Toward a meta-framework for conducting mixed methods representation analyses to optimize meta-inferences

Abstract

The purpose of this article is to propose a meta-framework for conducting what we term mixed methods representation analyses (MMRA). We define MMRA as the appropriate selection of sampling design (i.e., the sampling frame [random] or sampling boundary [purposive]; sampling combination, comprising the mixing dimension [partial/fully], time dimension [concurrent/sequential], emphasis dimension [dominant/equal status], and relationship among/between samples [identical/parallel/nested/multilevel]; sample size; and number of sampling units [e.g., of people, cases, words, texts, observations, events, incidents, activities, experiences, or any other object of study]) in order to obtain representation and concomitantly meta-inferences consistent with the study's generalization goal(s). Thus, the goal of conducting MMRA is to attain representation and interpretive consistency in order to enhance the rigor of mixed methods research studies.