Also found in: Dictionary, Wikipedia.
- any method that involves the examination of similarities and differences between phenomena or classes of phenomena with the aim of:
- establishing classifications and typologies of social phenomena; and
- the testing of hypotheses about casual relations by examining the empirical association and temporal ordering of factors.
- specifically cross-cultural or cross-societal (including historical) comparison of similarities and differences between social phenomena with the above aims (see CROSS-CULTURAL COMPARISON, HUMAN RELATIONS AREA FILES).
An early systematization of the comparative method was provided by J. S. MILL. The three most used of these methods within sociology are outlined in Fig. 5. Explicit use of Mill's methods is seen in classical sociological studies such as DURKHEIM's Suicide (1897) and WEBER's Protestant Ethic and the Spirit of Capitalism (1904-05). The method of concomitant variation receives further elaboration and systematization in modern statistical analysis (especially see STATISTICS AND STATISTICAL ANALYSIS, CORRELATION).
Compared with the experimental method proper, the problems which occur when the comparative method is used arise from an in ability to manipulate truly ‘independent’ variables. The main problem is the possible influence of unknown variables which, in the natural settings observed, may affect in unknown ways the variables for which a direct causal or concomitant relation is suggested.
The early use of the comparative method in cross-cultural analysis was in the work of evolutionary sociologists who have often been accused of suspect judgements of similarity and difference, and of studying units out of context.
Generally, the use of the comparative method is not invalidated by such problems, but they do underline the difficulties which can attend the use of the method, especially in its cross-cultural forms.
A more root-and-branch objection to the use of the comparative method in sociology arises from theorists who emphasize the importance of MEANINGFUL UNDERSTANDING AND EXPLANATION, or VERSTEHEN. In extreme cases (e.g. WINCH,1958),no place is seen within sociology for the testing of general hypotheses of the conventional scientific kind. However, most sociologists reject the RELATIVISM involved in this view and continue to examine general hypotheses.
With implications for both experimental and quasi-experimental methods, modern philosophical analysis has rejected Mill's view that INDUCTION AND INDUCTIVE LOGIC and the comparative method can provide the conclusive proof possible in deductive LOGIC (see EMPIRICISM, FALSIFICATIONISM). However, this does not undermine the usefulness of the comparative method any more than it undermines the experimental; it merely points to there being no recipe for establishing causality.