Standard deviations of yarn break data, taken by uniform increments of spindle hours, increased linearly with average rate of breakage and were larger than expected for a Poisson distribution. Transforming the numbers of breaks to log10 (1 + number of breaks) stabilized the variance. In a specific case, analysis of data for 720 spindle hours per yarn taken by 12 increments of 60 hours yielded 95% confidence limits describing the average number of breaks per yarn as within 30% of the observed number of breaks.
Get full access to this article
View all access options for this article.
References
1.
Barnes, J.C. , Elting, J.P., and Byatt, W.J., TEXTILE RESEARCH JOURNAL29, 844-845 (1959).
2.
Bartlett, M.S. , Biometrics3, 39-52 (1947).
3.
Li , Jerome, C.R., "Introduction to Statistical Inference," Ann. Arbor, Mich., Edwards Bros., Inc., 1957.
4.
Snedecor, G.W. , "Statistical Methods," 5th ed., Ames, Iowa, Iowa State College Press, 1956.
5.
Stewart, Ruby E., and Parsons, L.E., TEXTILE RESEARCH JOURNAL19, 283-287 (1949).
6.
Tallant, J.D. , Fiori, L.A., Little , H.W., and Leitz, L.A., Textile Inds.124, 1929-1933 (1960).
7.
Tallant, John D. and Fiori, Louis A., TEXTILE RESEARCH JOURNAL this issue, pp. 866-872.
8.
Woo, H.K.C., TEXTILE RESEARCH JOURNAL28, 612 (1958).
