Research Designs for Political Science: Contrivance and Demonstration in Theory and Practice

Research Designs for Political Science: Contrivance and Demonstration in Theory and Practice

Research Designs for Political Science: Contrivance and Demonstration in Theory and Practice

Research Designs for Political Science: Contrivance and Demonstration in Theory and Practice


Research design in political science has become too technical, mechanical, and uninvolved, argues David A. Bositis as he explains the need for an alternative design regimen that provides the means for real engagement in empirical research. Bositis' view offers a theoretical, imaginative, manipulative, and engaging alternative approach to political science research. His text is divided into three main parts. Part 1, "Epistemological Issues in the Design of Research," introduces the work's themes and context. Included in this discussion are research design as persuasive strategy, a critique of survey research, and a theory of contrivance and demonstration in design. In part 2, "Contrivance and Demonstration in Practice," Bositis examines the practice of design in political science, first through a discussion of the history of experimentation and then through an examination of both an integrated design and participant observation approach. Part 3 "The Politics of Research Design," offers an analysis of the politics and ethics that inform design choices. An extensive bibliography of nearly 400 entries is one of the most complete listings of experiments in political science to be found anywhere.


There are two fundamental standards by which every design must be judged: internal validity and generality (Campbell and Stanley, 1963:5). Internal validity is a judgment on findings of causal effect; thus, a design is internally valid if the information deriving from the design provides the basis for a strong finding of causal effect (or no effect as the case may be). the finding of a causal effect in a sample of observations does not, per se, indicate wide applicability of that finding. applicability, is a question of generality, of the aggregates and environments to which the finding can be generalized; this is a crucial judgment because a causal effect that applies to a single circumstance is of little intrinsic interest to science, regardless of its idiographic (unique and nonrecurrent) significance. There are many singular events of great interest to the political science community. Ronald Reagan's election and the U.S. defeat in Vietnam are individual events of enduring interest. However, political science, like all science, is nomothetic: it seeks to establish abstract general laws for indefinitely repeatable events and processes.

Internal Validity

A demonstration of causal effect requires evidence that the cause is antecedent to the effect, that the cause and the effect vary concomitantly, and that no extraneous factor is plausibly the source of this observed concomitant variation. the judgment of what is plausible is often difficult. a well-developed body of theory (when and if available) can occasionally provide help in making such a judgment, but usually one must rely upon the features of the design to address the problems of confounding extraneous variance.

In the temporal relationship between cause and effect, the cause is antecedent to the effect. This is true by definition and may seem trivial. It should be noted, however, that many research designs in political science are cross-sectional; all observations are made at a single point in time. With cross-sectional designs, the features of the design do not provide any evidence of a time structure. Under . . .

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.