Key points are not available for this paper at this time.
Summary Limits for the ratio of the means of two normal distributions are obtained by a method which appears to be analogous to that used in finding fiducial limits for the difference of two means with possibly unequal variances (the Fisher-Behrens problem). The properties of these limits are compared with those obtained using Fieller’s solution of this problem, for the case of known variances and zero correlation. The extension to more general cases is also discussed.
Monica A. Creasy (1954) studied this question.