In this paper it is considered how data from different origins and of different qualities can be combined in order to arrive at a more complete picture of phenomena considered in empirical research. It is shown how qualitative and quantitative data can be compared in meaningful ways, and how they can be analyzed together in statistical analyses. In addition to respondent-based information registry data may be used. This makes it possible to add information that are unaffected from memory distortions, non-response problems and from social desirability bias, but registry information have their particular weaknesses that need to be taken into account. Finally aggregate-level data may be used. Although they are only second best if standing alone, aggregated data may act as contextual information that may be used for performing multi-level analyses. Finally a study on the role of social factors in the course of breast cancer as an example is described where data from different sources are combined in order to tackle complex health-related research questions.
Keywords: Qualitative data, quantitative data, secondary data, mixed methods research, registry data, aggregate-level data.