SocioLex-CZ: Normative estimates for socio-semantic dimensions of meaning for 2,999 words and 1,000 images.

When we encounter words, we activate not only the social information provided by the speaker, but also the rich semantics of the words' meaning, and quantifying this information is a key challenge for the cognitive and behavioural sciences. Although there are many resources available that quantify affective and sensorimotor information, there are relatively few resources available that provide information on social dimensions of meaning. We present the SocioLex-CZ norms, where the primary focus is on socio-semantic dimensions of meaning. Across two experiments, we introduce normative estimates along five dimensions-gender, political alignment, location, valence and age-for a large set of Czech words (Experiment 1) and images (Experiment 2) from 1,709 participants. We provide a series of analyses demonstrating that the norms have good reliability, and present exploratory analyses examining how the variables interact with one another within and between words/images. These norms present a valuable dataset that quantifies socio-semantic representations at scale, which we hope will be used for a range of novel and multidisciplinary applications, thereby opening up new pathways for innovative research. We make the data, code and analysis available at https://osf.io/pv9md/ and also provide an interactive web app at https://tinyurl.com/sociolex-cz-app .