GPOP provides longitudinal and cross-sectional data about the content, structure and evolution of public opinions across the world. This on-going data collection and data producing effort is one of the largest in the field. Considering the project is still in its early stages, it sets out to systematically add and integrate updated versions of the dataset.
One of the key contributions of this project is the interpretation of public opinion as multi-dimensional. Different components can be analysed as separate time series or they can be combined into more general public opinion measures. The project provides a tentative list of public opinion components. This list is neither exhaustie nor definitive, but provides a comprehensive outlook of the potential constructs of public opinion.
Download the GPOP codebook here
The project typically releases time-series data with yearly public opinion values. These measures stem from thousands of survey marginals and are constructed using a dyadic ratios algorithm. The output data is available in both smoothed and unsmoothed formats.
The project provides its data in a standardised and harmonised format, namely as CSV and Stata files, together with a corresponding analytics report and a set of data collection guidelines. For now, we provide data sets that are drawn from a pre-set selection of marginals and variables. In the near future, the project will integrate the dyadic ratios algorhitm into the website so users will have the option to select their own set of variables and calculate their corresponding public opinion measure.