Reference class

from Wikipedia, the free encyclopedia

Reference class denotes the class of events and objects or facts to which a statistical hypothesis or assessment, in particular the relative frequency , or a relative attribute (such as "(relatively) large") relates.

In the frequency interpretation of the concept of probability, there is no probability without a specified reference class. This view excludes the likelihood of single events or individuals, since by definition these cannot be assigned to a reference class (which would not only consist of themselves and therefore would not allow any further conclusions).

Reference class problem

So that a statement like "This dog is big" can be evaluated with regard to its truth or epistemic justification, the reference class must be specifiable. In more complex cases, however, the relevant reference class is often not clearly determinable. For example, a person who wants to quantify their life expectancy can refer to different statistics that relate to the life expectancy of their occupational group, a certain social milieu or certain aspects of lifestyle. How small or large should the relevant case group be narrowed down (e.g. to all city dwellers in general, all in the respective country or all in the respective city)? In the philosophy of science or philosophy of probability, this topic is discussed as the "problem of the reference class".

See also

Individual evidence

  1. See Alan HájekInterpretations of Probability: 3.4 Frequency Interpretations. In: Edward N. Zalta (Ed.): Stanford Encyclopedia of Philosophy . .
  2. See also the compact presentation by Peter Baumann: Epistemology , Metzler, Stuttgart-Weimer, 75ff.
  3. Cf. Antony Eagle:  Chance versus Randomness: 4.2 The Reference Class Problem. In: Edward N. Zalta (Ed.): Stanford Encyclopedia of Philosophy . ; Jan-Willem Romeijn:  Philosophy of Statistics: 2.1 Physical probability and classical statistics. In: Edward N. Zalta (Ed.): Stanford Encyclopedia of Philosophy .

literature

  • Donald Gillies: Philosophical Theories of Probability . London 2000.
  • Hans Reichenbach : Probability theory: an investigation into the logical and mathematical foundations of the calculation of probability. Leiden 1935.