• Choose a category
• All
• Classic-Fiction

By Root 1532 0

for the failures of their predictions appear to be biased in favor of their initial theories, see Philip Tetlock, “Theory-Driven Reasoning about Plausible Pasts and Probable Futures in World Politics: Are We Prisoners of Our Preconceptions?” American Journal of Political Science, Vol. 43, No. 2 (April 1999), pp. 348-349. Researchers should also be on guard against other cognitive biases, including the bias toward over-confidence in one’s causal theories, a preference for uni-causal explanations, and a tendency toward assuming that causes resemble consequences in terms of scale, scope, or complexity.

219

Researchers also often find that their preliminary knowledge of the values of the independent and dependent variables is mistaken, particularly if it is based on news accounts or secondary sources that do not use precise definitions. Thus even these variables can provide some use-novelty for researchers; however, as we note in our chapter on congruence testing, tests of the congruence of independent and dependent variables, even with the advantage of use-novelty, are challenging and often less conclusive than process-tracing tests.

220

William Wohlforth suggests this practice in “Reality Check: Revising Theories of International Politics in Response to the End of the Cold War,” World Politics, Vol. 50, No. 4 (July 1998), pp. 650-680.

221

Alexander L. George and Richard Smoke, Deterrence in American Foreign Policy: Theory and Practice (New York: Columbia University Press, 1974). See also the discussion of George and Smoke in the Appendix, “Studies That Illustrate Research Design.”

222

Bruce Jentleson, Ariel Levite, and Larry Berman, eds., Foreign Military Intervention: The Dynamics of Protracted Conflict (New York: Columbia University Press, 1992). This book is discussed in the Appendix, “Studies That Illustrate Research Design.”

223

Sidney Verba, “Some Dilemmas of Political Research,” World Politics, Vol. 20, No. 1 (October 1967), pp. 113-114.

224

Ibid.

225

Gary King, Robert O. Keohane, and Sidney Verba, Designing Social Inquiry: Scientific Inference in Qualitative Research (Princeton, N.J.: Princeton University Press, 1994), pp. 29, 20.

226

Marianne Weber, Max Weber: A Biography, trans. Harry Zohn (New Brunswick, N.J.: Transaction Press, 1988), p. 278, cited in David Laitin, “Disciplining Political Science,” American Political Science Review, Vol. 89, No. 2 (June 1995), p. 455.

227

See Chapter 11 for a more detailed discussion of this research as an example of typological theory. See also Andrew Bennett, Joseph Lepgold, and Danny Unger, “Burden-Sharing in the Persian Gulf War,” International Organization, Vol. 48, No. 1 (Winter 1994), pp. 39-75.

228

See Stephen W. Van Evera, Guide to Methods for Students of Political Science (Ithaca, N.Y.: Cornell University Press, 1997), p. 34.

229

See, for example, Colin Elman, “Horses for Courses: Why Not Neorealist Theories of Foreign Policy,” Security Studies, Vol. 6, No. 1 (Autumn 1996), pp. 7-53. Elman critiques neorealist theories for claiming to eschew any testable predictions on individual states’ foreign policies.

230

In a similar formulation, Stephen Van Evera suggests that the probity of an empirical test depends on the certainty and uniqueness of the predictions a theory makes regarding the test. “Hoop tests” are those in which the predictions of a theory are certain but not unique. Failing such a test is damaging to a theory, but passing it is not definitive. “Smoking gun tests” are those in which a theory is unique but not certain. Passing such a test is strong corroboration, but failing it does not undermine a theory. “Doubly decisive” tests, when predictions are both unique and certain, are those in which either passage or failure is definitive. (Van Evera gives the example here of a bank camera, which can both convict those guilty of robbery and exculpate the innocent.) “Straw-in-the-wind” tests, with predictions of low certainty and uniqueness, are not definitive regardless of the outcome. See Van Evera, Guide to Methods, pp. 31-32.

231

Colin