Short course at London School of Hygiene & Tropical Medicine
Advanced Stata: Programming and other techniques to make your life
easier
For details please go to:
http://www.lshtm.ac.uk/prospectus/short/sasta1.htm
The aim of the course is to teach competent Stata users the techniques
that allow you to get the most out of Stata and speed up the output of
your work.
As well as being a powerful tool for statistical analysis, Stata offers
a variety of commands for manipulating your data and for formatting,
arranging and exporting your results.
The course is aimed at researchers and other professionals, from any
discipline, who regularly use Stata for analysis but want to learn how
to work more efficiently. It would be particularly suited to those who
are about to embark on large analyses and who would like a quick guide
on how to automate the repetitive parts of the process.
The course is taught by research staff from the Centre for Population
Studies, who regularly use Stata for large-scale analyses using multiple
data sources. The examples used in the course are drawn from the
background of the tutors and are, therefore, from the population and
health sciences. However, none require any specialist knowledge of the
field.
Much of the material in the course has been developed with students and
staff from the UK and overseas. Most teaching is hands-on, using Stata
to tackle a series of exercises designed to illustrate the use of
particular commands in order to solve a variety of problems. These
exercises are supplemented by short lectures and a very comprehensive
set of notes. There is a strong emphasis throughout on providing
information that can be built on to tackle new problems and to be
applied in different situations.
We use Stata 10 in a Windows environment; users of other operating
systems should note that, although almost everything is the same, there
are some differences between operating systems and these are not covered
in the course.
Who Should Attend
The course is designed for people who want to be more efficient in
their use of Stata. Those who are already experienced in using Stata for
data analysis will benefit most from the course. As a minimum, you
should be able to use Stata for an analysis of some sort (linear
regression, for example) and be comfortable writing comprehensive do
files.
If you are already familiar with the merge, collapse, reshape and
append commands, have used foreach or forvalues loops and can understand
a simple Stata program, then this course may be too basic for you. If,
on the other hand, none of that sounds familiar, then this could be
exactly what you need!
If you have any doubt about whether the course would be suitable for
you, please do get in touch with the organisers who will be able to
advise you further.
Applicants should have a good command of English.
Course Content
The course can be divided broadly into three sections:
Data handling and manipulation
Stata has some powerful but simple commands for managing and
manipulating data. We cover the commands needed to combine data from
different datasets (appending and merging) and to rearrange data from
wide format to long format. We will also cover searching for duplicate
records, and managing these, reordering the variables in the dataset,
generating summary variables and summary datasets.
Accessing and outputting results
One of the commonest complaints about Stata is the difficulty of
producing well-formatted results output. The output on the screen
typically contains more detail than is required, and the formatting is
often sub-standard. Cutting and pasting is tedious when there are many
results to present and inefficient when analyses have to be repeated.
It is possible to automatically output well formatted, concise and
relevant results from Stata. You can set up do files which write the
results you want to the screen, or external files, in the required
format. These results files can then be automatically updated each time
the analysis is repeated. There are several ways to do this using either
additional user-written Stata commands or with some simple programming.
This course introduces both approaches.
Programming Stata
Data cleaning, data management and the initial stages of many analyses
can be repetitive and time consuming. Many of these repetitive tasks can
be automated in Stata, which not only speeds up the process but also
reduces the chances of making an error. We cover the use of basic
programming techniques to assist you in quickly and efficiently carrying
out repetitive tasks.
Aims & Objectives
By the end of the course you should be able to:
1. Generate variables that contain summaries of the data
2. Create a summary dataset using collapse
3. Navigate a dataset using _n, _N and subscripted variables
4. Rearrange a dataset using reshape
5. Combine multiple datasets using merge and append
6. Identify duplicate observations
7. Export data to a spreadsheet
8. Create tailor-made publication quality graphs
9. Understand what macros and scalars are
10. Be able to use foreach and forvalues loops
11. Understand and use if statements
12. Understand how Stata stores estimation results
13. Be able to access and use stored estimation results
14. Know how to export results using user-written commands: estout,
outreg, tabout
15. Understand how Stata programs work
16. Be able to write and use a simple Stata program
17. Be able to write a do-file which exports results using file write
|