Stata for Regression Analysis

Prerequisites

This course is for new users who have learnt how to get data into Stata, and know how to operate basic syntax having completed the Basics course or through self-study with the student version or manuals. Emphasis will be on examples of running applied analyses of regression models for continuous, binary and ordinal outcomes using standard Stata procedures. Guidance will also be provided on further addons that may be of interest.

• Working knowledge of Stata basics as covered on the Part 1 & 2 course.
• Familiarity with simple statistical concepts.

Number of sessions: 2

• Linear regression for continuous outcomes
• Bootstrapping standard errors
• Logistic regression
• Logistic regression for clustered data
• Ordinal regression for multi-category outcomes that can be ordered
• Logit and probit models
• Nominal regression (in brief) for unordered multicategory outcomes
• Categorical covariates using xi:

Presentations, demonstrations and practicals.

Stata IC v. 10 on PWF Windows

• This course does include some guidance on statistics, not purely software use.
• Practicals will be run on PCs using Stata for Windows.

Two half day sessions

• Scientific Computing
• Statistical and Mathematical Software