Datasets and Code for the Paper: **"What’s the measure? An empirical investigation of self-ratings on response scales"** The repository also contains data and analyses for a preliminary study (here referred to as study 3) which is not reported in the paper. ---------- All variables are described in the Codebook: - File: Codebook.docx - Feel free to contact the authors ( sven.hilbert@ur.de, florian.pargent@psy.lmu.de, elisabeth.kraus@ur.de ) if more information is needed. ---------- Datasets: - Files: study1.csv, study2.csv, study3.csv - Only studies 1 and 2 are reported in the paper. Study 3 was originally the first study conducted and also part of earlier drafts of the manuscript. Based on negative reviewer feedback, study 3 was removed from the current version to keep the paper more concise. - Datasets for studies 1 and 2 also contain variables which were not used for the analyses reported in the paper. ---------- Analysis Scripts: - Files: analysis_study1.R, analysis_study2.R, analysis_study3.R - For the original study 3, combined models were estimated for man and women including the interaction with gender. This modeling strategy was changed for studies 1 and 2 in the paper, where separate models were fitted for men and women, as the proportional odds model with interaction does not allow for different threshold patterns for men and women (the thresholds are only scaled by the interaction). - Scripts for studies 1 and 2 also contain analyses which were not reported in the paper. - Note that analysis scripts from different studies use the same object names, so be careful when comparing analyses (e.g. clear your workspace before switching between different studies). ---------- The "Registrations" tab contains a rudimentary partial preregistration for study 3 (which is not referred to in the paper).