This is the repository for the Academic Priors Project, a research project aiming to understand whether cognitive biases or individual differences in cognitive style can influence academics' beliefs about controversial topics in their fields. The current link to our data dashboard (a broad, exploratory overview of one part of our data) is [https://thawing-plains-08488.herokuapp.com/][1]. We recommend you bookmark *this* OSF page if you would like to stay informed about updates or publications. The repository folders below are as follows # Additional plots These are plots of specific relationships in the data, such as between a controversial theme and the various research areas and methods used be respondents. The folder is divided into subfolders to distinguish between different plots types and different sets of associations (e.g., theme vs cognitive trait; theme vs research area; theme vs research method). # Analyses This folder contains R Markdown scripts (`.Rmd`) and PDFs produced by those scripts, which provide details on our reported analyses as well as various supplementary analyses. Each script corresponds to a subsection of the Results and containts a list of contents. # Data Our survey data is contained in `survey.csv`, the variable names/column names of which are explained in `variable_names.csv`. This folder also includes the research topics identified by respondents (`topics.csv`); the aggregated counts per research topic (`research-topics-summary.csv`); and the list of journals identified in Web of Science (`journals.csv`). We note that the survey data have been censored as follows for open sharing (whereas our analyses reported in the manuscript use the full data set). The data are anonymous, but as the survey asked about several demographic details (age, gender, academic rank), we have decided to replace some of the demographic variables with missing (`NA`) values whenever one of the responses were relatively rare. Further, whereas in general we excluded any research topic reported by fewer than 10 people, in the case of these less-common demographic groups, we excluded any topic mentioned by fewer than 200 people. `similarity.csv` contains the aggregated cosine similarity metrics across survey and bibliometric data. However, we do not made raw bibliometric data available, as we do not have participant consent to make it public. # Materials This folder contains examples of the questionnaires used to measure cognitive traits. # Model fits Some of the regression models take a considerable amount of time to run. Thus, we have saved the models as `rds` files so that they can be loaded directly, without having ot re-run them. [1]: https://thawing-plains-08488.herokuapp.com/