Sampling designs for analyzing publicness: Alternatives and their strengths and weaknesses

Abstract

Research comparing public and private organizations and otherwise analyzing "publicness" involves complex challenges. These include the challenge of designing and attaining adequate samples to represent the two complex categories of "public" and "private," as well as dimensions of publicness, and subcategories and control variables needed for valid comparisons. This review of sampling alternatives begins with discussion of the strengths and weaknesses of the presumably optimal design, a national probability sample of organizations. Due to the expensive nature of such a design - the discussion concentrates on the one such study ever conducted - the discussion then considers the strengths and weaknesses of the purposive samples and samples of opportunity that most research have used. In spite of limitations in representativeness and in accounting for all variables needed to eliminate alternative interpretations, studies using such samples can be aggregated to support conclusions about differences between public and private organizations that are by now well-founded. Researchers should continue to seek opportunities for the optimal large representative samples. Lacking such opportunities, researchers can contribute usefully to analysis of publicness by carefully designing their studies to make them consistent with previous studies and to support aggregation with previous studies. © The Author 2011. Published by Oxford University Press on behalf of the Journal of Public Administration Research and Theory, Inc. All rights reserved.