Agenda proposition

[didierbrassard]

Brief introduction on regression

• notation
• assumptions
• family/link functions
• relaxing assumptions (mixed models, non linearity, ...)

Marginal effect + demonstration and/or

1. regression with categorical covariate (ie, like t test)
2. regression with continuous covariate
3. regression with continuous covariate + flexible modelling (spline)
4. different family (eg, logistic regression)

At each step:

• Assumptions
• Parameters
• Visualization

Brief intro to issue of missingness

• brief introduction : why missing + why and when deal with missingness
• type of missingness
• intro to 2-3 type of way to deal with missing + pro/cons

multiple imputation + demo

• idea behind MI
• MI pitfalls
• hands on practice

[didierbrassard]

Example draft introduction document a880de9