Mixed methods research requires integration of qualitative and quantitative data. However, there is debate about how to define integration and what is required for integration to occur. This paper describes a mixed methods research project which revisits datasets from different eras, which were originally instigated for different purposes and had different theoretical frameworks. Using selected results about the relationship between teenagers' vocational aspirations and adult occupations we show how the research topic and question integrated the projects within a constructivist theoretical position. We argue that reanalysing and comparing historical datasets can provide new insights into a topic even with minimal integration of the data. We recommend a broad definition of integration and reflexive research practice to encourage innovation and diversity in mixed methods research, particularly with regard to reanalysis of datasets from different eras.