Consensus-based TRansparent, sUStainable, and Transferable (TRUST) best-practice guidelines for visual word recognition research

Reading is a fundamental life skill, and at its basis is the ability to process single words. Researchers often study single-word processing using lexical decision tasks. However, most research is conducted in a single language by separate research groups, limiting comparability, replicability, and generalizability. In addition, best-practice knowledge is often only transmitted informally within labs, leading to heterogeneity and reduced reproducibility across labs. Here, we, the TRUST Consortium, established a consensus-based research guideline for current best practices in psycholinguistic reading research by integrating expert knowledge. We used German as a case study and formulated guidelines for visual lexical decision experiments. We reached consortium-level consensus through multiple feedback circles and quantified agreement using importance ratings. The guidelines are openly available as a living guideline document on a webpage (\href{https://trust-guidelines-v01.readthedocs.io/en/latest/}{\textbf{trust-guidelines-v01.readthedocs.io}}), with git-versioning infrastructure that makes changes transparent and enables discussion. To validate the guidelines, we implemented the German Lexicon Project, providing an empirical feedback loop to test our suggestions. The living, open nature of the guidelines invites broader community feedback. It also enables future extension to other languages, groups (e.g., dyslexics), and methods (e.g., brain imaging), as it is openly available and includes instructions for hosting an adapted version at no cost. Thus, we provide a document intended to raise research standards in reading research worldwide, as a service to the community, potentially increasing replication rates and generalizability of relevant effects and phenomena.