Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: Pstest
From
Isobel Williams <[email protected]>
To
"[email protected]" <[email protected]>
Subject
st: Pstest
Date
Sun, 2 Mar 2014 10:35:04 +0000
Dear All,
I am performing -pstest- on my propensity-score logit model (estimated with -psmatch2-), and I am having a lot of trouble interpreting the results.
Is the t-test or the reduction in bias (after matching) a better means of evaluating my model?
Does the t-test only refer to whether or not the two groups differ after matching? What is the null hypothesis of the t-test?
I have one value of bias reduction for -2696.0, what does that mean?
Is the bias reduction more important than the matched-bias? What is an acceptable level of bias?
Some matched bias levels are above 10%, and some t-tests are at 0.0000- does this mean I should take these variables out, or completely change the model?
For -psmatch2-, I am using a logit model with a caliper of 0.2 based on common support.
Any advice/suggestions would be very helpful!
Isobel
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/