Comparative method
The comparative method is defined as a method for the systematic comparison of units of investigation with the general goal of empirical generalizability or the verification or formulation of hypotheses .
As a special characteristic of the political science discipline, this article now deals in more detail with the specific methodological bases of the comparison in the field of sociology / political science .
(See also: Comparative Politics .)
Purpose of comparison
In general, the comparison is used in social and political science disciplines:
- the formal description (comparison of structures, government systems, etc.),
- the explanation based on this (of causal relationships, cause-effect relationships in a narrower understanding),
- a forecast based on the analysis (to predict possible developments, for example through historical comparison), and
- the evaluative interpretation (especially for normative statements about the political system).
The reasons for a systematic comparison (e.g. of political systems, process structures, etc.) lie above all in the representation of the unknown and the highlighting of certain specifics or anomalies. Although it seems to lead the scientific terminology ad absurdum, it also invalidates the proverbial incomparability of apples and pears , since the mere comparison of them represents a comparative method aimed at specific particularities.
Furthermore, the comparative method is used for systematic categorization and typology through the consistent summary of characteristics.
A distinction is made between the following types :
In the classificatory type , all units of investigation are to be classified into exactly one group according to their distinguishing feature through the formation of distinct classes. Example: Classification of different countries according to their party system.
The extreme type is regarded as the most extreme comparative difference type , which defines the conceptual poles between which all possible phenomena can be classified in a kind of series. Example: Establishing a party spectrum with the corresponding extreme margins.
The ideal type, on the other hand, is completely constructed and cannot be empirically verified, since it does not (yet) appear in reality. Example: ideal of a socialist society.
method
According to John Stuart Mill (for the first time as early as 1843), a distinction is made between two fundamentally different research strategies: the concordance and the difference method. With the concordance method , the variables under consideration should be as similar as possible, but the remaining framework conditions should be completely different. We are looking for the cause of a certain phenomenon that occurs under completely different framework conditions. With the difference method, on the other hand, the operational variables are different, but the context is similar. The aim is to clarify why a certain phenomenon does not occur under similar framework conditions.
How a practical comparison is ultimately made depends primarily on what specifically found out, ergo which theory is to be developed. When building theories, the methodical trade-off (see Fig. 1) always moves in a kind of tension between the specificity and the range of a possible result arising from the investigation. In concrete terms, this means that a possible statement is either very far-reaching, but only a little in-depth, or appears very concrete and detailed, but only applies to very few cases. Maximizing both at the same time is not possible.
It should be noted that mostly only a small number of cases (country comparisons, individual case studies, etc.) are examined, reality has to be reduced to a measurable level of essential variables and quasi-experimental conditions are often assumed. This partly justifies a fundamental skepticism towards the results of the methodological comparison.
Problems of the comparative method
In addition to the methodical trade-off, two further general difficulties in dealing with the method explained above should be mentioned at this point.
On the one hand, it is very difficult to locate the comparative method in political science and other social science disciplines, as the comparison takes place in a certain way in every political science method (for example, also in the case of individual case studies up to comprehensive statistical surveys). Nevertheless, in his systematization , Arend Lijphart compared three other social science methods: the experiment , the case study and the statistical method .
On the other hand, the methodical approach itself appears to be at least problematic, since there are often only a few cases but an infinite number of variables (e.g. the framework conditions for a country comparison), quasi-experimental assumptions in sociology cannot be implemented in practice (e.g. constant laboratory-like conditions) and multi-causal relationships often reduced to simple cause and effect principles. Ultimately, contradicting hypotheses can even be proven, since the selection of the cases to be compared is always a subjective decision.
literature
- D. Berg-Schlosser / F. Müller-Rommel (eds.): Comparative Political Science . Leske + Buderich, Opladen 1987. ISBN 3810005649
- Detlef Jahn: Introduction to Comparative Political Science . VS-Verlag, Wiesbaden 2006. ISBN 3-8100-3894-6 . Especially chapter 5.
- Dieter Nohlen: Comparative Method . In: Nohlen / Schultze (Ed.): Lexicon of Political Science . 2nd Edition. Beck, Munich 2004, pp. 1042-1052. ISBN 3406511279
- Dieter Nohlen: Typus / Typologie . in: Jürgen Kriz u. a .: Political science methods . Beck, Munich 1994, pp. 491-496. ISBN 3406369049