In this experiment, the attempt was to solve the problem of whether or not

 there was a relationship between the use of the hormone Bovine Somatotropin and

increases in SCC.  In all models used to solve this problem, the link was

conclusive.  The original model predicted an difference of 40,000 in the somatic

cell counts between the control and experimental groups.  The fact that the only

co-variable that was significant in predicting the response was the pre-test seems

to contradict research, although there is nothing in the literature that states

what factors effect how an animal will respond to BST.  It would be interesting to

measure other factors and judge how they effect the cow¹s response to BST.

Regression does seem to be the most effective analysis method available.  It

allows the researcher to control for extraneous variables, and this is especially

useful in large animal studies in which the sample size is usually small.

Additionally, the model can be manipulated to specifically fit the data available,

another useful property when analyzing zoological studies.  


Back to the Table of Contents

Next Section: References


Html by: Todd Gureckis