Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: XTSet: Imputed longitudinal panel data

From   "Castle, Melissa H" <[email protected]>
To   "[email protected]" <[email protected]>
Subject   st: XTSet: Imputed longitudinal panel data
Date   Sun, 9 Jan 2011 15:57:10 -0500

Hi Stata-listers,

I¹m having a tough time setting my imputed dataset as time-series panel
data, and would greatly appreciate your help. Some background information:

I¹m working on a dataset that contains time-series panel data.  I recently
imputed this data in wide form using the MIM command in STATA 10 to create
10 imputed datasets.  I then used the ­reshape- command to convert it back
into long-form.  I need to STATA to recognize that my data is longitudinal
panel data.   The command I usually use for this is:
xtset c_id timevar

This usually works perfectly... But seems to be a little trickier when using
an imputed dataset.  When I use the above command, I get an error ³repeated
time values within panel².  I suspect this is because my panel-time variable
repeats for each iteration of imputations:
C_id    _mj     timevar
1       1        1
1       2        1
1       3        1
1       1        2
1       2        2
1       3        2
1       1        3
1       2        3
1       3        3
(Sample dataset: _mj = imputed dataset identifier).

So, I have a few questions:
* It seems that using XTSet isn¹t supported by MIM in Stata 10.  Is there
another way?
* The interwebs seem to suggest that this might be possible with the mi
command in Stata11, but I have yet to see a command implementing a panel and
time-series identifier.  Is this possible in Stata11?
* I tried to perform the command on each imputed database with an if
statement (xtset c_id timevar if _mj=1), but it says ³if not allowed². Is
there another way to apply a filter to the data?  Is this even a reasonable

I (and my thesis defense) would be wonderfully grateful for any help.


Melissa H. Castle
Doctoral Student
Clinical/Developmental Psychology
University of Pittsburgh

*   For searches and help try:

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