Tables of Limiting t Values for Probabilities to the Nearest .001 (n=2-16)
Report Number: AMRL TR 67-161
Author(s): Cotterman, Theodore E., Knoop, Patricia A.
Corporate Author(s): Aerospace Medical Research Laboratories
Laboratory: Aerospace Medical Research Laboratories
Date of Publication: 1968-05
Pages: 54
Contract: Laboratory Research - No Contract
DoD Project: 1710
Identifier: AD0673347
Abstract:
Tables of t to the nearest .00001 for probabilities within the range of .000 to .500, by increments of .001, are presented for samples of 2 through 16. The 't' values given are the limiting or boundary values having the indicated probability, rather than the closest approximation, thus making interpolation unnecessary in typical use. The tables are intended primarily as an aid in implementing a purposive probabilistic approach to psychological measurement in which the 't' distribution (or some other appropriate model) is used as a reference in deriving probabilistic behavioral measures. Secondarily, they also may be used to achieve greater precision in specifying the critical 't' for conventional tests of significance. The manner of use is briefly explained, and the way in which they were constructed is described. The FORTRAN computer program used in constructing them is included.
Provenance: RAF Centre of Aviation Medicine
Author(s): Cotterman, Theodore E., Knoop, Patricia A.
Corporate Author(s): Aerospace Medical Research Laboratories
Laboratory: Aerospace Medical Research Laboratories
Date of Publication: 1968-05
Pages: 54
Contract: Laboratory Research - No Contract
DoD Project: 1710
Identifier: AD0673347
Abstract:
Tables of t to the nearest .00001 for probabilities within the range of .000 to .500, by increments of .001, are presented for samples of 2 through 16. The 't' values given are the limiting or boundary values having the indicated probability, rather than the closest approximation, thus making interpolation unnecessary in typical use. The tables are intended primarily as an aid in implementing a purposive probabilistic approach to psychological measurement in which the 't' distribution (or some other appropriate model) is used as a reference in deriving probabilistic behavioral measures. Secondarily, they also may be used to achieve greater precision in specifying the critical 't' for conventional tests of significance. The manner of use is briefly explained, and the way in which they were constructed is described. The FORTRAN computer program used in constructing them is included.
Provenance: RAF Centre of Aviation Medicine