Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: xtlogit and number of observations


From   Traci A Schlesinger <tracias@Princeton.EDU>
To   statalist <statalist@hsphsun2.harvard.edu>
Subject   st: xtlogit and number of observations
Date   Wed, 14 May 2003 11:06:03 -0400

Fellow Statalisters:

i tried to post this several hours ago, but it still hasn't come up, so i've decided to try again.

I am using xtlogit to analyze pretrial release decisions, such as whether someone is denied bail or given a nonfinancial release. I have 5 years of individual level data which includes all individuals arrested for felonies in 54 counties. I am using a county level fixed effect. when i aggregate all 5 years together, the regression results that are returned to me are filled with dots where there should be numbers. however, when i run the regressions by year, i get normal results. or, if i run all 5 years together but by race group or crime type, i get normal results. In fact, it seems if i partition the data in any way that brings the number of observations below 40,000, i get results but if the number of observations is above 40,000, i get dots. Of course, other commands can handle more observations than this -- for instance, if i use xtreg to run by equation with all years, all race groups, and all crime types, i get sensible results. Does xtlogit have a lower limit to the number of observations it can handle?
if anyone could help be figure out what is going on -- or even better, how to fix it -- please let me know!
i've pasted the relevant parts from my log file below.
thanks!!!
traci

> xtlogit dbail black latino age agesq pripris prijail prifconv privconv pridco
> nv cjstatus priorfta chg1att totchgs1 murder rape robbery assault oviolent bu
> rglary theft oproperty drugt odrug weapons driving, fe i(county) or;

note: multiple positive outcomes within groups encountered.
note: 2 groups (451 obs) dropped due to all positive or
all negative outcomes.
Iteration 0: log likelihood = -8.988e+307

Conditional fixed-effects logit Number of obs = 40593
Group variable (i) : county Number of groups = 52

Obs per group: min = 109
avg = 780.6
max = 4992

LR chi2(25) = .
Log likelihood = . Prob > chi2 = .

------------------------------------------------------------------------------
dbail | OR Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
black | . . . . .
latino | . . . . .
age | . . . . .
agesq | . . . . .
pripris | . . . . .
prijail | . . . . .
prifconv | . . . . .
privconv | . . . . .
pridconv | . . . . .
cjstatus | . . . . .
priorfta | . . . . .
chg1att | . . . . .
totchgs1 | . . . . .
murder | . . . . .
rape | . . . . .
robbery | . . . . .
assault | . . . . .
oviolent | . . . . .
burglary | . . . . .
theft | . . . . .
oproperty | . . . . .
drugt | . . . . .
odrug | . . . . .
weapons | . . . . .
driving | . . . . .
------------------------------------------------------------------------------

. xtlogit dbail black latino age agesq pripris prijail prifconv privconv pridco
> nv cjstatus priorfta chg1att totchgs1 murder rape robbery assault oviolent bu
> rglary theft oproperty drugt odrug weapons driving if year == 1990, fe i(coun
> ty) or;

note: multiple positive outcomes within groups encountered.
note: 10 groups (514 obs) dropped due to all positive or
all negative outcomes.
Iteration 0: log likelihood = -1025.0744
Iteration 1: log likelihood = -850.28866
Iteration 2: log likelihood = -845.87709
Iteration 3: log likelihood = -845.86588
Iteration 4: log likelihood = -845.86588

Conditional fixed-effects logit Number of obs = 4163
Group variable (i) : county Number of groups = 28

Obs per group: min = 41
avg = 148.7
max = 690

LR chi2(25) = 420.16
Log likelihood = -845.86588 Prob > chi2 = 0.0000

------------------------------------------------------------------------------
dbail | OR Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
black | 1.395709 .2225102 2.09 0.037 1.021157 1.907643
latino | .9708441 .1930946 -0.15 0.882 .6574323 1.433666
age | 1.088391 .0428947 2.15 0.032 1.007484 1.175795
agesq | .9988118 .000582 -2.04 0.041 .9976716 .9999532
pripris | 1.314163 .0872487 4.12 0.000 1.153817 1.496791
prijail | .9792627 .0270809 -0.76 0.449 .9275979 1.033805
prifconv | 1.032692 .0390599 0.85 0.395 .9589049 1.112157
privconv | 1.003938 .0703026 0.06 0.955 .8751852 1.151632
pridconv | .9611685 .0638591 -0.60 0.551 .8438136 1.094845
cjstatus | 4.67771 .7036319 10.26 0.000 3.483322 6.281639
priorfta | 1.460088 .2553298 2.16 0.030 1.036399 2.056985
chg1att | 1.968277 .5003353 2.66 0.008 1.195945 3.239376
totchgs1 | 1.062049 .0453806 1.41 0.159 .9767272 1.154824
murder | 18.71818 9.604054 5.71 0.000 6.847358 51.16866
rape | 1.443282 .7784431 0.68 0.496 .5014762 4.153862
robbery | 2.036161 .8609694 1.68 0.093 .8889769 4.663734
assault | .9645549 .4031453 -0.09 0.931 .4251662 2.188241
oviolent | .5543121 .3317127 -0.99 0.324 .1715443 1.791151
burglary | .8193079 .3439418 -0.47 0.635 .3598458 1.865425
theft | .8282786 .343824 -0.45 0.650 .3671445 1.868598
oproperty | .4920875 .2233401 -1.56 0.118 .2021686 1.197763
drugt | .7131889 .2894573 -0.83 0.405 .3219089 1.580069
odrug | .5001843 .2156836 -1.61 0.108 .214825 1.164596
weapons | .3495524 .2299102 -1.60 0.110 .0963061 1.268735
driving | .5510242 .3453503 -0.95 0.342 .1613193 1.882153
------------------------------------------------------------------------------



*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/




© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index