This presentation stems from a survey addressed to researchers in academic institutions and funding agencies, aimed at understanding perceptions of research evaluation and excellence across Europe and beyond, conducted within the TIER (Towards an Inclusive Evaluation of Research) working group of CoARA and presented during an online TIER workshop, the 27th June, 2025. Specifically, these slides present an approach to detecting implicit biases — particularly those related to gender — in evaluation settings. The core finding is that language framing can be used as a proxy for measuring biases that would not otherwise be openly acknowledged. From the survey, which reached approximately 2,000 researchers, it emerged that respondents changed their assessments of stereotypical statements when the same content was expressed using different framings. This led us to investigate the underlying cognitive mechanisms more deeply, with the aim of identifying possible solutions for more equitable practices in research evaluation settings.