W. L. Stevens
Articles written in Journal of Genetics
Volume 32 Issue 1 February 1936 pp 51-64
Various methods are in use for calculating a quantity to measure the interference between recombination in non-adjacent segments. Three are investigated in this paper and are shown to be unsatisfactory, because they lead to results which depend not only on the two segments in question, but also on the location of other genes followed in the experiment. Two of them, including the most popular, have expectations less than unity in the absence of interference. Since equality to unity is taken as a criterion of non-interference, this has led to erroneous conclusions.
The exact method, which is given in this paper, while being free from the above objections, is just as simple in calculation. A formula is derived for the standard error of this estimate.
Coincidences have been recalculated for the
Volume 39 Issue 1 November 1939 pp 171-180
Volume 43 Issue 3 April 1942 pp 301-307
The use of the standard error of estimates of mutation rates, and other small proportions, is shown to be highly erroneous. A table is provided for rapidly making the exact tests of significance.
Volume 100, 2021
Continuous Article Publishing mode
Click here for Editorial Note on CAP Mode