Most of the social science research is based on the implied assumption that measurement error is the same across key socio-demographic groups and all differences in key statistics of interest are real. Nevertheless, there is evidence that this is not the case. In this paper, the authors tackle this important topic by investigating if data quality is different in three key groups analysed in social research: the unemployed, the low-income respondents and the elderly respondents. This paper investigates if the assumption of equal data quality across these three key variables in a popular online probability panel in the Netherlands. To do this the authors use multiple types of variables and multiple statistical approaches to evaluate data quality