Bookmark and Share

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: FE interaction model and dropped variable


From   Enrico Cavale <[email protected]>
To   [email protected]
Subject   st: FE interaction model and dropped variable
Date   Thu, 24 Mar 2011 23:07:59 +0100

Hello everybody,

I use daily-individual level data to measure the impact of a
treatment,  essentially  comparing  individual  behavior  during
treatment days to behavior during non-treatment days. Variation in the
timing of treatment days across different groups of individuals allows
to estimate models with calendar-date fixed effects to control for
day-specific attributes.

I want to estimate an interacted model with individual_by_year FE and
calendar_day FE where the regressors are the interactions of the
treatment dummy with month dummies. I had some memory issue of which
some of you are aware for a recent post "double format and memory
issue". So, I estimated a model with individual_by_year FE and
Year_by_week FE and stata returned to me the estimation for the 12
interactions variables. Later, following the advice of  Christopher
Baum in the other post, I could estimate my model with calendar_date
FE by first taking out the calendar_date means. The problem is that
one of the variable has been dropped. Do you have an idea of what is
going on?

Thanks,
Enrico

*
*   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/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index