Weeks 1 to 5

Home / Weeks 1 to 5

Weeks 1 to 5

Day, Time & Location

Lecture 1: Tuesday 10:00-10:50, Teviot Lecture Theatre (Doorway 5), Medical School.

Lecture 2: Wednesday 11:10-12:00, Lecture Theatre 4, Appleton Tower

1
Simple regression
This lecture will recap simple regression with a single continuous, and single binary predictor.
2
Multiple regression
In this lecture we will extend the regression model to include multiple independent variables, or predictors, and discuss the additional aspects of model interpretation that come with this.
3
Multiple regression
In this lecture we will continue from the last, discussing multiple regression.
4
Categorical predictors in regression models
Up to this point we have been focused on models including continuous predictors. In this lecture, we will begin from single binary predictors, and introduce the way in which we include categorical variables with 2+ levels. We will discuss dummy and effects coding, and discuss the differences with worked examples.
5
Model Assumptions
In this lecture we will recap the set of linear model assumption, introduce those specific to models with more than one predictor, and look at some examples.
6
Model Diagnostics
In this lecture we will recap the set of linear model assumption, introduce those specific to models with more than one predictor, and look at some examples.
Lecture Theatre 2, Appleton Tower
7
Model Selection
When we add multiple predictors to our models, we enter a situation where we may have a set of models with different sets of predictors. In this lecture we will discuss how to choose which of these models is the best representation of our data.
8
Interactions: Continuous*categorical
In the next three lectures we will discuss interactions between different types of variables. We will begin by discussing categorical by continuous variable interactions via a worked example.
9
Interactions: Continuous*continuous
In this lecture we will extend the previous discussion to continuous by continuous variable interactions, and discuss in more detail different types of interaction, and methods to probe identified interactions in order to interpret them more completely.
10
Interactions: Categorical*categorical (dummies)
In the final of three lectures, we will discuss categorical by categorical variable interactions. In the first instance we will focus on dummy coded variables, before comparing this to effects coded variables. This will lead us into a discussion of the relation between ANOVA and regression in the next lecture.