• Choose a category
• All
• Classic-Fiction

Bennett’s research on Soviet interventionism also employed an additional kind of congruence test. The research objective was to test a relatively new theory, learning theory, as an explanation for patterns of Soviet military intervention. This required first establishing whether there was any unexplained variance after accounting for the combined effects of more established theories. Bennett thus canvassed these theories and assessed their individual and collective congruence with both the rise and fall of Soviet interventionism. Bennett concluded that these theories collectively provided a more complete explanation of the rise of Soviet interventionism in the 1970s than of its fall in the 1980s (which is consistent with the fact that many analysts in the late 1970s expected such interventionism to continue to increase). This test suggested that it was not possible to reject out of hand that a learning explanation might account for some of the variance in Soviet policies.

Multivariate congruence testing can be complex, but it is also a familiar form of historical analyses and arguments. One historian may argue that the structure of the international system and the bipolar distribution of power between the United States and the Soviet Union made the Cold War inevitable. Another may argue that the Cold War arose from not just the distribution of power, but also from the specific domestic political dynamics in the United States and Soviet Union and despite the lack of any immediate danger of a military invasion by one superpower against the other. A third might argue that this balance of contributing and counteracting forces underdetermines the emergence of the Cold War unless one takes Stalin’s personality into account.

Two injunctions can help clarify such debates. First, it is important to consider a wide range of potentially causal factors, to specify the predicted contributing and counteracting effects of each, and to identify where underlying causal arguments are complementary and competing. Second, it is useful to guard against the bias of what has been termed “explanatory overdetermination.”385 When called upon to predict events, theorists and experts often give underdetermined accounts, yet when these same observers are asked to explain past events, their accounts make these events seem overdetermined. For example, almost no scholars predicted the collapse of the Soviet Union and the end of the Cold War, but afterwards many scholars pointed to numerous, seemingly overdetermining “causes” of these outcomes. Careful use of congruence testing, and inclusion of all the candidate theories, might instead lead to the conclusion that these outcomes were underdetermined, or at least that their timing and particular course could have been quite different if a few variables had been changed.

We now discuss how researchers can assess their preliminary findings of congruity.

How Plausible is the Claim of Congruity?

The possibility that consistency between the values of the independent and dependent variable in a given case is not spurious—and possibly causal—gains a measure of support if the relationship can be supported by a general law or statistical generalization. For example, a causal inference drawn from the observed consistency between an independent cognitive variable(s) such as the actor’s belief and some aspect of that individual’s behavior can be supported by psychological theories of cognitive balance that call attention to the fact that individuals generally (at least under certain conditions) strive to achieve consistency between their beliefs and their actions. This, of course, is a very general theory. If more specific generalizations or theories could be adduced, the imputation of a causal relation would be strengthened. Typically, the stronger and more precise the version of a more general theory, the more confidence we ought to attach to claims that consistency is not spurious.386

Is the Independent Variable a Necessary Condition for the Outcome of the Dependent