- Unlike conventional experimentation, QE most often take place outside a laboratory
- Assignment of units to treatments is not random in QE
- QE involve one or more treatments and measurements before and after a treatment.
- Random assignment of units to treatment is sometimes not possible for ethical and logistical reasons. Thus, QE is employed
- QE's ontological assumption is that an external world of objects exists, and these objects are lawfully interrelated.
- The goal of QE is to identify manipula that can make a difference to specific outcomes. QE seek knowledge about dependable but probablistic causal relationships between manipula (independent variables) and measured outcomes, knowing that understanding and prediction will be enhanced by learning why particular consequences have occurred.
Reference
Cook, T.D. (1983) "Quasi-experimentation: its ontology, epistemology, and methodology" pp. 74-94, .(editor) Beyond Method, SAGE.