The problem of the unambiguity of phase-shift analysis II

For the discrimination of phase-shift solutions, the new so-calledτ-criterion, based on the calculation of some functionτ of the experimental data, is used. This method can be applied, in general, to an arbitrary ambiguous phase-shift analysis. Theτ-criterion is more powerful than theχ2-criterion, e.g. for rejecting a less probable set it is necessary to have a smaller quantity of experimental points (in statistics this means that theτ-criterion gives a smaller Type II error). For the calculation of the quantityτ, the standard programme for phase-shift analysis with a small additional part can be used. A computer time of about only 5–10 min. on a computer of M-20 type is needed. A Type I error is then obtained from the tables of the normal distribution function.