|
|
|
Analytical Methods for Longitudinal Analysis – Tools and Analysis
|
Controlled interrupted time series designs, the strongest quasi-experimental designs, are crucial for policy
evaluation around the world. However, appropriate software to properly conduct segmented regression analyses
of interrupted time series studies remains too expensive and difficult to use for many researchers. In
particular, it is necessary to control for correlation between subsequent observations for arriving at
proper inferences from these study designs.
We have developed an Excel and R-based tool for performing segmented regression analyses of interrupted
time series studies. The R statistical package is freely-available on the Internet. While R can run appropriate
models, it is difficult to learn. We therefore have developed an Excel interface to the R-program. The
resulting analysis tool will use the familiar Excel spreadsheet data entry and a graphical user interface to
set statistical model parameters and obtain results from the R-program that will perform in the background.
A preliminary presentation that described development and use of the tool can be found
here. The program
and draft User’s Guide will be available soon.
|
|
LATEST UPDATES
MedIC
Initiative Courses conducted in Accra, Ghana
, November 16-25, 2008 and in Beijing, China, March 22-31, 2009
|
