I am not clear what precisely you seek. This example may or may not
help. The idea is to work with a composite variable formed by -egen-'s
-group()- function. If it helps, a bonus is that it can be applied to
two or more variables.
Nick
. webuse nlswork
(National Longitudinal Survey. Young Women 14-26 years of age in 1968)
. xtset id year
panel variable: idcode (unbalanced)
time variable: year, 68 to 88, but with gaps
delta: 1 unit
. xttab msp
Overall Between Within
msp | Freq. Percent Freq. Percent Percent
----------+-----------------------------------------------------
0 | 11324 39.71 3113 66.08 62.69
1 | 17194 60.29 3643 77.33 75.75
----------+-----------------------------------------------------
Total | 28518 100.00 6756 143.41 69.73
(n = 4711)
. xttab race
Overall Between Within
race | Freq. Percent Freq. Percent Percent
----------+-----------------------------------------------------
1 | 20180 70.72 3329 70.66 100.00
2 | 8051 28.22 1325 28.13 100.00
3 | 303 1.06 57 1.21 100.00
----------+-----------------------------------------------------
Total | 28534 100.00 4711 100.00 100.00
(n = 4711)
. egen both = group(msp race), label
(16 missing values generated)
. xttab both
Overall Between Within
both | Freq. Percent Freq. Percent Percent
----------+-----------------------------------------------------
0 1 | 6853 24.03 2046 43.43 57.56
0 2 | 4359 15.29 1036 21.99 72.65
0 3 | 112 0.39 31 0.66 68.24
1 1 | 13321 46.71 2747 58.31 78.32
1 2 | 3682 12.91 853 18.11 67.09
1 3 | 191 0.67 43 0.91 83.36
----------+-----------------------------------------------------
Total | 28518 100.00 6756 143.41 69.73
(n = 4711)
Thomas Klausch wrote:
Thanks Martin, this kinda works. However it is still a bit of a
nuissance. I take that there is no direct command such as xttab, but
just made for 2-way tables?
2009/8/3 Martin Weiss <[email protected]>:
Why not -collapse- the thing beforehand? If the characteristics are
time-invariant, nothing gets lost this way. Then use -tab- as in
*************
help tabulate twoway
*************
with all the tests available...
Thomas Klausch
I am trying to produce a 2 way table for time invariant
characteristics in panel data. For a one way table this can be done
using xttab as I understand. However is there a command for producing
a 2 way contingency table alongside suitable tests? Or do I need to
agregate data to the macro level before I can do this (which would be
inconvenient I suppose)?
*
* 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/
