Statistical analysis methods for avian reproduction experiments / D.A. MacLeod. : CW69-5/211E-PDF

"The situation with respect to the statistical analysis of data for avian reproduction experiments is a fairly confused one. An examination of the submissions from approximately 100 experiments revealed that a wide variety of different methods are employed, and that many of these are questionable. Assumptions are made that are not statistically valid, and tests are carried out that are inefficient at detecting the kind of treatment effects that are expected. These problems reduce the ability of the experiments to assess whether the test substances have the potential to cause reproductive effects. It was assumed that when a data set is analysed, the effect of the test substance should be tested at each dose level. The main objective of this report is to identify methods that are statistically valid and fully efficient at carrying out these tests"--Summary, p. i, ii.

Lien permanent pour cette publication :
publications.gc.ca/pub?id=9.857528&sl=1

• Format MARC XML
• Format MARC HTML

Demander des formats alternatifs