2025 Data Science Winter School

Date 8 December 2025
Delivered byTim Collier, LSHTM
Prerequisitesnone

This one-day introductory course is for people interested in using Stata effectively in their research. This course requires no prior knowledge of Stata but assumes an interest in research and in learning to use Stata efficiently.

Throughout the course there will be an emphasis on good practice for research, understanding work flow, good data management, efficiency and reproducibility of results.

This course introduces the Stata working environment and explains and compares the two main ways of working in Stata. We then introduce some essential tools for data management – getting data into Stata from Excel, merging datasets and creating new variables. The afternoon sessions focus on essential tools for descriptive statistics, common hypothesis tests, and a quick introduction to regression modelling (linear and logistic regression). 

The examples used throughout will be from the field of medical statistics. However, the underlying principles will have application across all areas of research. Participants will be able to take away a comprehensive set of course notes and data used on the course as well as files created throughout the day. There will be a Q&A session at the end of the course, but also opportunities to ask questions throughout the day.

Course Outline

We will cover:

• Introduction to the key Stata windows
• Working interactively in Stata via the Graphical User Interface
• Exploratory data analysis: getting to know your data and identifying errors using statistical and graphical summaries
• Saving commands and results
• Essential data management tools: importing and saving data, generating new variables, correcting errors, keeping data tidy

Learning Objectives

By the end of this course you will:

• be familiar with the main Stata windows;
• understand how to work in Stata via the Graphical User Interface and do-files and the advantages and disadvantages of the two approaches;
• be able to use a small toolkit of commands to carry out exploratory data analyses;
• be able to create and save Stata datasets;
• be able to create new variables and correct errors;
• understand how Stata handles data and the importance of good practice for data management;
• know how to use Stata’s online help facilities so that you will be able to continue learning beyond the course

Date 9 June 2025
Delivered byThomas Pical, Equancy
Prerequisitesnone

This course will introduce the basics of the Stata and Python language.

Python is a widespread and powerful programming language. There are many free libraries in different fields including statistics, econometrics, web scraping, machine learning, etc.

Python can be called from a running Stata session so that the extended functionality of the Python language can be exploited from within Stata. We call this Python integration, which was introduced in Stata 16. With this integration you can embed and run Python code interactively or in do-files and ado-files.

Course Outline

We will cover:

• Introduction to Python and a reminder of the basics
• Various ways to call Python from Stata, store information and load variables into Python
• Calling Stata from a Jupyter Notebook

Date10-11 June 2025
Delivered bySebastian Laurent, Lancaster University
PrerequisitesSome familiarity with Stata is desirable

The aim of this two-day course is to introduce participants to machine learning, a relatively new approach to data analytics at the intersection between statistics, computer science, and artificial intelligence.

Students will be taught how to master the theory and the techniques that allow turning information into knowledge and value by 'letting the data speak'. The teaching approach will be based on the graphical language and intuition more than on algebra. The course will make use of instructional as well as real-world examples, with a balance of theory and practical sessions using Stata.

Learning Objectives

By the end of this course you will have knowledge and understanding of:

• Implementing and optimising machine learning approaches
• Assessing model performance
• Selecting key features
• Using standard machine learning libraries

• Comparing estimators for static panel models for your research question 
• Testing for serial correlation

• The Arello Bond estimator and post-estimation diagnostic test 
• The Blundell Bond estimator and post estimation diagnostic tests
• Case study: the determinants of bank risk-taking in European banks.