Wrote: Etrit Rexhepi
On the eve of the local elections on October 12 in Kosovo, misinformation related to this process has been identified during the election campaign. A prominent form of inaccurate content circulating in the digital space during this period are fake polls that claim to show support for candidates in several municipalities of the country. Since the beginning of the campaign, hibrid.info has encountered, treated and evaluated seven polls published in six articles. The content was published by sources outside the media on social networks (TikTok and Facebook), with contradictory and often visually manipulated figures.
What is a survey?
Surveys are a scientific method for measuring public opinions and attitudes. They usually involve a small group of people chosen representatively from a larger population, who are asked the same questions. Through this process, their responses can provide a snapshot of the opinions and experiences of the larger group as a whole. Surveys are used to understand attitudes on various issues, to assess citizens’ perceptions of policies and candidates, or to gather information about the market and services. They also help provide an independent source of information about the public, allowing the voices of different groups in the population to be heard in a fair and balanced way.
The first fake poll published and addressed (September 15) was about the municipality of Kaçanik. The account “kaqaniku.press” published a table where Heset Malsiu was leading as an independent candidate with 36%, followed by Jeton Raka and Sabedin Vishi. This publication contained spelling errors, did not cite credible sources and was not found in any serious media, being proven to be fake (see here).
Three days later (September 18), another piece of content published by the TikTok account “deqanisot” claimed that over 16 people had participated in a survey for candidates in the municipality of Deçan, although the caption stated that only 1 people had participated. These contradictions and the lack of transparency about the author of the survey made this publication completely unfounded (see here).
The third case of the publication of a fake poll targeted the mayoral elections in Podujeva. A material attributed to the “Polis Institute” was distributed on TikTok, according to which Shpejtim Bulliqi, the candidate from the Vetevendosje Movement (LVV), won with 57.1%, followed by Ekrem Hyseni of the Democratic League of Kosovo (LDK) with 27.9% and Isak Shabani from the Democratic Party of Kosovo (PDK) with 11.4%. Verification showed that this poll does not exist, while the graphics were manipulated, as the percentages did not match the visual representation (see here).
Gjakova was one of the municipalities with the largest number of inhabitants among the municipalities where fake polls were identified. The page named “Universal” published a post, referring to an “SDSG institute” as the author of a poll that showed Ardian Gjini, the candidate of the Alliance for the Future of Kosovo (AAK), as the winner in Gjakova. The research showed that SDSG exists as an organization in the US, but deals with sustainable energy and not with public opinion measurement (see here).
Two posts claiming that the latest polls showed Agim Aliu of PDK in Ferizaj and Imri Ahmeti of LDK in Lipjan were the winners without a runoff were similarly treated. The content was shared by a page called “Kosova e Re” on Facebook. None of this data appears in credible sources, while any information on the methodology and samples used was missing (see here).
The last alleged poll, which was covered by hibrid.info, was a publication (September 29) on Facebook, by the page “Anamorava Lajm” which claimed to present support for the candidates of Viti, with the participation of 1305 citizens in 19 localities. According to the post, Arsim Ademi (LVV) was leading with 39.46%, followed by Sokol Haliti (LDK) with 37.55%, Bekim Azizi (PDK) with 15.33% and Labinot Bislimi (AAK) with 7.66%. However, the page is not a professional institution for polls, while there is a lack of clear data on the methodology, time of implementation and the measurement period. Furthermore, previous posts from the same page contained contradictory data on the number of localities and participants. Based on these elements, this poll cannot be considered credible and hibrid.info assesses it as “Disinformation” (see here).
These alleged polls share a common element of a lack of credible sources, transparency, and methodology. Many of them contain contradictory figures, visual manipulations, or fabricated references to institutes that do not conduct polls.
Based on the work methodology, in addition to the last survey discussed, hibrid.info has assessed these publications as "Fake News".
Fake polls as a form of opinion manipulation on social media were also detected during the period before and during the election campaign of the parliamentary elections of February 9 this year. During this period, hibrid.info had verified 36 fake polls. Before the start of the election campaign, hibrid.info had warned readers about the dangers of incorrect information that may be published during this time, also mentioning fake polls as one form of misinformation content.