Taylor-Russell tablets

from Wikipedia, the free encyclopedia

The Taylor-Russell tables are used to determine the success rate of an instrument depending on the validity of the instrument, the selection rate and the base rate. The validity refers to the reliability with which the instrument differentiates between suitable and unsuitable persons (e.g. applicants). The basic or basic rate means the proportion of people in the population who would be suitable at all based on a minimum requirement. The selection rate or quota describes the proportion of people you want to select from the population using the test instrument. Taylor and Russell determine the success rate by how large the proportion of people among the selected is who is actually suitable. A good selection (decision) is made when this proportion is maximum. Depending on the basic rate, selection rate and validity of the method, the various success rates can be read from the tables by Taylor and Russell.

According to Taylor and Russell, the relationship between the quantities mentioned is as follows:

  • the higher the base rate, the higher the success rate
  • the lower the selection rate, the higher the success rate
  • a low basic rate in combination with a high selection rate requires the method to be highly valid

On the one hand, the boards can be used to determine whether the use of a test procedure has any significant advantage, i.e. whether it increases the success rate. On the other hand, it is possible to estimate what a change in the validity (e.g. by choosing a different diagnostic tool) or the base rate (e.g. increasing the base rate by preselecting applicants) means for the success rate.

Web links

literature

  • HC Taylor, JT Russell: The relationship of validity coefficients to the practical effectiveness of tests in selection: Discussion and tables. In: Journal of Applied Psychology, 23 , 1939, pp. 565-578.
  • Noack, H. & Petermann, F. (1995). Decision theory. In RS Jäger & F. Petermann (eds.), Psychologische Diagnostik (pp. 295–310). Weinheim: Psychology Publishing Union.