(Page 13 of 26) - Qualitative Comparative Analysis (QCA): State of the Art and Prospects authored by Rihoux, Beno�t. and Ragin, Charles.

presence of the outcome:

A ----> Y

absence of the outcome:

a ----> y

Obviously, the solutions incorporating don’t care combinations are remarkably
parsimonious, but are they plausible? Before addressing this question, it is important to
point out that given the evidence in Table 2, a conventional quantitative analysis of the data
would quickly lead to the identification of condition A as the proper explanation of outcome
Y. After all, as the table shows, whenever A is present, Y is present; whenever A is absent,
Y is absent. None of the other causal conditions displays this simple relationship. Thus, the
QCA solution incorporating don’t care combinations dovetails with the results of a
conventional quantitative analysis of the same data.

The plausibility of this solution, however, depends upon the results of the researcher’s
counterfactual analysis. Consider the analysis of the presence of outcome (Y). Without
incorporating don’t care combinations, the solution is A

⋅

c; with don’t care combinations,

it is A. It follows that six don’t care combinations have been incorporated into the
parsimonious solution: A

⋅

d, A

⋅

D, A

⋅

d, A

⋅

D, A

⋅

d, and A

⋅

D. In

essence, the conclusion that A is the sole cause of Y, based on the analysis framed by the
truth table, assumes that if any of these six combinations of conditions could be found, they
would also display the outcome (Y).

In other words, the analysis of six counterfactual

cases undergirds the conclusion that A by itself causes Y, which is a dramatic use of
simplifying assumptions. For this reason, it is common in presentations of QCA to
emphasize the fact that researchers must evaluate any "remainders" incorporated as
"simplifying assumptions" into the solution of a truth table. This admonition is equivalent
to advising QCA researchers to conduct all the requisite counterfactual analyses.

Too often researchers by-pass counterfactual analyses because these assessments are tedious
and time consuming. Instead, they embrace parsimony and automatically use all the
simplifying assumptions incorporated into the most parsimonious solution they can produce.

Of course, if a "found" combination were to disconfirm the assumptions behind the solution (A

causes Y), then it would be incumbent upon the researcher to explain the inconsistency, based on
in-depth analysis of the found case(s) displaying the combination in question.