Social science surveys often collect data from similar samples or using similar survey designs, but in different countries or at different times. When combined, these data offer additional insights. When this is the goal, it is necessary to first harmonise the data in order to make them comparable with each other. We offer tailor-made advice on harmonising your data.
How can we make data that was collected with different instruments comparable? We provide advise on the harmonisation of substantive variables measured with different instruments, in particular in the following areas:
- Assessing the comparability of different instruments
- Assessing the consequences of lack of comparability
- Harmonisation of substantive variables measured with different instruments; especially latent constructs, such as values, attitudes, intentions, emotions, or personality traits
In survey research, socio-structural varaibles are important evaluation criteria, as they make relevant social groups and thus the social structure of societies visible. We offer information and advice on the coding of socio-structural variables as well as on their harmonisation across samples, e.g. for cross-national comparison. Our focus is on the measurement of education, occupational coding as well as the measurement of socio-cultural background.
We have compiled various sets of information and hands-on guidelines on frequently occurring consulting topics. These already answer many questions that may arise when planning a study and collecting data. You can find this information under basic knowledge (link will follow shortly).
Our consulting services in the form of 'helping you to help yourself' are free of charge. Additionally, we are happy to support you with more in-depth services, which are fee-based. You will find further information in our pricing overview (link will follow shortly).
- Development of new measurement instruments
- Cognitive Pretesting
- Translation of questionnaires in international studies
Related articles in the GESIS Survey Guidelines: