|
|
Hi,
I am interested in comparing results from an intent-to-treat (ITT)
analysis to results from an analysis that estimated the complier average
causal effect (CACE).
The test statistic I used to obtain the CACE effect is as follows:
CACE =the proportion of compliers in the treatment group* times the
complier treatment group mean + the proportion of non-compliers in the
treatment group * the non-complier treatment group mean - the control
group mean / by the proportion of compliers in the treatment group.
I would like to estimate the standard error for this statistic to
evaluate significance and found a reference to estimating SEs using the
delta method or bootstrapping techinques but I'm not familiar with how
to do this exactly.
If anyone has estimated the S.E. for the CACE and has syntax for this I
would very much appreicate it. Any other help or suggestions on how to
to this are also highly appreciated.
Thanks in advance,
Nicki
***********************************************************************
This email may contain confidential material.
If you were not an intended recipient,
please notify the sender and delete all copies.
We may monitor email to and from our network.
***********************************************************************
>>> Error in line 4 of spssx-l.mailtpl: unknown formatting command <<<
-> .................... <-
|
Locked
