Experiment Design of a Balanced Eye-Tracking Study for Measuring the Cognitive Impact of the Stimulus Word “Disinformation”
Published Thursday 8 October 2026 at 09:36
The article by Ruslana Margova and Irina Temnikova, titled “Experiment Design of a Balanced Eye-Tracking Study for Measuring the Cognitive Impact of the Stimulus Word ‘Disinformation’,” presents the methodological design of a controlled eye-tracking experiment investigating how the word “disinformation” affects the way Bulgarian readers visually process news content. Importantly, the paper reports the experimental framework rather than experimental results; the actual study is planned as future work.
The central research question is whether the explicit presence of the word “disinformation” in Bulgarian news headlines influences readers’ visual attention and eye-movement behaviour. The authors hypothesize that the term may function as an emotional trigger and cognitive anchor, potentially increasing cognitive load and prompting more systematic processing. The planned experiment therefore focuses on indicators such as fixation duration, gaze duration, saccades, and regression frequency.
A key contribution of the study is its balanced experimental design. The researchers created matched pairs of authentic Bulgarian news headlines and leads of up to 250 characters: one version contains the word disinformation, while a minimally modified version removes it without replacing it with a synonym. The planned stimulus set consists of 30 text pairs, and political news is excluded in order to reduce confounding effects related to political attitudes, prior beliefs, and difficulties in establishing factuality.
The experiment is designed for a Tobii Pro Spectrum eye tracker operating at up to 600 Hz in a controlled environment. The initial pilot is planned with 10 participants, followed by a larger sample. Participants will be adults aged 18–79, with recruitment balanced as far as possible according to gender, education, prior knowledge, and other background characteristics. Participants will not initially be told that the study specifically focuses on the word disinformation, helping to reduce changes in reading behaviour caused by awareness of the experiment’s purpose.
An important methodological aspect is the attempt to separate the effect of the stimulus word from other factors that influence reading. The design therefore controls for text length, readability, linguistic complexity, prior knowledge, topic interest, previous exposure to the news item, reading and media literacy, and participant fatigue. After reading, participants will also complete a questionnaire assessing perceived factuality, prior knowledge, interest, and previous exposure to each news story.
The paper is particularly relevant for Bulgarian as a low-resource language. The authors note that most previous eye-tracking research on misinformation has focused on English and that, to their knowledge, there is currently no publicly available Bulgarian eye-tracking corpus specifically dedicated to misinformation or news credibility. Bulgarian also presents language-specific characteristics—including Cyrillic script and rich inflectional morphology—that make it problematic to assume that findings from English can be directly transferred to Bulgarian readers.
The authors are explicit about the study’s limitations. Most importantly, simply comparing the presence versus absence of “disinformation” does not fully separate its emotional effect from its lexical properties, such as word length, frequency, and morphological complexity. A future version of the experiment could therefore introduce a third condition using a neutral, structurally comparable word such as information. Other limitations include the relatively small manually collected corpus, possible prior familiarity with some news items, and the exclusion of highly political material.
Overall, the paper’s main contribution is not a finding about how people react to the word disinformation, but a reproducible methodological framework for testing that question experimentally. The authors intend the resulting eye-tracking dataset and accompanying corpus to be released as open resources after the study is completed, potentially supporting future research on misinformation processing, media literacy, and cognitive responses to news in Bulgarian.
Suggested citation
Margova, Ruslana, and Irina Temnikova. 2026. “Experiment Design of a Balanced Eye-Tracking Study for Measuring the Cognitive Impact of the Stimulus Word ‘Disinformation’.” Proceedings of the Seventh International Conference on Computational Linguistics in Bulgaria (CLIB 2026), 296–303.
Article: Full-text PDF
Proceedings: CLIB Conference Proceedings
