Sampling Theoretically for Comparison

Sampling has historically been one of the major challenges of the comparative research approach. These sampling challenges primarily result from the way researchers select the cases/samples for the study. In this regard, researchers have to a large extent tended to employ non-probability convenience and purposive sampling techniques. Even though it may be argued that these sampling approaches need not be theory driven as samples tend to evolve in the process of research, more often than not, these sampling methods, especially in comparative research designs, while skewing research attention towards over-researched countries and cases, wealthy nations and incomparable cases, also introduce an element of bias into sampling and therefore into research findings. Thus, this paper argues for a move away from the simplicity of purposive and convenience sampling, to one of the more robust forms of theoretical sampling, in order to improve the research rigour associated with the comparative methodological approach. This paper accordingly postulates this may be achieved by engaging in some form of theoretical sampling. In this regard, this paper describes a two-phase method for generating comparative samples from theories, involving six distinct steps.