CzeLex: A megastudy dataset of concreteness, imageability, age of acquisition, valence, and gender ratings for 9601 Czech words

Understanding how semantics are processed at the lexical level has been a key aim for the cognitive sciences, with large-scale semantic rating studies in different languages playing a crucial role in facilitating the advancement and development of theories of lexical processing, as well as providing valuable methodological resources that can be used in empirical investigations. In this paper, we present the CzeLex database, a megastudy of 9601 Czech words that have been normed along five psycholinguistic dimensions from a large sample of 4405 Czech-speaking participants. We provide individual rating data at the item and participant levels, as well as aggregated data from both mean ratings and latent means derived using cumulative-link mixed modeling, along with a comprehensive analysis comparing aggregation methods. We demonstrate that the dataset has good reliability and that specific dimensions predict lexical processing measures in a lexical decision task, regardless of the aggregation method used. These norms are the largest available subjective rating resource for Czech and further contribute to the growing body of languages with openly available data on these dimensions. We hope that the dataset will be a valuable resource for researchers working on Czech, in addition to opening up new opportunities for cross-linguistic semantic investigations. All data, code, and analyses are available at https://osf.io/tqxfm .