to face at the respondent's home, but you can do it by phone.

I give you an example: it can be a sample of 30,000 respondents, which shows that a party or candidate has some astronomical scores. But the detail is not offered that the 30,000 were selected from the loyal viewers of a certain television station. These are intentional mistakes and I cannot deny that there are these unethical practices. A serious institute does not afford to do the pigs.

However, the institutes to probe the public opinion cannot live in Romania only from opinion polls on political, electoral issues. The Romanian politician is interested in what the world thinks about him especially before the elections. Between the elections it is no longer willing that they spend money for it. Thus, an institute loses its credibility to win other contracts between elections, research on socio-economic issues, marketing research, etc.

“I remained with a great scare ”

– How can the public distinguish between a misleading and a correct one?
-I would trust in a survey that I give as much details as possible, even if I do not understand them all. For example, you must pay attention to the details regarding the technical data, the sample and the interview method. The period of achievement is important, as well as the distance to the elections. Then, he can try to compare the results of a survey with the results of other surveys, made in the same period, with comparable samples. It is also important the formulation of questions – it can be a semi -closed question, with answer variants, or an open one, without answer variants. I would also look at the reputation of the Research Institute.

– You have recently published the volume “Tour two that was. You say, among other things, that politicians did not understand the lesson of this moment. What did you stay with after this period?
– I remained first of all with a great scare at that time. About how things were foreshadowed in the 2nd round, especially taking into account the data of some polls attesting that things could take an unfortunate turn for the European, Euroatlantic Road.

Yes, the political parties did not understand anything about the message of disappointment, despair and frustration of an important segment of the population with voting rights. Frustration in relation to a false that exists in the Romanian society. It was spoken of a false between pro-European and philorruşi. The data to which we had access to some opinion polls that were published starting with 2022 attests that a false type between pro-European and philosophers has no empirical foundation. Because most of the people with voting rights in Romania are pro-Nato, anti-Russia and enthusiasts of the European Union. There is, in fact, a false between the privileged political class and the majority of the population. People are disappointed because they do not feel represented.