Statistical Modelling with Stata

This course was originally written to teach researchers to analyse data using Stata, which was the analysis program of choice in the Centre for Epidemiology for many years. We are now encouraging the use of R within the Centre in parallel with Stata, and David Selby has kindly rewritten all of the Stata practicals so that they can be run in R. This is still a work in progress: not all Stata concepts map cleanly to R versions, so both the Stata and R versions will need some editing to produce something that is both meaningful and identical in both cases.

If you are following the course remotely, you may want to use the University's Remote Access PCs to run Stata on. There are instructions for doing this here, although they will obviously only work if you have a valid IT account with the University of Manchester.

Session Number	Title	Content	Lecture	Slides	Handout	Solution	Solution Do-file	R Version of Practical	Solution to R Practical	Datafiles	Further Reading
1	Introduction to Stata	Stata's windows Command Syntax	Lecture	Slides	Handout & Practical	Solution	Do file to produce solution		Solution Using R
2	Summarising Data	Types of data Graphical Summaries Numerical Summaries	Lecture	Slides	Handout Practical	Solution	Do file to produce solution	Practical Using R	Solution Using R	Height and Weight Dataset
3	Sampling and Confidence Intervals	Types of sampling Estimation from random samples Sampling Error Reference Ranges Confidence Intervals Sample Size	Lecture	Slides	Handout Practical	Solution		Practical Using R	Solution Using R		The tyranny of power: is there a better way to calculate sample size? Uncertainty beyond sampling error
4	Hypothesis Testing and Power	Hypothesis Tests Power calculations	Lecture	Slides	Handout Practical	Solution	Do file to produce solution	Practical Using R	Solution Using R	Height and Weight Dataset	Statistical tests, P values, confidence intervals, and power: a guide to misinterpretations The ASA's Statement on p-Values: Context, Process, and Purpose Scientific method: Statistical errors Why We (Usually) Don't Have to Worry About Multiple Comparisons The Harm Done to Reproducibility by the Culture of Null Hypothesis Significance Testing Some problems of Statistical consultancy [stars] are more appropriate to a hotel guide book than a serious scientific paper
5	Linear Models 1	Assumptions Interpretation Inference Goodness of Fit Diagnostics	Lecture	Slides	Handout & Practical	Solution	Do file to produce solution	Practical Using R	Solution Using R	Anscombe's Data constvar.dta Housing Data Lifeline Data Wood's '73 Data	Graphs in Statistical Analysis
6	Linear Models 2	Categorical variables Interactions Confounding Variable Selection	Lecture	Slides	Handout & Practical	Solution	Do file to produce solution	Practical Using R	Solution Using R	Cadmium Data Growth Data Hald Data Soap Data	Erroneous analyses of interactions in neuroscience: a problem of significance The difference between "significant" and "not significant" is not itself statistically significant.
7	Modelling Binary Outcomes	Limits of Linear Regression Generalised linear models Logistic Regression Logistic Regression Diagnostics Sensitivity and specificity Alternative Models	Lecture	Slides	Handout & Practical	Solution	Do file to produce solution	Practical Using R	Solution Using R	CHD dataset Pain dataset	EXPLAINED VARIATION FOR LOGISTIC REGRESSION
8	Modelling Categorical Outcomes	Nominal Outcomes Multinomial Regression Lincom Ordinal Outcomes	Lecture	Slides	Handout Practical	Solution	Do file to produce solution	Practical Using R	Solution Using R	Alligators Data Housing Data Politics Data
9	Modelling Counts	Poisson Regression Constraints Overdispersion Negative Binomial Regression	Lecture	Slides	Handout Practical	Solution	Do file to produce solution	Practical Using R	Solution Using R	Alligators Data Negative Binomial Regression Data Ships Data
10	Survival Analysis	Censoring Survival Curves and Life Tables Comparing Survival Curves Parametric Regression Cox Regression	Lecture	Slides	Handout Practical	Solution	Do file to produce solution	Practical Using R	Solution Using R	Leukaemia Data	Motion Sickness Incidence Paper
11	Refinements of the Stata Language	Graphics Summarising Data More Syntax Looping Reshaping data	Lecture	Slides	Handout & Practical	Solution	Do file to produce solution	Practical Using R	Solution Using R

Exam 2008
Exam 2009
Exam 2010
Exam 2011