Russel-Research-Method-in-Anthropology

Qualitative Data Analysis II: Models and Matrices 545
tion, the alternatives are to change the explanation to include the new case or
redefine the phenomenon to exclude the nuisance case. Ideally, the process
continues until a universal explanation for all known cases of a phenomenon
is attained. (Explaining cases by declaring them all unique is not an option of
the method. That’s a convenient way out, but it doesn’t get us anywhere.)
Charles Ragin (1987, 1994) formalized the logic of analytic induction,
using an approach based on Boolean logic. Boolean variables are dichotomous:
true or false, present or absent, and so on. This seems simple enough,
but it’s going to get very complicated, very quickly, so pay attention. Remember,
there is no math in this. It’s entirely qualitative. In fact, Ragin (1994) calls
his Boolean method of induction qualitative comparative analysis, orQCA.
Suppose you have four dichotomous variables, including three independent,
or causal, variables, and one independent, or outcome, variable. With one
dichotomous variable, A, there are 2 possibilities: A and not-A. With two
dichotomous variables, A and B, there are 4 possibilities: A and B, A and not-
B, not-A and B, not-A and not-B. With three dichotomous variables, there are
8 possibilities; with four there are 16 . . . and so on.
We’ve seen all this before: in the discussion about factorial designs of
experiments (chapter 5); in the discussion of how to use the number of subgroups
to figure out sample size (chapter 6); in the discussion of how to determine
the number of focus groups you need in any particular study (chapter 9);
and in the discussion of factorial questionnaires (chapter 10). The same principle
is involved.
Thomas Schweizer (1991, 1996) applied this Boolean logic in his analysis
of conflict and social status in Chen Village, China. In the 1950s, the village
began to prosper with the application of technology to agriculture. The Great
Leap Forward and the Cultural Revolution of the 1960s, however, reversed the
village’s fortunes. Chan et al. (1984) reconstructed the recent history of Chen
Village, focusing on the political fortunes of key actors there.
Schweizer coded the Chan et al. text for whether each of 13 people in the
village experienced an increase or a decrease in status after each of 14 events
(such as the Great Leap Forward, land reform and collectivization, the collapse
of Red Brigade leadership, and an event known locally as ‘‘the great
betrothal dispute’’). Schweizer wound up with a 13-actor-by-14-event matrix,
where a 1 in a cell meant that an actor had success in a particular event and a
0 meant a loss of status in the village.
When Schweizer looked at this actor-by-event matrix he found that, over
time, nine of the actors consistently won or consistently lost. That means that
for nine of the villagers, there was just one outcome, a win or a loss. But four
of the actors lost sometimes and won other times. For each of these four peo-