Abstract

The design of comparative research is conditioned largely by four rules:

(1) Comparability - Two or more instances of a phenomenon may be compared if and only if there exists some variable, say V , common to each instance.

(2) Mill's 1st Canon - No second variable, say U is the cause or effect of V if it is found when V is found

(3) Mill's 2nd Canon - No second variable U is the cause or effect of V if it is found when V is not.

(4) Rule of One Variable - No second variable U is definitely the cause or effect of V if there exists a third variable, say W, that is present or absent in the same circumstances as U

The paper suggests that these four rules are the logical foundation of comparative analysis.many other rules exist, but almost all occupy a logically inferior status to these rules. many are derivable from these rules. The paper study rules 1-4 keeping three objectives in mind:

First, to what degree are these rules adequateguides to inference?.

Second, to what degree are some of the rules derived from rules 1-4 validlydeduced?.

Third, and most important, to what degree are rules 1-4 sufficient guides to inference?