st: Causal Inference course in Swansea October 2012

[<Sarah.Miller@lshtm.ac.uk>]

Two-day course: 'A short course on Concepts and Methods in Causal
Inference'
 
-Dates: 11th and 12th October 2012
-Times: 9.00am – 5.30pm on both days
-Location: Swansea University, Wales
-Bookings: http://www.ncrm.ac.uk/training/show.php?article=3812 
-Course leaders: Bianca De Stavola, Rhian Daniel and Richard Silverwood
from the PATHWAYS node of the ESRC National Centre for Research Methods
(NCRM), London School of Hygiene and Tropical Medicine
 
-Project: http://pathways.lshtm.ac.uk/ 
-Contact: pathways@lshtm.ac.uk 
-Twitter: @pathwaysNCRM
 
At whom is the course aimed?
The course is aimed at researchers working in the social or medical
sciences. No prior knowledge of causal inference methods is assumed, but
familiarity with standard statistical techniques such as linear and
logistic regression is required. The computer practicals will be carried
out using the Stata software, familiarity with which would be highly
beneficial.
 
Course overview
Lecture 1 gives an introduction to key concepts in causal inference: to
the language of counterfactuals (part A) and to causal diagrams (part
B), and is accompanied by a pen & paper practical. Lectures 2-5 cover in
more detail particular statistical methods used in causal inference,
grouped according to the sort of causal question they address, and the
structure of the data available to answer them. Each of these lectures
is followed by a computer practical using Stata.
The two days split into `simple’ causal questions (on the first day)
and `complex’ causal questions (on the second). By `simple’ causal
questions, we mean questions of the type `what is the causal effect of a
single exposure A, such as educational achievement, on a single outcome
Y, such as blood pressure?’. By `complex’, on the other hand, we
mean causal questions concerning the effect of time-changing exposures,
and questions concerning pathways, such as how much of the causal effect
of A on Y is mediated by a third variable M?
 
 
 
(Thank you to Nick Cox, of Durham, for suggesting statalist)
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/