Our fifth annual Stata Winter School comprises a series of four separate short courses that allows the flexibility to attend one, a combination of or all courses consecutively.
The courses forming the 2019 Stata Winter School are:
Click here to view the complete course information, schedule information, and suggested pre-course reading list.
9 & 10 December 2019
This two-day course assumes no or little prior experience with Stata. We therefore start right at the beginning, with a gentle introduction to Stata using the Graphical User Interface. The focus of this course is on learning how to use Stata to produce a clean, complete, analysis ready dataset. Along the way we will learn how to load data into Stata from different formats, familiarise ourselves with our data, correct errors, combine datasets, and generate new variables. Throughout the course we emphasise good research practice with the aim of enabling you to produce reliable and reproducible results. The course is applied and hands-on, with students following along from-the-front demonstrations and with plenty of opportunities for questions. The course will be delivered in eight 1.5 hour sessions over the two days. A comprehensive set of course notes will be given to all participants along with data used during the course and a set of example do-files.
We start with a brief introduction to the course content and aims and to the Stata interface. We begin to learn how to use Stata through the Graphical User Interface (GUI). The GUI is a user-friendly way of working in Stata using drop-down menus and dialog boxes. As we work through a series of exercises we review the command syntax and begin to learn the structure of Stata commands. We also cover interpreting error messages and how to use Stata’s help facilities. At the end of the session we learn how to save commands that have been submitted so that they can be reused late.
One of the most important stages of any statistical analysis is that of getting to know your data. We cover a number of key commands for exploring datasets and for summarizing variables of different types. We particularly focus on understanding what a dataset contains, the distributions of variables and identifying potential errors – which we will then see how to correct in later sessions.
In this session we learn how to import data from an Excel file into Stata and how to then save that data as a Stata dataset. Often in a research project you will be required to combine data from different files into a single dataset for analysis – we introduce two key commands for carrying out such a task. The goal here is to produce a single analysis dataset.
Housekeeping is the process of creating and maintaining a tidy, user-friendly dataset. Taking time to do this pays dividends later when you come to the analysis. We cover labelling variables, labelling the values of a variable, dropping unwanted variables and naming of variables.
This is another key stage in any statistical analysis. We will introduce commands for creating new variables (sometimes called derived variables) and for correcting errors and modifying existing variables. In this session we will focus on commands for dealing with numeric variables.
In this session we will cover some of Stata’s commands and functions for dealing with string variables, i.e. variables that consist of non-numeric values. We will also spend some time dealing with dates and learning how to covert a date that is in human readable form into a date that Stata can use.
In this final taught session we will cover the creation of summary datasets (often useful when needing to produce a graph) and reshaping of datasets.
The final session consists of a data management challenge designed to help reinforce what has been taught over the previous seven sessions. The challenge will require importing and combining a number of Excel worksheets, checking and creating some new variables and producing a clean dataset ready for analysis.
11 December 2019
This one-day course assumes no prior experience with Stata, though some experience would be helpful. The focus of this course is on learning how to use Stata efficiently to visualise data effectively in figures and tables. The course is applied and hands-on, with students following along from-the-front demonstrations. There will be a comprehensive set of course notes to take away, along with data from the course and example do-files.
In this first session we work interactively, creating graphs using the Graphical User Interface (GUI). We will start simply and gradually build a graph adding in options to create a publication ready figure. We cover box-and-whisker plots and also Stata’s twoway family of graphs which allow overlaying of plots to produce complex figures. Once we are happy with the graph we will learn how to take the command generated through the GUI and save it in a do-file ready for later use.
In this session we focus on building graph commands within do-files. We will learn how to lay out a long command in a do-file and how to then reproduce the graph. Being able to do this helps you to work much more efficiently – graphs can easily be reproduced, edited, and old graph commands can be recycled to produce new figures. We will learn a number of key options for improving the look of a graph and will also briefly visit Stata’s built Graph Editor which allows interactive editing of a Stata graph. We will also cover exporting graphs from Stata into a number of different formats including producing high resolution images.
In this session we move from figures to tables. We cover one-way and two-way frequency tables, as well as tables of summary statistics. This will be demonstrated using the GUI and through the command syntax working in a do-file. We will cover saving results in logfiles and copying tables from Stata into Word and Excel and show some tips for producing clear tables. There is a brief introduction to exporting results directly to Excel using the
In this final supervised session there will be a number of graph or table challenges that will help reinforce what has been covered in the first three sessions. Students usually pick one or two of the challenges to work on through the session.
12 & 13 December 2019
Saturday, 14 December 2019
This course is aimed for researchers from any field, with a basic knowledge of Stata, who are interested in present results effectively from any regression model-fitting using Stata commands designed for this purpose.
Researchers from any field start to squirm when asked to give a simple explanation of the practical meaning of a non-linear relationship and/or an interaction from any type of regression model. Michael Mitchell’s book titled ‘Interpreting and Visualizing Regression Models Using Stata’ presented techniques that make answering those questions easy. This one-day course is mainly based in Mitchell’s book to introduce the way to present results from any model-fitting in a wide variety of settings in practice using the Stata commands
|Time||Session / Description|
|08:50 - 9:20||Arrival and Registration|
The number of seats available is restricted. Please register early to guarantee your place.