Education bias in probability-based surveys 

SBI contributors 

Prof. Dr. Michael Bergmann, Dr. Arne Bethmann

Project description

Survey data are only as good as the populations they represent. Yet in Germany, individuals with lower levels of formal education are consistently under-represented. Our review of 67 surveys conducted between 2000 and 2023 across 19 survey programmes shows that this bias is particularly strong in web-based self-administered and telephone interviews. 

As survey organisations move towards self-administered modes to cut costs, the problem is likely to worsen. This trend raises important concerns about data quality, since education bias can distort findings and undermine the reliability of analyses based on probability-based surveys. 

We are investigating the mechanisms behind non-participation among less educated individuals and exploring countermeasures that could strengthen representation. By better understanding these dynamics, our goal is to help survey projects reduce systematic bias and maintain the credibility of their findings. 

Status

Active

Selected publications 

Stein, A., Gummer, T., Naumann, E., Rohr, B., Silber, H., Auriga, R., Bergmann, M., Bethmann, A., Blohm, M., Cornesse, C., Christmann, P., Coban, M., Décieux, J., Gauly, B., Hahn, C., Helmschrott, S., Hochman, O., Lemcke, J., Naber, D., Pötzschke, S., Roßmann, J., Schanze, J., Schmidt, T., Schneider, S., Spangenberg, H., Rettig, T., Trappmann, M., Weinhardt, M., Weiß, B. (2025). Education bias in probability-based surveys: Evidence from Germany between 2000-2023. International Journal of Social Research Methodology, 1-18. DOI: 10.1080/13645579.2025.2508889