Announcing a meeting of the Statistical Society of Australia, W.A. Branch.
6:00 ᴘᴍ (AWST) on Tuesday 9th June 2020
Zoom Online Meeting (Join from 5:45 ᴘᴍ for general socialising)
Generalizations of Orthogonal Components in Analysis of Variance
Dr Brenton R Clarke
Mathematics and Statistics, ITMAS, College of Health, Science, Engineering, and Education, Murdoch University
In this lecture I discuss the vector matrix approach to Analysis of Variance, beginning with a model for the two-way layout, and then introducing the orthogonal components. The linear model for the two-way layout is discussed, and the orthogonal components are displayed, using Helmert matrices and Kronecker products. These representations are elucidated in my book “Linear ModelsThe Theory and Application of Analysis of Variance”, which was published through the publishing house Wiley in 2008. The history and education of this approach for the two-way layout was detailed earlier in Clarke (2002). That paper was inspired by the work of Irwin (1934) who tried to give students of the then new field of Analysis of Variance an understanding of the components in ANOVA by writing out long hand the contrasts making up the independent components in an ANOVA for the randomised complete block design, and even the latin square. The approach in Clarke (2002) does this more succinctly using partitions of the Helmert matrix and uses matrix algebra. In the last months I was contacted by a mathematician Reza Farhadian who approached me regarding a generalization of the Helmert Contrasts. In Farhadian and Clarke (2020) we now have both the contrast approach and the matrix approach to generalize the Helmert Contrasts. I will illustrate the matrix approach in R. The ideas of generalizing the contrasts, for instance, in the area of residuals, is not new. There are the “BLUS” residuals and “recursive residuals”, which are all linked in terms of uncorrelated residuals. It can be noted that REML is the likelihood based on these error contrasts. A short discussion of these and their generalizations will ensue.
Clarke, B.R. (2002) A representation of orthogonal components in analysis of variance, International Mathematical Journal, 1, 133-147.
Clarke, B.R. (2008) Linear Models The Theory and Application of Analysis of Variance, Wiley, Hoboken, N.J.
Farhadian, R. and Clarke, B.R. (2020) A note on the Helmert transformation, Communications in Statistics- Theory and Method, submitted
Irwin, J.O. (1934) On the independence of constituent items in the analysis of variance, J. Roy. Statist. Soc., Suppl., 1, 236-251
ABOUT THE SPEAKER:
Brenton roamed the world when he was young, and settled in WA where he is supported by his wife and family. He is an experienced statistician, having studied or worked in 6 universities, not to mention various sabbaticals in various parts of the world. This is his 36th year working at Murdoch University.
ONLINE MEETING DETAILS:
This seminar will be presented online using Zoom (we recommend you download the Zoom App before the meeting start time). Once the seminar begins, participates will be asked to mute themselves. The meeting will be interactive, and viewers will be able to ask questions.
Instructions for connecting will be sent to your email upon registration. There is no close-off time for registrations though it is recommended to register in advance.
For further information please contact the Branch Secretary, Rick Tankard, Murdoch University.
He can be reached by email at firstname.lastname@example.org or by phone at (08) 9360 2820.