Summary: KKV are trying to unite quantitative and qualitative methodologies under ‘one logic of inference.’ Two major components of research design: i) it should be important in the real world; ii) it should make a specific contribution to an identifiable scholarly literature in order to aid in explaining aspects in the world. Theories should be falsifiable, not necessarily parsimonious (it implies a specific ontology, something Hall (in Aligning Ontology and Methodology) picks up on). Process of data collection should be recorded to allow for replication, as much data as possible should be collected, methods should be reliable.
Important Insight: One logic of inference, and the application of a positivistic approach to all aspects of research – KKV offer suggestions on how to improve research questions, theory, data quality and existing data – all based on systematizing the approach.
Critique: Research must be “important in the real world?” According to who? Also, the ‘specific contribution’ is just Kuhnian mopping up, there’s no room for innovation here, only operating within the status quo.
Chapter 2: Descriptive Inference
Summary: Description and explanation are interactive, and both depend on inference, but there is a difference between mere description and descriptive inference. ‘Interpretation/description is valid in and of itself, but evaluating the veracity of any interpreted claim must follow the logic of scientific inference.’ They recognize the importance of case studies in yielding valid causal inferences.
Important Insight: Descriptive inference helps to distinguish systemic components from non-systemic components – could be linked well with Przeworski & Teune’s Most Different Systems Design.
Critique: There’s nothing really specific in this chapter, if you buy the whole project it follows. However, I will say that KKV’s emphasis on the need to focus on the outcomes we wish to describe or explain may shed some light on why qualitative approaches have to face up to the ‘selection on the DV’ critique…