Identification

Charles Manski gave a brilliant example of what identification is all about:

Suppose that you observe the almost simultanous movement of a man his image in a mirror. Does the mirror image cause the man's movements or reflect them? If you don't understand something of optics and human behavior, you will not be able to tell. (Manski, 1995, p. 1)

Methodological research in the social sciences uses statistical theory. The empirical problem is to infer some feature of a population described by a probability distribution. The data available to the researcher (be it field or experimental data) are observations extracted from the population by some sampling process. In this framework, the statistical and identification problems can be separated:

  • Identification is about the conlusions that could be drawn if one could use the sampling process to obtain an unlimited number of observations.

  • Statistical inference is about the (generally weaker) conclusions that can be drawn from a finite number of observations.

    Identification problems cannot be solved by gathering more of the same kind of data. They can be alleviated only by invoking stronger assumptions or by initiating new sampling processes that yield different kinds of data.

    See also: econometrics, selection problem

    Literature: Manski (1995)

    Entry by: Joachim Winter

    June 17, 1999
    Direct questions and comments to: Glossary master