Doctoral Dissertations

Date of Award


Degree Type


Degree Name

Doctor of Philosophy


Animal Science

Major Professor

H. V. Shirley

Committee Members

R. R. Shrode, R. L. Tugwell, L. M. Josephson


Data collected during the Twelfth, Thirteenth and Fourteenth Tennessee Random Sample Laying tests (1969-1971) were analyzed by com-puting conventional product-movement coefficients of correlation among production variables for each strain and treatment subclass in each test to determine the magnitude of these relationships. Then various coeffi-cients of correlation between income over feed and chick cost and other production variables for each strain, protein level and cage-density subclass in each test were computed and subjected to tests of homogene-ity. There were strain differences and treatment differences in many of the correlations calculated, as indicated by the tests of homogeneity. It was obvious from the results that both sample size and magnitude of correlations being compared play a part in determining the outcome of tests of homogeneity. Correlations larger than .9 which were very similar in value were frequently declared nonhomogeneous even when sample sizes were quite large. In many instances, smaller correlations were declared homogeneous, even though appreciably different from one another. Although comparison of correlation estimates was not a stated objective of these random sample tests, the results of this empirical study indicate that consideration of homogeneity (or lack of it) among separate estimates is essential if one is to draw valid conclusions concerning correlation between variables, especially if investigation of correlation is a major objective of an experiment with quantitative variables such as those represented in the data from random sample laying tests. No particular treatment combination, among those studied here, was consistently superior to any other in increasing correlations of income over feed and chick cost with other variables.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."