Opinion polls and polls are an crucial component of any election campaign. That is why Krzysztof Stanowski's words on "buyed polls" are all the more disturbing. Is this likely and is it possible to manipulate, for example, a ranking of assurance in politicians? We ask Tomasz Sienietycz, the owner of the investigation Partner investigation company.
Konrad Bagiński: Krzysztof Stanowski at the debate “Super Express” suggested (while talking to Adrian Zandberg) that a candidate outside the mainstream did not stand a chance in the election. And he said, "The probes are bought." What do you think?
Tomasz Sienietycz: I heard this message and I feel that Mr. Stanowski utilized a shortcut. On the another hand, with all this buying of polls... it's a message that's a small unprovable. That's one. Secondly, I don't feel offended here due to the fact that my polls don't buy. I'm speaking ridiculously, trying to get a small bit into Mr. Stanowski's rhetoric. There is no denying that polls can of course play a function in creating candidates.
I remember, until I announced the list of candidates, I chose any names. all 4th I did a survey that was about divided into the PiS electorate and the electorate of today's ruling coalition. I utilized to give a fewer names and see who people would vote for. The candidate most poorly elected by respondents fell out and I selected another for another wave of investigation after 3 months. For example, it was noted that at 1 point the erstwhile talker of the Sejm Elizabeth Witek began to fall. It worked perfectly with her lesser vulnerability in the media. Your marshal disappeared from the tv screens, and it immediately affected the ratings.
So we can safely say that the exhibition of candidates in the media, and it helps the biggest, very much. After all, if we show the laundry powder thousands of times, the more people will buy it.
A fewer days ago, I spoke to Michał Tragar from the Center for Civic Education, which he releases before the election of “Electoral Lantern”. It turned out to me and many of my friends that the Adrian Zandberg mentioned was highly advanced on my list of consensus. So why is it so low in polls? It turns out that a match of views is not the most crucial thing, but the image of this candidate is inactive important. He, I believe, may be considered a small antipathic.
Of course, it's very influential. I just don't see Mr. Zandberg as antipathic. On the contrary, as a man who is affirmative towards people. I'd alternatively fishy another candidates. But to him I would have a completely different remark: see that he only appears in the media erstwhile there is an election coming. Besides, it's quiet about him.
But speaking of these words by Mr Stanowski, I do not say that any polls would affect Adrian Zandberg's position. It's more like his absence from the media is affecting the polls. We had the same thing with Kamala Harris and the American elections. It was hard to believe he had a chance with Donald Trump. After 2 years of being vice president, she was simply hidden in the back seat by Biden. Why?
It was written that Mrs. Kamela Harris was chaotic and was arguing people. If she stopped appearing in the media, how could she compete with Trump? Whether he said something wise or more controversial, he was inactive loud and constantly shown.
Do the parties actually commission specified ‘internal’ studies? due to the fact that we have a couple of polls a week now, so what's the point of repeating the same research?
I've never been commissioned by anyone, and I'm just acting on a small different rules than another companies that print polls. I do not cooperate with any political party, I am not a associate of any party, and so I do not know what interior polls are conducted by political groups, who are doing them and according to what methodology.
And if a organization had asked for this, would you have done specified a poll unpublished, for the interior usage of a party?
Well, that kind of depends on what it is, because, for example, Mr. Maciak doesn't necessarily have the limits of absurdity. But in another cases it would be a purely commercial relationship. If anyone comes forward and I know I can do it with a clear conscience, I will. There's nothing incorrect with that.
Of course, due to the fact that I besides wonder if it makes sense to do specified polls at all these days, due to the fact that as journalists we see that polling is most frequently ordered by the media, who are not truly curious in conquering or conquering candidates, but to make this message credible.
This is different due to the fact that any want to be credible, but any represent the interests of certain political groups. In the old days, the media represented the interests of the only leading party. Not 1 now, but a few. There's definitely social and political relations, possibly business.
To be honest, I don't truly believe in this completely impartial message. Not in all the media, of course, but in part.
Can you theoretically manipulate a poll? The X organization may want her candidate to be higher in the polls. That's an extra plus for him, people want to be on the winning side.
Absolutely. individual erstwhile wanted me to just run the questionnaire he brought me. I refused to do specified a study. Signing under the results would be an act of suicide for a investigation firm, even a tiny 1 like mine, or possibly above all.
Can the poll be manipulated?
Yes, through appropriate test settings as such. But you can besides manipulate the explanation of results.
If I wanted to order a poll, and to make it mine, I would have asked a twelve questions, most of which would direct the subject to the fact that any candidate is bad and any is better. And then I'd ask a fewer questions about who the individual would like to vote for. I think this is how I could conquer my candidate and weaken my competition.
Yeah, we could do that. You can ask questions biasedly. The order of questions is very important. This is the second possible form of manipulation. And the biggest field to show off is the selection of the example and the determination of scales, i.e. the selection of specified factors, which should reflect the reality well, are typical of the full population of Poland. Another issue is the appropriate wording and setting of the cafeteria of answers.
What's that?
Order and method of formulation of the answer to choose. They can be formulated and configured to manipulate the results.
But I will stress: I do not want to enter Mr Stanowski's rhetoric that the polls are "buyed". possibly they are. I'm not sure. I could besides hazard a claim (these are my guesses, I have no evidence of it) that there are factions in the staffs of various parties that poll to play something for their people. For example, push your candidate through, usage different tools to increase his chances in the eyes of the party's leadership. Which does not always should be good for these political groups.
There's 1 more thing I'm wondering. Political polls in Poland have not been made since today, and I do not callback a large number of cases where they powerfully dispel the actual results of the elections. Wouldn't that indicate the reliability of most polls?
Such a turnaround happens, but it is not or does not should be the consequence of any conscious manipulation. You mentioned earlier about the alleged polling average. I don't truly believe in this average due to the fact that it's crucial to averaging these results. Let me give you an example. As studies were carried out before the last parliamentary elections, most investigation companies showed a group of about 8% of respondents who declared to participate in the elections, but did not yet know who they would vote for. All of this group ‘distributed’ proportionally between the parties.
I argued with that due to the fact that it's never like these people are going to share evenly. My company besides conducted investigation before these elections. 3 groups hit a point, 1 got a small scattered, although within the mistake limits, and in 1 batch the mistake was statistically significant. Why? due to the fact that this group of 8 percent people who declared that they did not know who they would support voted almost entirely for the KO and the 3rd Road.
One could conclude that people who said they didn't know who they were going to vote for were going to vote for opposition against those who then ruled. I'm the only 1 with no proof of that. On this principle, I could say now – if the Coalition now rules on October 15, the undecided will vote for Nawrocki. But that would be a conjecture that is not acceptable to the researcher, and that would not be justified in any way. It's not to be counted, the behaviour of any voters is simply part of a riddle that slips distant from socio-demographic rules.
What could origin the polls to dispel the results now?
Election committees very frequently confuse elections to parliament with presidential elections. And let me remind you that in the parliamentary elections we always have 1 round, and here we practically always have two. That means the votes and preferences are different. The people who voted for a candidate who did not enter the second circular then have a dilemma. Whether to go to the election or to cast empty votes, or to vote for a candidate that my dreamer would theoretically cast his vote for. It's quite a few reasoning again.
We frequently ignore the fact that people who live abroad and have the right to vote can play a peculiar function in the presidential elections. In 2020 it was over half a million people, in 2023 it was over 600,000. This is simply a group that is not measured in any studies and can turn upside down not only polls, but besides election results. The participation of these individuals may increase the survey mistake by more than 2 percent points, which is almost as much as the statistical mistake of this kind of research.
