How language shapes research assessment: A comparative survey of Italy and Hungary

This poster presents findings from a comparative study on multilingualism and language bias in research assessment, carried out by CNR (Italy) and HUN-REN, the Hungarian Research Network, within the CoARA Working Group on Multilingualism and Language Biases in Research Assessment.

The study starts from a paradox: global science relies on English as a shared language, yet frameworks such as the Helsinki Initiative and UNESCO's Recommendation on Open Science recognize multilingualism as a core condition for inclusive science. Two comparative surveys ( 910 respondents at CNR and 211 at HUN-REN, using the same instrument, translated and adapted) show that language disadvantage is not an individual matter but a systemic one: it varies by country and, more strikingly, by discipline, with SSH researchers consistently more exposed than STEM colleagues in both national contexts. The effect ripples through grant participation, manuscript rejection rates, teaching practices, and the uptake of language-support services and generative AI tools.

The results indicate that publication norms and assessment criteria implicitly reward one language over others, and that tackling this bias is essential for fair and genuinely comprehensive research evaluation, in line with CoARA's goals.

The full questionnaire (6 sections, 33 questions, with branching logic) is available on Zenodo and can be freely reused and adapted by other institutions (https://zenodo.org/records/17512397).