The data obtained on the Walker yarn abrader in the determination of abrasion-resistance of yarns form frequency distributions which are decidedly positively skewed. The mean of such distributions, being affected by the extremes, tends to be atypical. The modal class is too broad for pre cise differentiation and the first breaks are too variable to be of use. The median, which is independent of the ex treme values, is more representative. Furthermore, by com puting the average median of several sets of yarns rather than determining the lot median, a saving of time of approxi mately 50 percent is effected.
Get full access to this article
View all access options for this article.
References
1.
"A.S.T.M. Manual on Presentation of Data," Philadelphia , American Society for Testing Materials, April, 1943, p. 1.
2.
Worthing, A.G., and Geffner, J., "Treatment of Experimental Data," New York, John Wiley and Sons, 1943, pp. 177 ff.
3.
Davies, G.R., and Crowder, W.F., "Methods of Statistical Analysis in the Social Sciences," New York, John Wiley and Sons, 1933 , p. 303.
