Nick I was using -storecmd- to do -xtintreg- but I obtained slightly
different results after the -storecmd- statement and then if I repeat the
analysis with a straight command. For some reason, with "storecmd", the
integration method used is
"ghermite" (not the default), whereas with the straight command, it
correctly uses the default "aghermite". But I never issued an option for
"ghermite" so what's going on?
Al
. storecmd INTREG8 xtintreg zmiss z ind2 ind3 ind4 if icyt==8 ,i(jsub) nolog
Random-effects interval regression Number of obs =
97
Group variable (i): jsub Number of groups =
28
Random effects u_i ~ Gaussian Obs per group: min =
3
avg =
3.5
max =
4
Wald chi2(3) =
9.37
Log likelihood = -98.397601 Prob > chi2 =
0.0248
----------------------------------------------------------------------------
--
| Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+--------------------------------------------------------------
--
ind2 | .2779403 .1801249 1.54 0.123 -.075098
.6309787
ind3 | .4799128 .1793074 2.68 0.007 .1284768
.8313488
ind4 | .0490643 .1817517 0.27 0.787 -.3071625
.4052911
_cons | 1.64652 .1439518 11.44 0.000 1.36438
1.92866
-------------+--------------------------------------------------------------
--
/sigma_u | .5890853 .0847491 6.95 0.000 .4229802
.7551905
/sigma_e | .5086366 .0529546 9.61 0.000 .4048475
.6124256
-------------+--------------------------------------------------------------
--
rho | .5728955 .0834249 .4078475
.7259425
----------------------------------------------------------------------------
--
. ereturn list
scalars:
e(ll) = -98.39760067570394
e(df_m) = 3
e(chi2) = 9.367763093475308
e(n_quad) = 12
e(N_g) = 28
e(g_min) = 3
e(g_avg) = 3.464285714285714
e(g_max) = 4
e(N_unc) = 78
e(N_lc) = 19
e(N_rc) = 0
e(N_int) = 0
e(N) = 97
e(chi2_c) = 48.31552213790093
e(sigma_u) = .5890853482329951
e(sigma_e) = .5086365726862585
e(rho) = .5728955056304219
macros:
e(cmd) : "xtintreg"
e(ivar) : "jsub"
e(predict) : "tobit_p"
e(depvar) : "zmiss z"
e(chi2_ct) : "Wald"
e(properties) : "b V"
e(title) : "Random-effects interval regression"
e(distrib) : "Gaussian"
e(intmethod) : "ghermite"
e(crittype) : "log likelihood"
e(chi2type) : "Wald"
matrices:
e(b) : 1 x 6
e(V) : 6 x 6
functions:
e(sample)
.
. xtintreg zmiss z ind2 ind3 ind4 if icyt==8 ,i(jsub) nolog
Random-effects interval regression Number of obs =
97
Group variable (i): jsub Number of groups =
28
Random effects u_i ~ Gaussian Obs per group: min =
3
avg =
3.5
max =
4
Wald chi2(3) =
9.25
Log likelihood = -98.427307 Prob > chi2 =
0.0261
----------------------------------------------------------------------------
--
| Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+--------------------------------------------------------------
--
ind2 | .2834601 .1838606 1.54 0.123 -.0769
.6438203
ind3 | .4859499 .1829623 2.66 0.008 .1273503
.8445495
ind4 | .0547933 .1856597 0.30 0.768 -.3090932
.4186797
_cons | 1.663773 .1530167 10.87 0.000 1.363866
1.96368
-------------+--------------------------------------------------------------
--
/sigma_u | .6051366 .0952469 6.35 0.000 .4184562
.7918171
/sigma_e | .513304 .0509305 10.08 0.000 .4134821
.6131259
-------------+--------------------------------------------------------------
--
rho | .5815579 .0908475 .4012919
.7459393
----------------------------------------------------------------------------
--
Observation summary: 19 left-censored observations
78 uncensored observations
0 right-censored observations
0 interval observations
. ereturn list
scalars:
e(rc) = 0
e(ll) = -98.42730712890625
e(converged) = 1
e(rank) = 6
e(k) = 6
e(k_eq) = 3
e(k_dv) = 2
e(ic) = 4
e(df_m) = 3
e(chi2) = 9.249642461173721
e(p) = .0261494544480191
e(n_quad) = 12
e(N_g) = 28
e(g_min) = 3
e(g_avg) = 3.464285714285714
e(g_max) = 4
e(N_unc) = 78
e(N_lc) = 19
e(N_rc) = 0
e(N) = 97
e(rho) = .5815578938064966
e(sigma_e) = .5133039972278113
e(sigma_u) = .605136618421632
e(chi2_c) = 40.36503330986928
e(N_int) = 0
macros:
e(cmd) : "xtintreg"
e(ivar) : "jsub"
e(predict) : "xtintreg_p"
e(depvar) : "zmiss z"
e(chi2_ct) : "Wald"
e(properties) : "b V"
e(title) : "Random-effects interval regression"
e(distrib) : "Gaussian"
e(intmethod) : "aghermite"
e(chi2type) : "Wald"
e(opt) : "ml"
e(ml_method) : "d2"
e(user) : "xtintreg_d2"
e(crittype) : "log likelihood"
e(technique) : "nr"
matrices:
e(b) : 1 x 6
e(V) : 6 x 6
e(gradient) : 1 x 6
e(ilog) : 1 x 20
functions:
e(sample)
.
. which storecmd
.\storecmd.ado
*! 1.0.0 NJC 8 December 1999 with code from Jeroen Weesie's -keyb-
program storecmd
*! 1.0.0 NJC 8 December 1999 with code from Jeroen Weesie's -keyb-
version 7.0
gettoken cname 0 : 0
if `"`0'"' == "" {
error 198
}
if `"$BUF_DISP"' != "" {
di in wh `". `0'"'
}
local cmd `"`0'"'
* put command in keyboard buffer
push `cmd'
* execute command
`cmd'
* save to characteristic if it worked
char _dta[`cname'] `"`cmd'"'
end
*
* 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/