Graphics have been published on social media claiming to represent public opinion measurements on the eve of the June 7 parliamentary elections.
Two of them were published on Facebook and one on TikTok.
The three graphics include the acronyms of only four or five parties. In two of them, the Democratic Party of Kosovo is shown first, while in another, the Vetevendosje Movement is shown first.
The truth?
Two of the graphics provide no details at all about the methodology used to conduct a poll. Another falsely cites a well-known organization for measuring public opinion in Kosovo.
What was claimed?
In the first case, the Facebook account “Gjergj Dedaj” published a graphic on May 9 showing the acronyms of the four largest Albanian parties in the recent elections. According to the graphic, the Democratic Party of Kosovo (PDK) comes out the largest with 36 percent, followed by the Democratic League of Kosovo (LDK) with 29 percent, the Self-Determination Movement (LVV) with 26 percent, and the Alliance for the Future of Kosovo (AAK) with 9 percent.
Post of the account in question:
A Google image search has led to several other posts with the same image. It was posted on TikTok and Facebook, posts that have generated hundreds of comments (see here and here). One took the publication for granted, while the other asked if followers believed the "poll".
In the second case, the "lajme.politika.sh" account on TikTok has published an image that claims to be a measurement of public opinion on the question "If parliamentary elections were held next Sunday, which political party would you vote for?? "
The image shows the ranking of the four largest parties, with PDK in first place with 38.4 percent, LVV in second place with 35.7%, LDK in third place with 17.6% and AAK in fourth place with 6.3%. It is noted that PSD does not cross the threshold and that it has 2.0% in the measurement.
In addition to spelling errors, the survey results show the dates June 1-10 as the dates for conducting the survey. The publication was made on May 25. It also said that the measurement was conducted through telephone interviews and there were 1200 respondents. It also said that the margin of error is 2.8%. At the end, the address of the company “UBO Consulting” is also given, but it does not match the one provided by the company itself on its website and social media accounts.
TikTok has labeled the publication as containing content generated with artificial intelligence.
While a Google search with the words "UBO Consulting poll for the June 7 elections"doesn't lead to any real media releases in the last week."
The third case is on Facebook. In the group “Lëvizja VETËVENDOSJE Velbert/Düsseldorf 🇦🇱🇩🇪” the account “Vesel Berisha” has published a graph which resembles the presentation of public opinion measurements. According to the publication, LVV has 69%, LDK 13, DPK 14 and AAK 4.
The content has generated over 2,600 reactions and more than 500 comments.
In the claimed measurements, the sample reaches 100 percent with no more than five entities, although in the June 7 elections the ballot contains 21 political entities. In the last elections (December 28th) the four largest parties together received just over 90 percent and almost 10 percent was divided among other parties, mainly community parties.
What are surveys?
A survey is a preliminary research to identify a situation, opinions or attitudes about something specific. In this statistical method, results are drawn based on the study of a part of the population, called a sample.
Based on an article by the Pew Research Institute (a research institute for fact-based surveys and studies), it clarifies what information should be sought about how a survey was conducted.
Who conducted the survey and who sponsored it?
Who was interviewed?
How was the survey conducted?
How were people selected for the survey?
When was the survey conducted?
How many people were interviewed?
Was the survey weighted? If so, how?
Do the survey questions seem neutral and fair?
These are missing from both publications, while the one that claims to have mentioned them has data inconsistencies and TikTok has confirmed that it is AI-generated content.
Hibrid.info also searched the Google search engine with the keywords "Kosovo survey”, limiting the results to publications from the last 30 days (see here). Apart from a few publications of unreliable surveys and fact-checkers’ treatment of these claims, the results of this search did not turn up any credible articles that would confirm the existence of this or any other alleged survey.
Fake polls Several election polls have now been circulating. Hibrid.info has identified and addressed hundreds of fake publications from various publishers on social media, which have falsely presented themselves as credible polls (see some here, here and here).
Analyze
Three graphics are circulating on Facebook and TikTok that claim to present election polls for June 7.
Verification shows that they lack a clear methodology, contain inconsistencies, and in one case have been labeled by TikTok as AI-generated content.
Based on the evaluation methodology, hibrid.info assesses the content in question as "fake news".
rationale
The “Fake News” rating is given to informational content that is entirely fabricated or that contains factually incorrect claims or information. Content that is rated as fake news can be defined as having been created and distributed with the intent to misinform the public, i.e., to present a claim that is completely false as fact.