These are occasional lectures on miscellaneous topics. The topic may occupy one lecture but some may require two. The topics are a personal selection and represent things which are perhaps not so thoroughly covered in the usual statistical curricula, or where I think a different emphasis may pay dividends.
The treatment will often be quite mathematical and this is deliberate. While our subject is based in mathematics, everyday practice in medical statistics can leave the practicing biostatistician distanced from their roots. Even those studying for a PhD will often focus on a few relevant areas and mathematical techniques. These lectures will be a refreshment for the soul.
The first few topics, and possibly subsequent ones, will be quite old-fashioned. While not deliberate this has not been avoided. Some things which have been around for years, and proved their value, are now so much part of the furniture that a refresher on their background is not without merit.
The lectures will be in-person and each set of slides will be accompanied by a document that provides more details and a more thorough narrative.
While this is apparently a straightforward business, I think that standard justifications often miss the real benefit of the procedure, namely the effect on interpretation.
There are many kinds of residuals for Normal-theory models and most of them are names on the help screen of many statistical packages. The background, derivation and uses of these are worth revisiting.
The Normal plotting of residuals was omitted from Topic 2, as a thorough treatment deserves its own topic.