What is good research? A seemingly simple question reveals itself difficult to answer. The current methods for identifying research quality come with validity issues due to a lack of conceptual scrutiny. I argue that research quality is a context dependent latent construct. While identifying research quality is a difficult task, it is not impossible. The methodological toolbox of the social sciences provides instruments to capture such latent constructs. I propose a method to conceptualise research quality in its context and argue that it is fruitful to combine peer review and indicator-based evaluations rather than playing them off against each other. Similarly, instead of juxtaposing notions of quality of different stakeholders, it is more promising to start from the scholars’ notions of quality and to add other stakeholders’ notions of quality in a communicative process arriving at a context-specific definition of research quality.