Abstract
A procedure is presented which demonstrates estimation of and adjustment for residual effects in changeover designs. The method utilizes all data collected in an experiment by including treatments imposed on animals prior to initiation of data collection. Estimation is achieved via general linear models. An example is given of a nutrition experiment conducted with dairy cattle. Such analyses should increase efficacy of changeover designs and reduce concern by researchers about biased estimates of direct effects which could result from residual effects. Methods from popular computer programs for estimating direct effect treatment means are compared. Practical problems encountered in computing standard errors of mean estimates in mixed linear models.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Recommended Citation
Littell, Ramon C.; Wilcox, Charles J.; Van Horn, H. H.; and Tomlinson, A. P.
(1995).
"ESTIMATION OF AND ADJUSTMENT FOR RESIDUAL EFFECTS IN DAIRY FEEDING EXPERIMENTS UTILIZING CHANGEOVER DESIGNS,"
Conference on Applied Statistics in Agriculture.
https://doi.org/10.4148/2475-7772.1336
ESTIMATION OF AND ADJUSTMENT FOR RESIDUAL EFFECTS IN DAIRY FEEDING EXPERIMENTS UTILIZING CHANGEOVER DESIGNS
A procedure is presented which demonstrates estimation of and adjustment for residual effects in changeover designs. The method utilizes all data collected in an experiment by including treatments imposed on animals prior to initiation of data collection. Estimation is achieved via general linear models. An example is given of a nutrition experiment conducted with dairy cattle. Such analyses should increase efficacy of changeover designs and reduce concern by researchers about biased estimates of direct effects which could result from residual effects. Methods from popular computer programs for estimating direct effect treatment means are compared. Practical problems encountered in computing standard errors of mean estimates in mixed linear models.