Sources of Bias and Quality of Data in Social Science Research

Surveys supply us with the statistical data for economic and administrative decisions. Errors in survey research may lead to wrong and costly decisions. For a better control of the quality of research, methodological standards and validity criteria should be communicated more clearly. In this paper, sources of bias in interviews, sampling, and data analysis are discussed. An often neglected sorce of bias is the fraudulent manipulation of data. How can faked data be detected? In a final section new tests using the Benford distribution are explored. However, more research is needed to ascertain the validity of Benford tests