The SPost package by Scott Long and Jeremy Freese is a suite of post-estimation commands used to compute additional tests and effects representations for a variety of regression models (see http://www.indiana.edu/~jslsoc/spost.htm). To facilitate and automate the task of processing result from SPost for inclusion in reports and publications, estadd provides tools to integrate SPost results with estout or esttab.
Supported commands are brant, fitstat, listcoef, mlogtest, prchange, prvalue, and asprvalue from SPost for Stata 9 or newer (spost9_ado). SPost for Stata 8 (spostado) is not supported. See the SPost section in estadd's documentation for further details. Below is a range of examples covering various models and applications.
The general procedure to tabulate results from an SPost command in esttab or estout is to
For example, to tabulate a number of fitstat information measures for a linear regression model, type:
. spex regjob2
(Academic Biochemists / S Long)
. regress job fem phd ment fel art cit
Source | SS df MS Number of obs = 408
-------------+------------------------------ F( 6, 401) = 17.78
Model | 81.0584763 6 13.5097461 Prob > F = 0.0000
Residual | 304.737915 401 .759944926 R-squared = 0.2101
-------------+------------------------------ Adj R-squared = 0.1983
Total | 385.796392 407 .947902683 Root MSE = .87175
------------------------------------------------------------------------------
job | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
fem | -.1391939 .0902344 -1.54 0.124 -.3165856 .0381977
phd | .2726826 .0493183 5.53 0.000 .1757278 .3696375
ment | .0011867 .0007012 1.69 0.091 -.0001917 .0025651
fel | .2341384 .0948206 2.47 0.014 .0477308 .4205461
art | .0228011 .0288843 0.79 0.430 -.0339824 .0795846
cit | .0044788 .0019687 2.28 0.023 .0006087 .008349
_cons | 1.067184 .1661357 6.42 0.000 .7405785 1.39379
------------------------------------------------------------------------------
. estadd fitstat, bic
AIC: 2.580 AIC*n: 1052.793
BIC: -1371.725 BIC': -60.162
BIC used by Stata: 1080.872 AIC used by Stata: 1052.793
added scalars:
e(aic0) = 2.5803757
e(aic_n) = 1052.7933
e(bic0) = -1371.7248
e(bic_p) = -60.162312
e(statabic) = 1080.8722
e(stataaic) = 1052.7933
. esttab, cells(none) scalars(aic0 aic_n bic0 bic_p)
-------------------------
(1)
job
-------------------------
N 408
aic0 2.580
aic_n 1052.8
bic0 -1371.7
bic_p -60.16
-------------------------
If you are working with multiple models you can either add results to each model individually after estimation (as above), or you can first estimate and store a set of models and then apply estadd to all of them in one call using the colon syntax. Here is an example of the latter, using eststo to store the models:
. spex regjob2
(Academic Biochemists / S Long)
. eststo: regress job fem phd ment
Source | SS df MS Number of obs = 408
-------------+------------------------------ F( 3, 404) = 23.77
Model | 57.8903644 3 19.2967881 Prob > F = 0.0000
Residual | 327.906027 404 .811648583 R-squared = 0.1501
-------------+------------------------------ Adj R-squared = 0.1437
Total | 385.796392 407 .947902683 Root MSE = .90092
------------------------------------------------------------------------------
job | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
fem | -.1769641 .0915984 -1.93 0.054 -.3570331 .003105
phd | .3307536 .0495896 6.67 0.000 .2332678 .4282395
ment | .0015841 .0007207 2.20 0.029 .0001673 .0030009
_cons | 1.171768 .1635376 7.17 0.000 .8502769 1.493259
------------------------------------------------------------------------------
(est1 stored)
. eststo: regress job fem phd ment fel art cit
Source | SS df MS Number of obs = 408
-------------+------------------------------ F( 6, 401) = 17.78
Model | 81.0584763 6 13.5097461 Prob > F = 0.0000
Residual | 304.737915 401 .759944926 R-squared = 0.2101
-------------+------------------------------ Adj R-squared = 0.1983
Total | 385.796392 407 .947902683 Root MSE = .87175
------------------------------------------------------------------------------
job | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
fem | -.1391939 .0902344 -1.54 0.124 -.3165856 .0381977
phd | .2726826 .0493183 5.53 0.000 .1757278 .3696375
ment | .0011867 .0007012 1.69 0.091 -.0001917 .0025651
fel | .2341384 .0948206 2.47 0.014 .0477308 .4205461
art | .0228011 .0288843 0.79 0.430 -.0339824 .0795846
cit | .0044788 .0019687 2.28 0.023 .0006087 .008349
_cons | 1.067184 .1661357 6.42 0.000 .7405785 1.39379
------------------------------------------------------------------------------
(est2 stored)
. estadd fitstat, bic: *
. esttab, cells(none) scalars(aic0 aic_n bic0 bic_p)
--------------------------------------
(1) (2)
job job
--------------------------------------
N 408 408
aic0 2.639 2.580
aic_n 1076.7 1052.8
bic0 -1359.9 -1371.7
bic_p -48.30 -60.16
--------------------------------------
. eststo clear
A key difference between the two approaches is that with the first method output from estadd fitstat is displayed, whereas execution with the second syntax is silent.
The default for estadd prchange is to return a matrix called e(dc) containing the 0 to 1 change effects for binary variables and the standard deviation change effects for continuous variables in the first row, followed by additional rows containing separate results for the different effect types computed by prchange. To tabulate the contents of the first row simply refer to dc in esttab or estout. Example:
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. quietly logit lfp k5 k618 age wc hc lwg inc, nolog
. estadd prchange
logit: Changes in Probabilities for lfp
min->max 0->1 -+1/2 -+sd/2 MargEfct
k5 -0.6361 -0.3499 -0.3428 -0.1849 -0.3569
k618 -0.1278 -0.0156 -0.0158 -0.0208 -0.0158
age -0.4372 -0.0030 -0.0153 -0.1232 -0.0153
wc 0.1881 0.1881 0.1945 0.0884 0.1969
hc 0.0272 0.0272 0.0273 0.0133 0.0273
lwg 0.6624 0.1499 0.1465 0.0865 0.1475
inc -0.6415 -0.0068 -0.0084 -0.0975 -0.0084
NotInLF inLF
Pr(y|x) 0.4222 0.5778
k5 k618 age wc hc lwg inc
x= .237716 1.35325 42.5378 .281541 .391766 1.09711 20.129
sd_x= .523959 1.31987 8.07257 .450049 .488469 .587556 11.6348
added scalars:
e(predval) = .57779419
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 6 x 7 (main, min->max, 0->1, -+1/2, -+sd/2, Marg
> Efct)
e(pattern) : 1 x 7
e(X) : 4 x 7 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
. esttab, aux(dc) nopar wide
-----------------------------------------
(1)
lfp
-----------------------------------------
lfp
k5 -1.463*** -0.185
k618 -0.0646 -0.0208
age -0.0629*** -0.123
wc 0.807*** 0.188
hc 0.112 0.0272
lwg 0.605*** 0.0865
inc -0.0344*** -0.0975
_cons 3.182***
-----------------------------------------
N 753
-----------------------------------------
dc in second column
* p<0.05, ** p<0.01, *** p<0.001
To change the defaults for the contents of the first row of e(dc) use the c() option (for continuous variables) and the b() option (for binary variables). For example, to tabulate the marginal effects for continuous variables and the 0 to 1 change effects for binary variables (see the helpfile for the list of available effects types), type:
. estadd prchange, c(margefct) replace
logit: Changes in Probabilities for lfp
min->max 0->1 -+1/2 -+sd/2 MargEfct
k5 -0.6361 -0.3499 -0.3428 -0.1849 -0.3569
k618 -0.1278 -0.0156 -0.0158 -0.0208 -0.0158
age -0.4372 -0.0030 -0.0153 -0.1232 -0.0153
wc 0.1881 0.1881 0.1945 0.0884 0.1969
hc 0.0272 0.0272 0.0273 0.0133 0.0273
lwg 0.6624 0.1499 0.1465 0.0865 0.1475
inc -0.6415 -0.0068 -0.0084 -0.0975 -0.0084
NotInLF inLF
Pr(y|x) 0.4222 0.5778
k5 k618 age wc hc lwg inc
x= .237716 1.35325 42.5378 .281541 .391766 1.09711 20.129
sd_x= .523959 1.31987 8.07257 .450049 .488469 .587556 11.6348
added scalars:
e(predval) = .57779419
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 6 x 7 (main, min->max, 0->1, -+1/2, -+sd/2, Marg
> Efct)
e(pattern) : 1 x 7
e(X) : 4 x 7 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
margefct for continuous variables
. esttab, aux(dc) nopar wide
-----------------------------------------
(1)
lfp
-----------------------------------------
lfp
k5 -1.463*** -0.357
k618 -0.0646 -0.0158
age -0.0629*** -0.0153
wc 0.807*** 0.188
hc 0.112 0.0272
lwg 0.605*** 0.148
inc -0.0344*** -0.00840
_cons 3.182***
-----------------------------------------
N 753
-----------------------------------------
dc in second column
* p<0.05, ** p<0.01, *** p<0.001
Alternatively, if you want to tabulate the different effect types computed by prchange separately, address the rows in e(dc) using dc[#] where # is the row number or dc[name] where name is the row name. Example:
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. quietly logit lfp k5 k618 age wc hc lwg inc, nolog
. estadd prchange
logit: Changes in Probabilities for lfp
min->max 0->1 -+1/2 -+sd/2 MargEfct
k5 -0.6361 -0.3499 -0.3428 -0.1849 -0.3569
k618 -0.1278 -0.0156 -0.0158 -0.0208 -0.0158
age -0.4372 -0.0030 -0.0153 -0.1232 -0.0153
wc 0.1881 0.1881 0.1945 0.0884 0.1969
hc 0.0272 0.0272 0.0273 0.0133 0.0273
lwg 0.6624 0.1499 0.1465 0.0865 0.1475
inc -0.6415 -0.0068 -0.0084 -0.0975 -0.0084
NotInLF inLF
Pr(y|x) 0.4222 0.5778
k5 k618 age wc hc lwg inc
x= .237716 1.35325 42.5378 .281541 .391766 1.09711 20.129
sd_x= .523959 1.31987 8.07257 .450049 .488469 .587556 11.6348
added scalars:
e(predval) = .57779419
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 6 x 7 (main, min->max, 0->1, -+1/2, -+sd/2, Marg
> Efct)
e(pattern) : 1 x 7
e(X) : 4 x 7 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
. esttab, cells("dc[2] dc[3] dc[4] dc[5] dc[6]")
-----------------------------------------------------------------------------
(1)
lfp
dc[2] dc[3] dc[4] dc[5] dc[6]
-----------------------------------------------------------------------------
k5 -.6360998 -.3498737 -.3427888 -.1848925 -.3568748
k618 -.1277862 -.0156047 -.0157506 -.0207876 -.0157519
age -.4372017 -.002954 -.015336 -.1231976 -.0153371
wc .1880592 .1880592 .1944887 .0884042 .1969329
hc .0271984 .0271984 .0272506 .0133135 .0272572
lwg .6624324 .1499499 .14648 .0864619 .1475137
inc -.6415044 -.0068042 -.008403 -.0974665 -.0084031
-----------------------------------------------------------------------------
N 753
-----------------------------------------------------------------------------
. esttab, cells("dc[min->max] dc[0->1] dc[-+1/2] dc[-+sd/2] dc[MargEfct]")
-----------------------------------------------------------------------------
(1)
lfp
min->max 0->1 -+1/2 -+sd/2 MargEfct
-----------------------------------------------------------------------------
k5 -.6360998 -.3498737 -.3427888 -.1848925 -.3568748
k618 -.1277862 -.0156047 -.0157506 -.0207876 -.0157519
age -.4372017 -.002954 -.015336 -.1231976 -.0153371
wc .1880592 .1880592 .1944887 .0884042 .1969329
hc .0271984 .0271984 .0272506 .0133135 .0272572
lwg .6624324 .1499499 .14648 .0864619 .1475137
inc -.6415044 -.0068042 -.008403 -.0974665 -.0084031
-----------------------------------------------------------------------------
N 753
-----------------------------------------------------------------------------
The procedure to prepare results from prvalue for tabulation is to first collect a series of predictions by repeated calls to estadd prvalue and then apply estadd prvalue post to rearrange results and post them in e(). Use the label() option to label the single predictions. Example:
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. logit lfp k5 k618 age wc hc lwg inc, nolog
Logistic regression Number of obs = 753
LR chi2(7) = 124.48
Prob > chi2 = 0.0000
Log likelihood = -452.63296 Pseudo R2 = 0.1209
------------------------------------------------------------------------------
lfp | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
k5 | -1.462913 .1970006 -7.43 0.000 -1.849027 -1.076799
k618 | -.0645707 .0680008 -0.95 0.342 -.1978499 .0687085
age | -.0628706 .0127831 -4.92 0.000 -.0879249 -.0378162
wc | .8072738 .2299799 3.51 0.000 .3565215 1.258026
hc | .1117336 .2060397 0.54 0.588 -.2920969 .515564
lwg | .6046931 .1508176 4.01 0.000 .3090961 .9002901
inc | -.0344464 .0082084 -4.20 0.000 -.0505346 -.0183583
_cons | 3.18214 .6443751 4.94 0.000 1.919188 4.445092
------------------------------------------------------------------------------
. estadd prvalue, x(age=35 k5=2 wc=0 hc=0 inc=15) label(family type 1)
logit: Predictions for lfp
Confidence intervals by delta method
95% Conf. Interval
Pr(y=inLF|x): 0.1318 [ 0.0556, 0.2081]
Pr(y=NotInLF|x): 0.8682 [ 0.7919, 0.9444]
k5 k618 age wc hc lwg
x= 2 1.3532537 35 0 0 1.0971148
inc
x= 15
added matrices:
e(_estadd_prvalue) : 1 x 12
e(_estadd_prvalue_x) : 1 x 7
. estadd prvalue, x(age=50 k5=0 k618=0 wc=1 hc=1) label(family type 2)
logit: Predictions for lfp
Confidence intervals by delta method
95% Conf. Interval
Pr(y=inLF|x): 0.7166 [ 0.6333, 0.7999]
Pr(y=NotInLF|x): 0.2834 [ 0.2001, 0.3667]
k5 k618 age wc hc lwg
x= 0 0 50 1 1 1.0971148
inc
x= 20.128965
updated matrices:
e(_estadd_prvalue) : 2 x 12
e(_estadd_prvalue_x) : 2 x 7
. estadd prvalue, label(average family)
logit: Predictions for lfp
Confidence intervals by delta method
95% Conf. Interval
Pr(y=inLF|x): 0.5778 [ 0.5392, 0.6164]
Pr(y=NotInLF|x): 0.4222 [ 0.3836, 0.4608]
k5 k618 age wc hc lwg
x= .2377158 1.3532537 42.537849 .2815405 .39176627 1.0971148
inc
x= 20.128965
updated matrices:
e(_estadd_prvalue) : 3 x 12
e(_estadd_prvalue_x) : 3 x 7
. estadd prvalue post
scalars:
e(N) = 753
macros:
e(depvar) : "lfp"
e(cmd) : "estadd_prvalue"
e(model) : "logit"
e(properties) : "b"
matrices:
e(b) : 1 x 6 (predictions)
e(se) : 1 x 6 (standard errors)
e(LB) : 1 x 6 (lower CI bounds)
e(UB) : 1 x 6 (upper CI bounds)
e(Category) : 1 x 6 (outcome values)
e(X) : 7 x 3 (k5, k618, age, wc, hc, lwg, inc)
. esttab, ci wide nostar ///
> keep(inLF:) eqlabels(none) varwidth(15)
---------------------------------------------------
(1)
lfp
---------------------------------------------------
family type 1 0.132 [0.0556,0.208]
family type 2 0.717 [0.633,0.800]
average family 0.578 [0.539,0.616]
---------------------------------------------------
N 753
---------------------------------------------------
95% confidence intervals in brackets
The procedure for asprvalue is analogous (however, note that asprvalue does not provide standard errors or confidence intervals).
If you want to tabulate differences in predictions, first apply prvalue (or asprvalue) with the save option and then estadd prvalue (or estadd asprvalue) with the diff option. Example:
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. quietly logit lfp k5 k618 age wc hc lwg inc, nolog
. quietly prvalue, x(k5=0 wc=0) save
. estadd prvalue, x(k5=0 wc=1) label(k5 = 0) brief diff
logit: Change in Predictions for lfp
Current Saved Change 95% CI for Change
Pr(y=inLF|x): 0.7758 0.6069 0.1689 [ 0.0830, 0.2549]
Pr(y=NotInLF|x): 0.2242 0.3931 -0.1689 [-0.2549, -0.0830]
added matrices:
e(_estadd_prvalue) : 1 x 12
e(_estadd_prvalue_x) : 1 x 7
. quietly prvalue, x(k5=1 wc=0) save
. estadd prvalue, x(k5=1 wc=1) label(k5 = 1) brief diff
logit: Change in Predictions for lfp
Current Saved Change 95% CI for Change
Pr(y=inLF|x): 0.4449 0.2633 0.1815 [ 0.0763, 0.2868]
Pr(y=NotInLF|x): 0.5551 0.7367 -0.1815 [-0.2868, -0.0763]
updated matrices:
e(_estadd_prvalue) : 2 x 12
e(_estadd_prvalue_x) : 2 x 7
. quietly prvalue, x(k5=2 wc=0) save
. estadd prvalue, x(k5=2 wc=1) label(k5 = 2) brief diff
logit: Change in Predictions for lfp
Current Saved Change 95% CI for Change
Pr(y=inLF|x): 0.1565 0.0764 0.0801 [ 0.0156, 0.1445]
Pr(y=NotInLF|x): 0.8435 0.9236 -0.0801 [-0.1445, -0.0156]
updated matrices:
e(_estadd_prvalue) : 3 x 12
e(_estadd_prvalue_x) : 3 x 7
. estadd prvalue post
scalars:
e(N) = 753
macros:
e(depvar) : "lfp"
e(cmd) : "estadd_prvalue"
e(model) : "logit"
e(properties) : "b"
matrices:
e(b) : 1 x 6 (predictions)
e(se) : 1 x 6 (standard errors)
e(LB) : 1 x 6 (lower CI bounds)
e(UB) : 1 x 6 (upper CI bounds)
e(Category) : 1 x 6 (outcome values)
e(X) : 7 x 3 (k5, k618, age, wc, hc, lwg, inc)
. esttab, keep(inLF:) ci wide nostar ///
> mtitle("wc=1 - wc=0")
------------------------------------------------
(1)
wc=1 - wc=0
------------------------------------------------
inLF
k5 = 0 0.169 [0.0830,0.255]
k5 = 1 0.182 [0.0763,0.287]
k5 = 2 0.0801 [0.0156,0.145]
------------------------------------------------
N 753
------------------------------------------------
95% confidence intervals in brackets
The confidence bounds computed by prvalue are saved by estadd prvalue post in e(LB) and e(UB). The following example illustrates how to tabulate these results:
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. quietly logit lfp k5 k618 age wc hc lwg inc, nolog
. estadd prvalue, x(age=35 k5=2 wc=0 hc=0 inc=15) ///
> label(family type 1) bootstrap
logit: Predictions for lfp
Bootstrap confidence intervals using percentile method
(1000 of 1000 replications completed)
95% Conf. Interval
Pr(y=inLF|x): 0.1318 [ 0.0629, 0.2220]
Pr(y=NotInLF|x): 0.8682 [ 0.7780, 0.9371]
k5 k618 age wc hc lwg
x= 2 1.3532537 35 0 0 1.0971148
inc
x= 15
added matrices:
e(_estadd_prvalue) : 1 x 12
e(_estadd_prvalue_x) : 1 x 7
. estadd prvalue, x(age=50 k5=0 k618=0 wc=1 hc=1) ///
> label(family type 2) bootstrap
logit: Predictions for lfp
Bootstrap confidence intervals using percentile method
(1000 of 1000 replications completed)
95% Conf. Interval
Pr(y=inLF|x): 0.7166 [ 0.6305, 0.7994]
Pr(y=NotInLF|x): 0.2834 [ 0.2006, 0.3695]
k5 k618 age wc hc lwg
x= 0 0 50 1 1 1.0971148
inc
x= 20.128965
updated matrices:
e(_estadd_prvalue) : 2 x 12
e(_estadd_prvalue_x) : 2 x 7
. estadd prvalue, label(average family) bootstrap
logit: Predictions for lfp
Bootstrap confidence intervals using percentile method
(1000 of 1000 replications completed)
95% Conf. Interval
Pr(y=inLF|x): 0.5778 [ 0.5389, 0.6205]
Pr(y=NotInLF|x): 0.4222 [ 0.3795, 0.4611]
k5 k618 age wc hc lwg
x= .2377158 1.3532537 42.537849 .2815405 .39176627 1.0971148
inc
x= 20.128965
updated matrices:
e(_estadd_prvalue) : 3 x 12
e(_estadd_prvalue_x) : 3 x 7
. estadd prvalue post
scalars:
e(N) = 753
macros:
e(depvar) : "lfp"
e(cmd) : "estadd_prvalue"
e(model) : "logit"
e(properties) : "b"
matrices:
e(b) : 1 x 6 (predictions)
e(se) : 1 x 6 (standard errors)
e(LB) : 1 x 6 (lower CI bounds)
e(UB) : 1 x 6 (upper CI bounds)
e(Category) : 1 x 6 (outcome values)
e(X) : 7 x 3 (k5, k618, age, wc, hc, lwg, inc)
. esttab, cells("b LB UB") ///
> keep(inLF:) eqlabels(none) varwidth(15)
------------------------------------------------------
(1)
lfp
b LB UB
------------------------------------------------------
family type 1 .1318369 .062898 .2219745
family type 2 .7166017 .6304579 .7994323
average family .5777942 .5389454 .6205298
------------------------------------------------------
N 753
------------------------------------------------------
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. quietly logit lfp k5 k618 age wc hc lwg inc, nolog
. estadd fitstat
Measures of Fit for logit of lfp
Log-Lik Intercept Only: -514.873 Log-Lik Full Model: -452.633
D(745): 905.266 LR(7): 124.480
Prob > LR: 0.000
McFadden's R2: 0.121 McFadden's Adj R2: 0.105
ML (Cox-Snell) R2: 0.152 Cragg-Uhler(Nagelkerke) R2: 0.204
McKelvey & Zavoina's R2: 0.217 Efron's R2: 0.155
Variance of y*: 4.203 Variance of error: 3.290
Count R2: 0.693 Adj Count R2: 0.289
AIC: 1.223 AIC*n: 921.266
BIC: -4029.663 BIC': -78.112
BIC used by Stata: 958.258 AIC used by Stata: 921.266
added scalars:
e(dev) = 905.26592
e(dev_df) = 745
e(lrx2) = 124.48049
e(lrx2_df) = 7
e(lrx2_p) = 8.923e-24
e(r2_mf) = .12088461
e(r2_mfadj) = .1053468
e(r2_ml) = .15237143
e(r2_cu) = .20445312
e(r2_mz) = .2171939
e(r2_ef) = .15493519
e(v_ystar) = 4.2026603
e(v_error) = 3.2898681
e(r2_ct) = .69322709
e(r2_ctadj) = .28923077
e(aic0) = 1.2234607
e(aic_n) = 921.26592
e(bic0) = -4029.6627
e(bic_p) = -78.112037
e(statabic) = 958.25844
e(stataaic) = 921.26592
e(n_rhs) = 7
e(n_parm) = 8
. eststo logit
. quietly probit lfp k5 k618 age wc hc lwg inc, nolog
. estadd fitstat
Measures of Fit for probit of lfp
Log-Lik Intercept Only: -514.873 Log-Lik Full Model: -452.695
D(745): 905.390 LR(7): 124.356
Prob > LR: 0.000
McFadden's R2: 0.121 McFadden's Adj R2: 0.105
ML (Cox-Snell) R2: 0.152 Cragg-Uhler(Nagelkerke) R2: 0.204
McKelvey & Zavoina's R2: 0.247 Efron's R2: 0.154
Variance of y*: 1.328 Variance of error: 1.000
Count R2: 0.687 Adj Count R2: 0.274
AIC: 1.224 AIC*n: 921.390
BIC: -4029.539 BIC': -77.988
BIC used by Stata: 958.382 AIC used by Stata: 921.390
added scalars:
e(dev) = 905.38993
e(dev_df) = 745
e(lrx2) = 124.35648
e(lrx2_df) = 7
e(lrx2_p) = 9.471e-24
e(r2_mf) = .12076418
e(r2_mfadj) = .10522638
e(r2_ml) = .15223182
e(r2_cu) = .2042658
e(r2_mz) = .24703499
e(r2_ef) = .15420358
e(v_ystar) = 1.328083
e(v_error) = 1
e(r2_ct) = .68658699
e(r2_ctadj) = .27384615
e(aic0) = 1.2236254
e(aic_n) = 921.38993
e(bic0) = -4029.5387
e(bic_p) = -77.988025
e(statabic) = 958.38245
e(stataaic) = 921.38993
e(n_rhs) = 7
e(n_parm) = 8
. eststo probit
. esttab, scalars(r2_mf r2_mfadj r2_ml r2_cu) wide mtitles
----------------------------------------------------------------------
(1) (2)
logit probit
----------------------------------------------------------------------
lfp
k5 -1.463*** (-7.43) -0.875*** (-7.70)
k618 -0.0646 (-0.95) -0.0386 (-0.95)
age -0.0629*** (-4.92) -0.0378*** (-4.97)
wc 0.807*** (3.51) 0.488*** (3.60)
hc 0.112 (0.54) 0.0572 (0.46)
lwg 0.605*** (4.01) 0.366*** (4.17)
inc -0.0344*** (-4.20) -0.0205*** (-4.30)
_cons 3.182*** (4.94) 1.918*** (5.04)
----------------------------------------------------------------------
N 753 753
r2_mf 0.121 0.121
r2_mfadj 0.105 0.105
r2_ml 0.152 0.152
r2_cu 0.204 0.204
----------------------------------------------------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. quietly logit lfp k5 k618 age wc hc lwg inc, nolog
. estadd listcoef, std
logit (N=753): Unstandardized and Standardized Estimates
Observed SD: .49562951
Latent SD: 2.0500391
Odds of: inLF vs NotInLF
-------------------------------------------------------------------------------
lfp | b z P>|z| bStdX bStdY bStdXY SDofX
-------------+-----------------------------------------------------------------
k5 | -1.46291 -7.426 0.000 -0.7665 -0.7136 -0.3739 0.5240
k618 | -0.06457 -0.950 0.342 -0.0852 -0.0315 -0.0416 1.3199
age | -0.06287 -4.918 0.000 -0.5075 -0.0307 -0.2476 8.0726
wc | 0.80727 3.510 0.000 0.3633 0.3938 0.1772 0.4500
hc | 0.11173 0.542 0.588 0.0546 0.0545 0.0266 0.4885
lwg | 0.60469 4.009 0.000 0.3553 0.2950 0.1733 0.5876
inc | -0.03445 -4.196 0.000 -0.4008 -0.0168 -0.1955 11.6348
-------------------------------------------------------------------------------
added matrices:
e(b_xs) : 1 x 7 (bStdX)
e(b_ys) : 1 x 7 (bStdY)
e(b_std) : 1 x 7 (bStdXY)
e(b_sdx) : 1 x 7 (SDofX)
. eststo logit
. quietly probit lfp k5 k618 age wc hc lwg inc, nolog
. estadd listcoef
probit (N=753): Unstandardized and Standardized Estimates
Observed SD: .49562951
Latent SD: 1.1524248
-------------------------------------------------------------------------------
lfp | b z P>|z| bStdX bStdY bStdXY SDofX
-------------+-----------------------------------------------------------------
k5 | -0.87471 -7.703 0.000 -0.4583 -0.7590 -0.3977 0.5240
k618 | -0.03859 -0.953 0.340 -0.0509 -0.0335 -0.0442 1.3199
age | -0.03782 -4.971 0.000 -0.3053 -0.0328 -0.2649 8.0726
wc | 0.48831 3.604 0.000 0.2198 0.4237 0.1907 0.4500
hc | 0.05717 0.461 0.645 0.0279 0.0496 0.0242 0.4885
lwg | 0.36563 4.165 0.000 0.2148 0.3173 0.1864 0.5876
inc | -0.02053 -4.297 0.000 -0.2388 -0.0178 -0.2072 11.6348
-------------------------------------------------------------------------------
added matrices:
e(b_xs) : 1 x 7 (bStdX)
e(b_ys) : 1 x 7 (bStdY)
e(b_std) : 1 x 7 (bStdXY)
e(b_sdx) : 1 x 7 (SDofX)
. eststo probit
. esttab, aux(b_std) nopar wide mtitles
----------------------------------------------------------------------
(1) (2)
logit probit
----------------------------------------------------------------------
lfp
k5 -1.463*** -0.374 -0.875*** -0.398
k618 -0.0646 -0.0416 -0.0386 -0.0442
age -0.0629*** -0.248 -0.0378*** -0.265
wc 0.807*** 0.177 0.488*** 0.191
hc 0.112 0.0266 0.0572 0.0242
lwg 0.605*** 0.173 0.366*** 0.186
inc -0.0344*** -0.195 -0.0205*** -0.207
_cons 3.182*** 1.918***
----------------------------------------------------------------------
N 753 753
----------------------------------------------------------------------
b_std in second column
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. quietly logit lfp k5 k618 age wc hc lwg inc, nolog
. estadd listcoef, quietly std
added matrices:
e(b_xs) : 1 x 7 (bStdX)
e(b_ys) : 1 x 7 (bStdY)
e(b_std) : 1 x 7 (bStdXY)
e(b_sdx) : 1 x 7 (SDofX)
. estadd listcoef, quietly fact nosd
added matrices:
e(b_fact) : 1 x 7 (e^b)
e(b_facts) : 1 x 7 (e^bStdX)
. estadd listcoef, quietly per nosd
added matrices:
e(b_pct) : 1 x 7 (%)
e(b_pcts) : 1 x 7 (%StdX)
. esttab, cell("b_std b_facts b_pcts b_sdx")
----------------------------------------------------------------
(1)
lfp
b_std b_facts b_pcts b_sdx
----------------------------------------------------------------
k5 -.3738985 .4646334 -53.53666 .523959
k618 -.0415725 .9183055 -8.169451 1.319874
age -.2475695 .6019823 -39.80177 8.072574
wc .1772225 1.438086 43.8086 .4500494
hc .0266231 1.056095 5.60953 .4884694
lwg .1733095 1.426596 42.65962 .5875564
inc -.1954974 .6697992 -33.02008 11.6348
----------------------------------------------------------------
N 753
----------------------------------------------------------------
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. quietly logit lfp k5 k618 age wc hc lwg inc, nolog
. estadd prchange
logit: Changes in Probabilities for lfp
min->max 0->1 -+1/2 -+sd/2 MargEfct
k5 -0.6361 -0.3499 -0.3428 -0.1849 -0.3569
k618 -0.1278 -0.0156 -0.0158 -0.0208 -0.0158
age -0.4372 -0.0030 -0.0153 -0.1232 -0.0153
wc 0.1881 0.1881 0.1945 0.0884 0.1969
hc 0.0272 0.0272 0.0273 0.0133 0.0273
lwg 0.6624 0.1499 0.1465 0.0865 0.1475
inc -0.6415 -0.0068 -0.0084 -0.0975 -0.0084
NotInLF inLF
Pr(y|x) 0.4222 0.5778
k5 k618 age wc hc lwg inc
x= .237716 1.35325 42.5378 .281541 .391766 1.09711 20.129
sd_x= .523959 1.31987 8.07257 .450049 .488469 .587556 11.6348
added scalars:
e(predval) = .57779419
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 6 x 7 (main, min->max, 0->1, -+1/2, -+sd/2, Marg
> Efct)
e(pattern) : 1 x 7
e(X) : 4 x 7 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
. eststo logit
. quietly probit lfp k5 k618 age wc hc lwg inc, nolog
. estadd prchange
probit: Changes in Probabilities for lfp
min->max 0->1 -+1/2 -+sd/2 MargEfct
k5 -0.6441 -0.3380 -0.3320 -0.1778 -0.3422
k618 -0.1221 -0.0150 -0.0151 -0.0199 -0.0151
age -0.4274 -0.0031 -0.0148 -0.1190 -0.0148
wc 0.1844 0.1844 0.1892 0.0858 0.1911
hc 0.0223 0.0223 0.0224 0.0109 0.0224
lwg 0.6649 0.1450 0.1423 0.0839 0.1431
inc -0.6425 -0.0068 -0.0080 -0.0932 -0.0080
NotInLF inLF
Pr(y|x) 0.4218 0.5782
k5 k618 age wc hc lwg inc
x= .237716 1.35325 42.5378 .281541 .391766 1.09711 20.129
sd_x= .523959 1.31987 8.07257 .450049 .488469 .587556 11.6348
added scalars:
e(predval) = .57816368
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 6 x 7 (main, min->max, 0->1, -+1/2, -+sd/2, Marg
> Efct)
e(pattern) : 1 x 7
e(X) : 4 x 7 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
. eststo probit
. esttab, aux(dc) nopar wide mtitles
----------------------------------------------------------------------
(1) (2)
logit probit
----------------------------------------------------------------------
lfp
k5 -1.463*** -0.185 -0.875*** -0.178
k618 -0.0646 -0.0208 -0.0386 -0.0199
age -0.0629*** -0.123 -0.0378*** -0.119
wc 0.807*** 0.188 0.488*** 0.184
hc 0.112 0.0272 0.0572 0.0223
lwg 0.605*** 0.0865 0.366*** 0.0839
inc -0.0344*** -0.0975 -0.0205*** -0.0932
_cons 3.182*** 1.918***
----------------------------------------------------------------------
N 753 753
----------------------------------------------------------------------
dc in second column
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. quietly logit lfp k5 k618 age wc hc lwg inc, nolog
. estadd prvalue, x(k5=0 wc=0) label(k5 = 0) brief
logit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.6069 [ 0.5567, 0.6570]
Pr(y=NotInLF|x): 0.3931 [ 0.3430, 0.4433]
added matrices:
e(_estadd_prvalue) : 1 x 12
e(_estadd_prvalue_x) : 1 x 7
. estadd prvalue, x(k5=1 wc=0) label(k5 = 1) brief
logit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.2633 [ 0.1932, 0.3335]
Pr(y=NotInLF|x): 0.7367 [ 0.6665, 0.8068]
updated matrices:
e(_estadd_prvalue) : 2 x 12
e(_estadd_prvalue_x) : 2 x 7
. estadd prvalue, x(k5=2 wc=0) label(k5 = 2) brief
logit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.0764 [ 0.0258, 0.1271]
Pr(y=NotInLF|x): 0.9236 [ 0.8729, 0.9742]
updated matrices:
e(_estadd_prvalue) : 3 x 12
e(_estadd_prvalue_x) : 3 x 7
. estadd prvalue, x(k5=3 wc=0) label(k5 = 3) brief
logit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.0188 [-0.0014, 0.0390]
Pr(y=NotInLF|x): 0.9812 [ 0.9610, 1.0014]
updated matrices:
e(_estadd_prvalue) : 4 x 12
e(_estadd_prvalue_x) : 4 x 7
. estadd prvalue post NoCollege
scalars:
e(N) = 753
macros:
e(depvar) : "lfp"
e(cmd) : "estadd_prvalue"
e(model) : "logit"
e(properties) : "b"
matrices:
e(b) : 1 x 8 (predictions)
e(se) : 1 x 8 (standard errors)
e(LB) : 1 x 8 (lower CI bounds)
e(UB) : 1 x 8 (upper CI bounds)
e(Category) : 1 x 8 (outcome values)
e(X) : 7 x 4 (k5, k618, age, wc, hc, lwg, inc)
results stored as NoCollege
. estadd prvalue, x(k5=0 wc=1) label(k5 = 0) brief replace
logit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.7758 [ 0.7077, 0.8439]
Pr(y=NotInLF|x): 0.2242 [ 0.1561, 0.2923]
added matrices:
e(_estadd_prvalue) : 1 x 12
e(_estadd_prvalue_x) : 1 x 7
. estadd prvalue, x(k5=1 wc=1) label(k5 = 1) brief
logit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.4449 [ 0.3331, 0.5567]
Pr(y=NotInLF|x): 0.5551 [ 0.4433, 0.6669]
updated matrices:
e(_estadd_prvalue) : 2 x 12
e(_estadd_prvalue_x) : 2 x 7
. estadd prvalue, x(k5=2 wc=1) label(k5 = 2) brief
logit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.1565 [ 0.0582, 0.2548]
Pr(y=NotInLF|x): 0.8435 [ 0.7452, 0.9418]
updated matrices:
e(_estadd_prvalue) : 3 x 12
e(_estadd_prvalue_x) : 3 x 7
. estadd prvalue, x(k5=3 wc=1) label(k5 = 3) brief
logit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.0412 [-0.0021, 0.0845]
Pr(y=NotInLF|x): 0.9588 [ 0.9155, 1.0021]
updated matrices:
e(_estadd_prvalue) : 4 x 12
e(_estadd_prvalue_x) : 4 x 7
. estadd prvalue post College
scalars:
e(N) = 753
macros:
e(depvar) : "lfp"
e(cmd) : "estadd_prvalue"
e(model) : "logit"
e(properties) : "b"
matrices:
e(b) : 1 x 8 (predictions)
e(se) : 1 x 8 (standard errors)
e(LB) : 1 x 8 (lower CI bounds)
e(UB) : 1 x 8 (upper CI bounds)
e(Category) : 1 x 8 (outcome values)
e(X) : 7 x 4 (k5, k618, age, wc, hc, lwg, inc)
results stored as College
. quietly prvalue, x(k5=0 wc=0) save
. estadd prvalue, x(k5=0 wc=1) label(k5 = 0) brief diff replace
logit: Change in Predictions for lfp
Current Saved Change 95% CI for Change
Pr(y=inLF|x): 0.7758 0.6069 0.1689 [ 0.0830, 0.2549]
Pr(y=NotInLF|x): 0.2242 0.3931 -0.1689 [-0.2549, -0.0830]
added matrices:
e(_estadd_prvalue) : 1 x 12
e(_estadd_prvalue_x) : 1 x 7
. quietly prvalue, x(k5=1 wc=0) save
. estadd prvalue, x(k5=1 wc=1) label(k5 = 1) brief diff
logit: Change in Predictions for lfp
Current Saved Change 95% CI for Change
Pr(y=inLF|x): 0.4449 0.2633 0.1815 [ 0.0763, 0.2868]
Pr(y=NotInLF|x): 0.5551 0.7367 -0.1815 [-0.2868, -0.0763]
updated matrices:
e(_estadd_prvalue) : 2 x 12
e(_estadd_prvalue_x) : 2 x 7
. quietly prvalue, x(k5=2 wc=0) save
. estadd prvalue, x(k5=2 wc=1) label(k5 = 2) brief diff
logit: Change in Predictions for lfp
Current Saved Change 95% CI for Change
Pr(y=inLF|x): 0.1565 0.0764 0.0801 [ 0.0156, 0.1445]
Pr(y=NotInLF|x): 0.8435 0.9236 -0.0801 [-0.1445, -0.0156]
updated matrices:
e(_estadd_prvalue) : 3 x 12
e(_estadd_prvalue_x) : 3 x 7
. quietly prvalue, x(k5=3 wc=0) save
. estadd prvalue, x(k5=3 wc=1) label(k5 = 3) brief diff
logit: Change in Predictions for lfp
Current Saved Change 95% CI for Change
Pr(y=inLF|x): 0.0412 0.0188 0.0224 [-0.0037, 0.0485]
Pr(y=NotInLF|x): 0.9588 0.9812 -0.0224 [-0.0485, 0.0037]
updated matrices:
e(_estadd_prvalue) : 4 x 12
e(_estadd_prvalue_x) : 4 x 7
. estadd prvalue post Difference
scalars:
e(N) = 753
macros:
e(depvar) : "lfp"
e(cmd) : "estadd_prvalue"
e(model) : "logit"
e(properties) : "b"
matrices:
e(b) : 1 x 8 (predictions)
e(se) : 1 x 8 (standard errors)
e(LB) : 1 x 8 (lower CI bounds)
e(UB) : 1 x 8 (upper CI bounds)
e(Category) : 1 x 8 (outcome values)
e(X) : 7 x 4 (k5, k618, age, wc, hc, lwg, inc)
results stored as Difference
. esttab, se nostar nonumber noobs mtitles ///
> keep(inLF:) eqlabels(none)
---------------------------------------------------
NoCollege College Difference
---------------------------------------------------
k5 = 0 0.607 0.776 0.169
(0.0256) (0.0348) (0.0439)
k5 = 1 0.263 0.445 0.182
(0.0358) (0.0570) (0.0537)
k5 = 2 0.0764 0.157 0.0801
(0.0259) (0.0502) (0.0329)
k5 = 3 0.0188 0.0412 0.0224
(0.0103) (0.0221) (0.0133)
---------------------------------------------------
Standard errors in parentheses
. eststo clear
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. quietly probit lfp k5 k618 age wc hc lwg inc, nolog
. estadd prvalue, x(k5=0 wc=0) label(k5 = 0) brief
probit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.6055 [ 0.5563, 0.6547]
Pr(y=NotInLF|x): 0.3945 [ 0.3453, 0.4437]
added matrices:
e(_estadd_prvalue) : 1 x 12
e(_estadd_prvalue_x) : 1 x 7
. estadd prvalue, x(k5=1 wc=0) label(k5 = 1) brief
probit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.2719 [ 0.2017, 0.3421]
Pr(y=NotInLF|x): 0.7281 [ 0.6579, 0.7983]
updated matrices:
e(_estadd_prvalue) : 2 x 12
e(_estadd_prvalue_x) : 2 x 7
. estadd prvalue, x(k5=2 wc=0) label(k5 = 2) brief
probit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.0692 [ 0.0140, 0.1244]
Pr(y=NotInLF|x): 0.9308 [ 0.8756, 0.9860]
updated matrices:
e(_estadd_prvalue) : 3 x 12
e(_estadd_prvalue_x) : 3 x 7
. estadd prvalue, x(k5=3 wc=0) label(k5 = 3) brief
probit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.0092 [-0.0065, 0.0249]
Pr(y=NotInLF|x): 0.9908 [ 0.9751, 1.0065]
updated matrices:
e(_estadd_prvalue) : 4 x 12
e(_estadd_prvalue_x) : 4 x 7
. estadd prvalue post NoCollege
scalars:
e(N) = 753
macros:
e(depvar) : "lfp"
e(cmd) : "estadd_prvalue"
e(model) : "probit"
e(properties) : "b"
matrices:
e(b) : 1 x 8 (predictions)
e(se) : 1 x 8 (standard errors)
e(LB) : 1 x 8 (lower CI bounds)
e(UB) : 1 x 8 (upper CI bounds)
e(Category) : 1 x 8 (outcome values)
e(X) : 7 x 4 (k5, k618, age, wc, hc, lwg, inc)
results stored as NoCollege
. estadd prvalue, x(k5=0 wc=1) label(k5 = 0) brief replace
probit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.7752 [ 0.7070, 0.8434]
Pr(y=NotInLF|x): 0.2248 [ 0.1566, 0.2930]
added matrices:
e(_estadd_prvalue) : 1 x 12
e(_estadd_prvalue_x) : 1 x 7
. estadd prvalue, x(k5=1 wc=1) label(k5 = 1) brief
probit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.4527 [ 0.3477, 0.5578]
Pr(y=NotInLF|x): 0.5473 [ 0.4422, 0.6523]
updated matrices:
e(_estadd_prvalue) : 2 x 12
e(_estadd_prvalue_x) : 2 x 7
. estadd prvalue, x(k5=2 wc=1) label(k5 = 2) brief
probit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.1602 [ 0.0547, 0.2658]
Pr(y=NotInLF|x): 0.8398 [ 0.7342, 0.9453]
updated matrices:
e(_estadd_prvalue) : 3 x 12
e(_estadd_prvalue_x) : 3 x 7
. estadd prvalue, x(k5=3 wc=1) label(k5 = 3) brief
probit: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.0309 [-0.0135, 0.0752]
Pr(y=NotInLF|x): 0.9691 [ 0.9248, 1.0135]
updated matrices:
e(_estadd_prvalue) : 4 x 12
e(_estadd_prvalue_x) : 4 x 7
. estadd prvalue post College
scalars:
e(N) = 753
macros:
e(depvar) : "lfp"
e(cmd) : "estadd_prvalue"
e(model) : "probit"
e(properties) : "b"
matrices:
e(b) : 1 x 8 (predictions)
e(se) : 1 x 8 (standard errors)
e(LB) : 1 x 8 (lower CI bounds)
e(UB) : 1 x 8 (upper CI bounds)
e(Category) : 1 x 8 (outcome values)
e(X) : 7 x 4 (k5, k618, age, wc, hc, lwg, inc)
results stored as College
. quietly prvalue, x(k5=0 wc=0) save
. estadd prvalue, x(k5=0 wc=1) label(k5 = 0) brief diff replace
probit: Change in Predictions for lfp
Current Saved Change 95% CI for Change
Pr(y=inLF|x): 0.7752 0.6055 0.1696 [ 0.0839, 0.2554]
Pr(y=NotInLF|x): 0.2248 0.3945 -0.1696 [-0.2554, -0.0839]
added matrices:
e(_estadd_prvalue) : 1 x 12
e(_estadd_prvalue_x) : 1 x 7
. quietly prvalue, x(k5=1 wc=0) save
. estadd prvalue, x(k5=1 wc=1) label(k5 = 1) brief diff
probit: Change in Predictions for lfp
Current Saved Change 95% CI for Change
Pr(y=inLF|x): 0.4527 0.2719 0.1808 [ 0.0803, 0.2814]
Pr(y=NotInLF|x): 0.5473 0.7281 -0.1808 [-0.2814, -0.0803]
updated matrices:
e(_estadd_prvalue) : 2 x 12
e(_estadd_prvalue_x) : 2 x 7
. quietly prvalue, x(k5=2 wc=0) save
. estadd prvalue, x(k5=2 wc=1) label(k5 = 2) brief diff
probit: Change in Predictions for lfp
Current Saved Change 95% CI for Change
Pr(y=inLF|x): 0.1602 0.0692 0.0910 [ 0.0217, 0.1604]
Pr(y=NotInLF|x): 0.8398 0.9308 -0.0910 [-0.1604, -0.0217]
updated matrices:
e(_estadd_prvalue) : 3 x 12
e(_estadd_prvalue_x) : 3 x 7
. quietly prvalue, x(k5=3 wc=0) save
. estadd prvalue, x(k5=3 wc=1) label(k5 = 3) brief diff
probit: Change in Predictions for lfp
Current Saved Change 95% CI for Change
Pr(y=inLF|x): 0.0309 0.0092 0.0216 [-0.0090, 0.0523]
Pr(y=NotInLF|x): 0.9691 0.9908 -0.0216 [-0.0523, 0.0090]
updated matrices:
e(_estadd_prvalue) : 4 x 12
e(_estadd_prvalue_x) : 4 x 7
. estadd prvalue post Difference
scalars:
e(N) = 753
macros:
e(depvar) : "lfp"
e(cmd) : "estadd_prvalue"
e(model) : "probit"
e(properties) : "b"
matrices:
e(b) : 1 x 8 (predictions)
e(se) : 1 x 8 (standard errors)
e(LB) : 1 x 8 (lower CI bounds)
e(UB) : 1 x 8 (upper CI bounds)
e(Category) : 1 x 8 (outcome values)
e(X) : 7 x 4 (k5, k618, age, wc, hc, lwg, inc)
results stored as Difference
. esttab, se nostar nonumber noobs mtitles ///
> keep(inLF:) eqlabels(none)
---------------------------------------------------
NoCollege College Difference
---------------------------------------------------
k5 = 0 0.606 0.775 0.170
(0.0251) (0.0348) (0.0437)
k5 = 1 0.272 0.453 0.181
(0.0358) (0.0536) (0.0513)
k5 = 2 0.0692 0.160 0.0910
(0.0282) (0.0539) (0.0354)
k5 = 3 0.00922 0.0309 0.0216
(0.00800) (0.0226) (0.0157)
---------------------------------------------------
Standard errors in parentheses
. eststo clear
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. quietly cloglog lfp k5 k618 age wc hc lwg inc, nolog
. estadd fitstat
Measures of Fit for cloglog of lfp
Log-Lik Intercept Only: -514.873 Log-Lik Full Model: -448.471
D(745): 896.943 LR(7): 132.804
Prob > LR: 0.000
McFadden's R2: 0.129 McFadden's Adj R2: 0.113
ML (Cox-Snell) R2: 0.162 Cragg-Uhler(Nagelkerke) R2: 0.217
Efron's R2: 0.160
Count R2: 0.687 Adj Count R2: 0.274
AIC: 1.212 AIC*n: 912.943
BIC: -4037.986 BIC': -86.435
BIC used by Stata: 949.935 AIC used by Stata: 912.943
added scalars:
e(dev) = 896.9429
e(dev_df) = 745
e(lrx2) = 132.80351
e(lrx2_df) = 7
e(lrx2_p) = 1.631e-25
e(r2_mf) = .1289672
e(r2_mfadj) = .11342939
e(r2_ml) = .16168879
e(r2_cu) = .21695524
e(r2_ef) = .15960051
e(r2_ct) = .68658699
e(r2_ctadj) = .27384615
e(aic0) = 1.2124076
e(aic_n) = 912.9429
e(bic0) = -4037.9857
e(bic_p) = -86.435051
e(statabic) = 949.93542
e(stataaic) = 912.9429
e(n_rhs) = 7
e(n_parm) = 8
. estadd listcoef
cloglog (N=753): Unstandardized and Standardized Estimates
Observed SD: .49562951
-------------------------------------------------------------
lfp | b z P>|z| bStdX SDofX
-------------+-----------------------------------------------
k5 | -1.00288 -7.101 0.000 -0.5255 0.5240
k618 | -0.05225 -1.197 0.231 -0.0690 1.3199
age | -0.04036 -5.047 0.000 -0.3258 8.0726
wc | 0.41893 2.877 0.004 0.1885 0.4500
hc | 0.05546 0.408 0.683 0.0271 0.4885
lwg | 0.58236 4.781 0.000 0.3422 0.5876
inc | -0.02493 -4.157 0.000 -0.2900 11.6348
-------------------------------------------------------------
added matrices:
e(b_xs) : 1 x 7 (bStdX)
e(b_sdx) : 1 x 7 (SDofX)
. esttab, cell("b b_xs b_sdx") scalars(r2_mf r2_mfadj r2_ml r2_cu)
---------------------------------------------------
(1)
lfp
b b_xs b_sdx
---------------------------------------------------
lfp
k5 -1.002878 -.525467 .523959
k618 -.0522477 -.0689604 1.319874
age -.0403616 -.3258222 8.072574
wc .4189326 .1885404 .4500494
hc .0554553 .0270882 .4884694
lwg .5823638 .3421716 .5875564
inc -.0249275 -.2900264 11.6348
_cons 1.554071
---------------------------------------------------
N 753
r2_mf 0.129
r2_mfadj 0.113
r2_ml 0.162
r2_cu 0.217
---------------------------------------------------
. spex binlfp2
(Data from 1976 PSID-T Mroz)
. quietly cloglog lfp k5 k618 age wc hc lwg inc, nolog
. estadd prvalue, x(age=35 k5=2 wc=0 hc=0 inc=15) ///
> label(family type 1) brief
cloglog: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.1716 [ 0.0931, 0.2500]
Pr(y=NotInLF|x): 0.8284 [ 0.7500, 0.9069]
added matrices:
e(_estadd_prvalue) : 1 x 12
e(_estadd_prvalue_x) : 1 x 7
. estadd prvalue, x(age=50 k5=0 k618=0 wc=1 hc=1) ///
> label(family type 2) brief
cloglog: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.6862 [ 0.5920, 0.7803]
Pr(y=NotInLF|x): 0.3138 [ 0.2197, 0.4080]
updated matrices:
e(_estadd_prvalue) : 2 x 12
e(_estadd_prvalue_x) : 2 x 7
. estadd prvalue, label(average family) brief
cloglog: Predictions for lfp
95% Conf. Interval
Pr(y=inLF|x): 0.5608 [ 0.5225, 0.5991]
Pr(y=NotInLF|x): 0.4392 [ 0.4009, 0.4775]
updated matrices:
e(_estadd_prvalue) : 3 x 12
e(_estadd_prvalue_x) : 3 x 7
. estadd prvalue post
scalars:
e(N) = 753
macros:
e(depvar) : "lfp"
e(cmd) : "estadd_prvalue"
e(model) : "cloglog"
e(properties) : "b"
matrices:
e(b) : 1 x 6 (predictions)
e(se) : 1 x 6 (standard errors)
e(LB) : 1 x 6 (lower CI bounds)
e(UB) : 1 x 6 (upper CI bounds)
e(Category) : 1 x 6 (outcome values)
e(X) : 7 x 3 (k5, k618, age, wc, hc, lwg, inc)
. esttab, ci wide nostar ///
> keep(inLF:) eqlabels(none) varwidth(15)
---------------------------------------------------
(1)
lfp
---------------------------------------------------
family type 1 0.172 [0.0931,0.250]
family type 2 0.686 [0.592,0.780]
average family 0.561 [0.522,0.599]
---------------------------------------------------
N 753
---------------------------------------------------
95% confidence intervals in brackets
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly ologit warm yr89 male white age ed prst
. estadd brant
Brant Test of Parallel Regression Assumption
Variable | chi2 p>chi2 df
-------------+--------------------------
All | 49.18 0.000 12
-------------+--------------------------
yr89 | 13.01 0.001 2
male | 22.24 0.000 2
white | 1.27 0.531 2
age | 7.38 0.025 2
ed | 4.31 0.116 2
prst | 4.33 0.115 2
----------------------------------------
A significant test statistic provides evidence that the parallel
regression assumption has been violated.
added scalars:
e(brant_chi2) = 49.181219
e(brant_df) = 12
e(brant_p) = 1.944e-06
added matrix:
e(brant) : 2 x 6 (chi2, p>chi2)
. esttab, cell("b t brant[chi2] brant[p>chi2]") ///
> scalars(brant_chi2 brant_df brant_p) ///
> eqlabels(none)
----------------------------------------------------------------
(1)
warm
b t chi2 p>chi2
----------------------------------------------------------------
yr89 .5239025 6.557071 13.01311 .0014936
male -.7332997 -9.343457 22.2379 .0000148
white -.3911595 -3.304247 1.267856 .530504
age -.0216655 -8.777619 7.383264 .0249313
ed .0671728 4.204878 4.310353 .1158828
prst .0060727 1.844178 4.331991 .1146358
cut1 -2.465362 -10.31909
cut2 -.630904 -2.70408
cut3 1.261854 5.392123
----------------------------------------------------------------
N 2293
brant_chi2 49.18
brant_df 12
brant_p 0.00000194
----------------------------------------------------------------
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly ologit warm yr89 male white age ed prst, nolog
. estadd fitstat
Measures of Fit for ologit of warm
Log-Lik Intercept Only: -2995.770 Log-Lik Full Model: -2844.912
D(2284): 5689.825 LR(6): 301.716
Prob > LR: 0.000
McFadden's R2: 0.050 McFadden's Adj R2: 0.047
ML (Cox-Snell) R2: 0.123 Cragg-Uhler(Nagelkerke) R2: 0.133
McKelvey & Zavoina's R2: 0.127
Variance of y*: 3.768 Variance of error: 3.290
Count R2: 0.432 Adj Count R2: 0.093
AIC: 2.489 AIC*n: 5707.825
BIC: -11982.891 BIC': -255.291
BIC used by Stata: 5759.463 AIC used by Stata: 5707.825
added scalars:
e(dev) = 5689.8246
e(dev_df) = 2284
e(lrx2) = 301.71628
e(lrx2_df) = 6
e(lrx2_p) = 3.508e-62
e(r2_mf) = .05035704
e(r2_mfadj) = .04735281
e(r2_ml) = .12329214
e(r2_cu) = .13304665
e(r2_mz) = .12682954
e(v_ystar) = 3.7677272
e(v_error) = 3.2898681
e(r2_ct) = .4317488
e(r2_ctadj) = .09324983
e(aic0) = 2.4892388
e(aic_n) = 5707.8246
e(bic0) = -11982.891
e(bic_p) = -255.29058
e(statabic) = 5759.4631
e(stataaic) = 5707.8246
e(n_rhs) = 6
e(n_parm) = 9
. eststo ologit
. quietly oprobit warm yr89 male white age ed prst, nolog
. estadd fitstat
Measures of Fit for oprobit of warm
Log-Lik Intercept Only: -2995.770 Log-Lik Full Model: -2848.611
D(2284): 5697.222 LR(6): 294.319
Prob > LR: 0.000
McFadden's R2: 0.049 McFadden's Adj R2: 0.046
ML (Cox-Snell) R2: 0.120 Cragg-Uhler(Nagelkerke) R2: 0.130
McKelvey & Zavoina's R2: 0.136
Variance of y*: 1.158 Variance of error: 1.000
Count R2: 0.429 Adj Count R2: 0.089
AIC: 2.492 AIC*n: 5715.222
BIC: -11975.494 BIC': -247.893
BIC used by Stata: 5766.861 AIC used by Stata: 5715.222
added scalars:
e(dev) = 5697.222
e(dev_df) = 2284
e(lrx2) = 294.31886
e(lrx2_df) = 6
e(lrx2_p) = 1.349e-60
e(r2_mf) = .0491224
e(r2_mfadj) = .04611816
e(r2_ml) = .12045924
e(r2_cu) = .12998962
e(r2_mz) = .1363472
e(v_ystar) = 1.1578727
e(v_error) = 1
e(r2_ct) = .42913214
e(r2_ctadj) = .08907446
e(aic0) = 2.4924649
e(aic_n) = 5715.222
e(bic0) = -11975.494
e(bic_p) = -247.89316
e(statabic) = 5766.8605
e(stataaic) = 5715.222
e(n_rhs) = 6
e(n_parm) = 9
. eststo oprobit
. esttab, scalars(r2_mf r2_mfadj r2_ml r2_cu) wide eqlabels(none) mtitles
----------------------------------------------------------------------
(1) (2)
ologit oprobit
----------------------------------------------------------------------
yr89 0.524*** (6.56) 0.319*** (6.80)
male -0.733*** (-9.34) -0.417*** (-9.16)
white -0.391*** (-3.30) -0.227** (-3.26)
age -0.0217*** (-8.78) -0.0122*** (-8.47)
ed 0.0672*** (4.20) 0.0387*** (4.15)
prst 0.00607 (1.84) 0.00328 (1.71)
cut1 -2.465*** (-10.32) -1.429*** (-10.29)
cut2 -0.631** (-2.70) -0.361** (-2.63)
cut3 1.262*** (5.39) 0.768*** (5.60)
----------------------------------------------------------------------
N 2293 2293
r2_mf 0.0504 0.0491
r2_mfadj 0.0474 0.0461
r2_ml 0.123 0.120
r2_cu 0.133 0.130
----------------------------------------------------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly ologit warm yr89 male white age ed prst, nolog
. estadd listcoef, std
ologit (N=2293): Unstandardized and Standardized Estimates
Observed SD: .9282156
Latent SD: 1.9410634
-------------------------------------------------------------------------------
warm | b z P>|z| bStdX bStdY bStdXY SDofX
-------------+-----------------------------------------------------------------
yr89 | 0.52390 6.557 0.000 0.2566 0.2699 0.1322 0.4897
male | -0.73330 -9.343 0.000 -0.3658 -0.3778 -0.1885 0.4989
white | -0.39116 -3.304 0.001 -0.1287 -0.2015 -0.0663 0.3290
age | -0.02167 -8.778 0.000 -0.3635 -0.0112 -0.1873 16.7790
ed | 0.06717 4.205 0.000 0.2123 0.0346 0.1094 3.1608
prst | 0.00607 1.844 0.065 0.0880 0.0031 0.0453 14.4923
-------------------------------------------------------------------------------
added matrices:
e(b_xs) : 1 x 6 (bStdX)
e(b_ys) : 1 x 6 (bStdY)
e(b_std) : 1 x 6 (bStdXY)
e(b_sdx) : 1 x 6 (SDofX)
. eststo ologit
. quietly oprobit warm yr89 male white age ed prst, nolog
. estadd listcoef
oprobit (N=2293): Unstandardized and Standardized Estimates
Observed SD: .9282156
Latent SD: 1.0760449
-------------------------------------------------------------------------------
warm | b z P>|z| bStdX bStdY bStdXY SDofX
-------------+-----------------------------------------------------------------
yr89 | 0.31881 6.805 0.000 0.1561 0.2963 0.1451 0.4897
male | -0.41703 -9.156 0.000 -0.2080 -0.3876 -0.1933 0.4989
white | -0.22650 -3.260 0.001 -0.0745 -0.2105 -0.0693 0.3290
age | -0.01222 -8.471 0.000 -0.2051 -0.0114 -0.1906 16.7790
ed | 0.03872 4.153 0.000 0.1224 0.0360 0.1137 3.1608
prst | 0.00328 1.705 0.088 0.0476 0.0031 0.0442 14.4923
-------------------------------------------------------------------------------
added matrices:
e(b_xs) : 1 x 6 (bStdX)
e(b_ys) : 1 x 6 (bStdY)
e(b_std) : 1 x 6 (bStdXY)
e(b_sdx) : 1 x 6 (SDofX)
. eststo oprobit
. esttab, aux(b_std) nopar wide eqlabels(none) mtitles
----------------------------------------------------------------------
(1) (2)
ologit oprobit
----------------------------------------------------------------------
yr89 0.524*** 0.132 0.319*** 0.145
male -0.733*** -0.188 -0.417*** -0.193
white -0.391*** -0.0663 -0.227** -0.0693
age -0.0217*** -0.187 -0.0122*** -0.191
ed 0.0672*** 0.109 0.0387*** 0.114
prst 0.00607 0.0453 0.00328 0.0442
cut1 -2.465*** -1.429***
cut2 -0.631** -0.361**
cut3 1.262*** 0.768***
----------------------------------------------------------------------
N 2293 2293
----------------------------------------------------------------------
b_std in second column
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex ordwarm2
(77 & 89 General Social Survey)
. eststo ologit: quietly ologit warm yr89 male white age ed prst, nolog
. eststo oprobit: quietly oprobit warm yr89 male white age ed prst, nolog
. estadd prchange male age prst: *
. esttab, main(dc) nostar not mtitles
--------------------------------------
(1) (2)
ologit oprobit
--------------------------------------
Avg|Chg|
male 0.0896 0.0819
age 0.0447 0.0404
prst 0.0108 0.00938
--------------------------------------
1SD
male 0.0746 0.0810
age 0.0360 0.0390
prst -0.00870 -0.00905
--------------------------------------
2D
male 0.105 0.0827
age 0.0533 0.0417
prst -0.0130 -0.00972
--------------------------------------
3A
male -0.0814 -0.0622
age -0.0401 -0.0301
prst 0.00977 0.00702
--------------------------------------
4SA
male -0.0979 -0.101
age -0.0492 -0.0506
prst 0.0119 0.0117
--------------------------------------
N 2293 2293
--------------------------------------
dc coefficients
. eststo clear
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly ologit warm yr89 male white age ed prst, nolog
. estadd prchange male age prst, outcome(2)
ologit: Changes in Probabilities for warm
Outcome: 2 (2D)
Min->Max 0->1 -+1/2 -+sd/2 MargEfct
male .10462105 .10462105 .10556346 .05362016 .10812605
age .1862759 .00289795 .00319454 .05328724 .00319461
prst -.06301633 -.00080028 -.00089544 -.01297233 -.00089543
1SD 2D 3A 4SA
Pr(y|x) .11125716 .32816544 .39936733 .16121005
yr89 male white age ed prst
x= .398604 .464893 .876581 44.9355 12.2181 39.5853
sd_x= .489718 .498875 .328989 16.779 3.16083 14.4923
added scalars:
e(predval) = .32816544
e(outcome) = 2
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 6 x 3 (main, Min->Max, 0->1, -+1/2, -+sd/2, Marg
> Efct)
e(pattern) : 1 x 3
e(X) : 4 x 6 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
. eststo ologit
. quietly oprobit warm yr89 male white age ed prst, nolog
. estadd prchange male age prst, outcome(2)
oprobit: Changes in Probabilities for warm
Outcome: 2 (2D)
Min->Max 0->1 -+1/2 -+sd/2 MargEfct
male .08271328 .08271328 .08378038 .0423308 .08521085
age .1476199 .00280122 .00249711 .04172876 .00249716
prst -.04764727 -.00058544 -.00067082 -.00971937 -.00067081
1SD 2D 3A 4SA
Pr(y|x) .11177309 .32895118 .39563131 .1636444
yr89 male white age ed prst
x= .398604 .464893 .876581 44.9355 12.2181 39.5853
sd_x= .489718 .498875 .328989 16.779 3.16083 14.4923
added scalars:
e(predval) = .32895118
e(outcome) = 2
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 6 x 3 (main, Min->Max, 0->1, -+1/2, -+sd/2, Marg
> Efct)
e(pattern) : 1 x 3
e(X) : 4 x 6 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
. eststo oprobit
. esttab, main(dc) nostar not stats(predval outcome) mtitles
--------------------------------------
(1) (2)
ologit oprobit
--------------------------------------
male 0.105 0.0827
age 0.0533 0.0417
prst -0.0130 -0.00972
--------------------------------------
predval 0.328 0.329
outcome 2 2
--------------------------------------
dc coefficients
. eststo clear
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly ologit warm yr89 male white age ed prst, nolog
. estadd prchange male age prst, split
ologit: Changes in Probabilities for warm
male
Avg|Chg| 1SD 2D 3A 4SA
0->1 .08961766 .07461427 .10462105 -.08137083 -.09786447
age
Avg|Chg| 1SD 2D 3A 4SA
Min->Max .18319855 .18012119 .1862759 -.17905769 -.18733941
-+1/2 .00266841 .00214228 .00319454 -.00240716 -.00292964
-+sd/2 .0446563 .03602537 .05328724 -.0401054 -.0492072
MargEfct .00266844 .00214226 .00319461 -.00240723 -.00292964
prst
Avg|Chg| 1SD 2D 3A 4SA
Min->Max .05186236 -.04070839 -.06301633 .04440692 .05931778
-+1/2 .00074795 -.00060046 -.00089544 .00067475 .00082116
-+sd/2 .01083777 -.00870322 -.01297233 .00977433 .0119012
MargEfct .00074795 -.00060046 -.00089543 .00067473 .00082116
1SD 2D 3A 4SA
Pr(y|x) .11125716 .32816544 .39936733 .16121005
yr89 male white age ed prst
x= .398604 .464893 .876581 44.9355 12.2181 39.5853
sd_x= .489718 .498875 .328989 16.779 3.16083 14.4923
added scalars:
e(predval) = .11125716
e(outcome) = 1
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 6 x 3 (main, Min->Max, 0->1, -+1/2, -+sd/2, Marg
> Efct)
e(pattern) : 1 x 3
e(X) : 4 x 6 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
results for outcome 1 stored as ologit_1
results for outcome 2 stored as ologit_2
results for outcome 3 stored as ologit_3
results for outcome 4 stored as ologit_4
. esttab, main(dc) nostar not stats(predval outcome) ///
> mtitles nonumbers
----------------------------------------------------------------
1SD 2D 3A 4SA
----------------------------------------------------------------
male 0.0746 0.105 -0.0814 -0.0979
age 0.0360 0.0533 -0.0401 -0.0492
prst -0.00870 -0.0130 0.00977 0.0119
----------------------------------------------------------------
predval 0.111 0.328 0.399 0.161
outcome 1 2 3 4
----------------------------------------------------------------
dc coefficients
. eststo clear
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly oprobit warm yr89 male white age ed prst, nolog
. estadd prchange male age prst, split
oprobit: Changes in Probabilities for warm
male
Avg|Chg| 1SD 2D 3A 4SA
0->1 .08185343 .08099359 .08271328 -.06222743 -.10147941
age
Avg|Chg| 1SD 2D 3A 4SA
Min->Max .16764789 .18767587 .1476199 -.14207253 -.19322326
-+1/2 .00241081 .00232452 .00249711 -.00180408 -.00301754
-+sd/2 .04038233 .03903589 .04172876 -.03013682 -.05062783
MargEfct .00241084 .00232452 .00249716 -.00180413 -.00301755
prst
Avg|Chg| 1SD 2D 3A 4SA
Min->Max .04505957 -.04247186 -.04764727 .03203639 .05808274
-+1/2 .00064762 -.00062443 -.00067082 .00048462 .00081059
-+sd/2 .00938461 -.00904983 -.00971937 .00702184 .01174738
MargEfct .00064762 -.00062443 -.00067081 .00048464 .0008106
1SD 2D 3A 4SA
Pr(y|x) .11177309 .32895118 .39563131 .1636444
yr89 male white age ed prst
x= .398604 .464893 .876581 44.9355 12.2181 39.5853
sd_x= .489718 .498875 .328989 16.779 3.16083 14.4923
added scalars:
e(predval) = .11177309
e(outcome) = 1
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 6 x 3 (main, Min->Max, 0->1, -+1/2, -+sd/2, Marg
> Efct)
e(pattern) : 1 x 3
e(X) : 4 x 6 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
results for outcome 1 stored as oprobit_1
results for outcome 2 stored as oprobit_2
results for outcome 3 stored as oprobit_3
results for outcome 4 stored as oprobit_4
. esttab, main(dc) nostar not stats(predval outcome) ///
> mtitles nonumbers
----------------------------------------------------------------
1SD 2D 3A 4SA
----------------------------------------------------------------
male 0.0810 0.0827 -0.0622 -0.101
age 0.0390 0.0417 -0.0301 -0.0506
prst -0.00905 -0.00972 0.00702 0.0117
----------------------------------------------------------------
predval 0.112 0.329 0.396 0.164
outcome 1 2 3 4
----------------------------------------------------------------
dc coefficients
. eststo clear
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly ologit warm yr89 male white age ed prst, nolog
. estadd prvalue, x(yr89=0 male=1 prst=20 age=64 ed=16) ///
> brief label(type1)
ologit: Predictions for warm
95% Conf. Interval
Pr(y=1SD|x): 0.2317 [ 0.1776, 0.2857]
Pr(y=2D|x): 0.4221 [ 0.3942, 0.4500]
Pr(y=3A|x): 0.2723 [ 0.2249, 0.3198]
Pr(y=4SA|x): 0.0739 [ 0.0523, 0.0954]
added matrices:
e(_estadd_prvalue) : 1 x 24
e(_estadd_prvalue_x) : 1 x 6
. estadd prvalue, x(yr89=1 male=0 prst=80 age=30 ed=24) ///
> brief label(type2)
ologit: Predictions for warm
95% Conf. Interval
Pr(y=1SD|x): 0.0164 [ 0.0106, 0.0222]
Pr(y=2D|x): 0.0781 [ 0.0554, 0.1008]
Pr(y=3A|x): 0.3147 [ 0.2636, 0.3658]
Pr(y=4SA|x): 0.5908 [ 0.5143, 0.6673]
updated matrices:
e(_estadd_prvalue) : 2 x 24
e(_estadd_prvalue_x) : 2 x 6
. estadd prvalue, x(yr89=0) brief label(type3)
ologit: Predictions for warm
95% Conf. Interval
Pr(y=1SD|x): 0.1336 [ 0.1176, 0.1496]
Pr(y=2D|x): 0.3577 [ 0.3348, 0.3806]
Pr(y=3A|x): 0.3737 [ 0.3517, 0.3957]
Pr(y=4SA|x): 0.1349 [ 0.1195, 0.1504]
updated matrices:
e(_estadd_prvalue) : 3 x 24
e(_estadd_prvalue_x) : 3 x 6
. estadd prvalue, x(yr89=1) brief label(type4)
ologit: Predictions for warm
95% Conf. Interval
Pr(y=1SD|x): 0.0837 [ 0.0711, 0.0963]
Pr(y=2D|x): 0.2802 [ 0.2571, 0.3032]
Pr(y=3A|x): 0.4277 [ 0.4046, 0.4507]
Pr(y=4SA|x): 0.2085 [ 0.1855, 0.2315]
updated matrices:
e(_estadd_prvalue) : 4 x 24
e(_estadd_prvalue_x) : 4 x 6
. estadd prvalue post
scalars:
e(N) = 2293
macros:
e(depvar) : "warm"
e(cmd) : "estadd_prvalue"
e(model) : "ologit"
e(properties) : "b"
matrices:
e(b) : 1 x 16 (predictions)
e(se) : 1 x 16 (standard errors)
e(LB) : 1 x 16 (lower CI bounds)
e(UB) : 1 x 16 (upper CI bounds)
e(Category) : 1 x 16 (outcome values)
e(X) : 6 x 4 (yr89, male, white, age, ed, prst)
. esttab, nostar unstack ///
> coeflabels(type1 "old working class men 1977" ///
> type2 "young prestigious women 1989" ///
> type3 "average individual 1977" ///
> type4 "average individual 1989") ///
> wrap varwidth(18)
----------------------------------------------------------------------
(1)
warm
1SD 2D 3A 4SA
----------------------------------------------------------------------
old working class 0.232 0.422 0.272 0.0739
men 1977 (8.40) (29.67) (11.25) (6.72)
young prestigious 0.0164 0.0781 0.315 0.591
women 1989 (5.56) (6.74) (12.07) (15.13)
average individual 0.134 0.358 0.374 0.135
1977 (16.37) (30.57) (33.30) (17.09)
average individual 0.0837 0.280 0.428 0.208
1989 (13.00) (23.81) (36.29) (17.77)
----------------------------------------------------------------------
N 2293
----------------------------------------------------------------------
t statistics in parentheses
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly oprobit warm yr89 male white age ed prst, nolog
. estadd prvalue, x(yr89=0 male=1 prst=20 age=64 ed=16) ///
> brief label(type1)
oprobit: Predictions for warm
95% Conf. Interval
Pr(y=1SD|x): 0.2370 [ 0.1821, 0.2918]
Pr(y=2D|x): 0.4006 [ 0.3750, 0.4261]
Pr(y=3A|x): 0.2931 [ 0.2844, 0.3017]
Pr(y=4SA|x): 0.0693 [ 0.0450, 0.0936]
added matrices:
e(_estadd_prvalue) : 1 x 24
e(_estadd_prvalue_x) : 1 x 6
. estadd prvalue, x(yr89=1 male=0 prst=80 age=30 ed=24) ///
> brief label(type2)
oprobit: Predictions for warm
95% Conf. Interval
Pr(y=1SD|x): 0.0088 [ 0.0040, 0.0136]
Pr(y=2D|x): 0.0870 [ 0.0754, 0.0985]
Pr(y=3A|x): 0.3338 [ 0.3083, 0.3593]
Pr(y=4SA|x): 0.5704 [ 0.4977, 0.6432]
updated matrices:
e(_estadd_prvalue) : 2 x 24
e(_estadd_prvalue_x) : 2 x 6
. estadd prvalue, x(yr89=0) brief label(type3)
oprobit: Predictions for warm
95% Conf. Interval
Pr(y=1SD|x): 0.1378 [ 0.1213, 0.1544]
Pr(y=2D|x): 0.3534 [ 0.3262, 0.3805]
Pr(y=3A|x): 0.3746 [ 0.3605, 0.3886]
Pr(y=4SA|x): 0.1342 [ 0.1182, 0.1502]
updated matrices:
e(_estadd_prvalue) : 3 x 24
e(_estadd_prvalue_x) : 3 x 6
. estadd prvalue, x(yr89=1) brief label(type4)
oprobit: Predictions for warm
95% Conf. Interval
Pr(y=1SD|x): 0.0794 [ 0.0660, 0.0929]
Pr(y=2D|x): 0.2872 [ 0.2615, 0.3128]
Pr(y=3A|x): 0.4180 [ 0.3990, 0.4370]
Pr(y=4SA|x): 0.2154 [ 0.1917, 0.2391]
updated matrices:
e(_estadd_prvalue) : 4 x 24
e(_estadd_prvalue_x) : 4 x 6
. estadd prvalue post
scalars:
e(N) = 2293
macros:
e(depvar) : "warm"
e(cmd) : "estadd_prvalue"
e(model) : "oprobit"
e(properties) : "b"
matrices:
e(b) : 1 x 16 (predictions)
e(se) : 1 x 16 (standard errors)
e(LB) : 1 x 16 (lower CI bounds)
e(UB) : 1 x 16 (upper CI bounds)
e(Category) : 1 x 16 (outcome values)
e(X) : 6 x 4 (yr89, male, white, age, ed, prst)
. esttab, nostar unstack ///
> coeflabels(type1 "old working class men 1977" ///
> type2 "young prestigious women 1989" ///
> type3 "average individual 1977" ///
> type4 "average individual 1989") ///
> wrap varwidth(18)
----------------------------------------------------------------------
(1)
warm
1SD 2D 3A 4SA
----------------------------------------------------------------------
old working class 0.237 0.401 0.293 0.0693
men 1977 (8.47) (30.74) (66.40) (5.59)
young prestigious 0.00879 0.0870 0.334 0.570
women 1989 (3.59) (14.72) (25.67) (15.37)
average individual 0.138 0.353 0.375 0.134
1977 (16.33) (25.49) (52.31) (16.44)
average individual 0.0794 0.287 0.418 0.215
1989 (11.57) (21.95) (43.16) (17.83)
----------------------------------------------------------------------
N 2293
----------------------------------------------------------------------
t statistics in parentheses
. spex nomocc2
(1982 General Social Survey)
. quietly mlogit occ white ed exper, nolog
. estadd fitstat
Measures of Fit for mlogit of occ
Log-Lik Intercept Only: -509.844 Log-Lik Full Model: -426.800
D(321): 853.601 LR(12): 166.087
Prob > LR: 0.000
McFadden's R2: 0.163 McFadden's Adj R2: 0.131
ML (Cox-Snell) R2: 0.389 Cragg-Uhler(Nagelkerke) R2: 0.409
Count R2: 0.501 Adj Count R2: 0.253
AIC: 2.628 AIC*n: 885.601
BIC: -1014.646 BIC': -96.246
BIC used by Stata: 946.722 AIC used by Stata: 885.601
added scalars:
e(dev) = 853.60095
e(dev_df) = 321
e(lrx2) = 166.08716
e(lrx2_df) = 12
e(lrx2_p) = 3.010e-29
e(r2_mf) = .16288035
e(r2_mfadj) = .13149821
e(r2_ml) = .38911114
e(r2_cu) = .40895353
e(r2_ct) = .50148368
e(r2_ctadj) = .25333333
e(aic0) = 2.627896
e(aic_n) = 885.60095
e(bic0) = -1014.6457
e(bic_p) = -96.246162
e(statabic) = 946.72228
e(stataaic) = 885.60095
e(n_rhs) = 3
e(n_parm) = 16
. eststo mlogit
. esttab, wide scalars(r2_ct r2_ctadj aic0 aic_n) mtitles
-----------------------------------------
(1)
mlogit
-----------------------------------------
Menial
white -1.774* (-2.35)
ed -0.779*** (-6.79)
exper -0.0357* (-1.98)
_cons 11.52*** (6.23)
-----------------------------------------
BlueCol
white -0.538 (-0.67)
ed -0.878*** (-8.74)
exper -0.0309* (-2.15)
_cons 12.26*** (7.35)
-----------------------------------------
Craft
white -1.302* (-2.01)
ed -0.685*** (-7.67)
exper -0.00797 (-0.63)
_cons 10.43*** (6.87)
-----------------------------------------
WhiteCol
white -0.203 (-0.23)
ed -0.426*** (-4.62)
exper -0.00106 (-0.07)
_cons 5.280** (3.14)
-----------------------------------------
Prof
o.white 0 (.)
o.ed 0 (.)
o.exper 0 (.)
o._cons 0 (.)
-----------------------------------------
N 337
r2_ct 0.501
r2_ctadj 0.253
aic0 2.628
aic_n 885.6
-----------------------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex nomocc2
(1982 General Social Survey)
. quietly mlogit occ white ed exper, nolog
. estadd listcoef, gt adjacent
mlogit (N=337): Factor Change in the Odds of occ
Variable: white (sd=.27642268)
Odds comparing |
Alternative 1 |
to Alternative 2 | b z P>|z| e^b e^bStdX
------------------+---------------------------------------------
BlueCol -Menial | 1.23650 1.707 0.088 3.4436 1.4075
Craft -BlueCol | -0.76416 -1.208 0.227 0.4657 0.8096
WhiteCol-Craft | 1.09904 1.343 0.179 3.0013 1.3550
Prof -WhiteCol | 0.20292 0.233 0.815 1.2250 1.0577
----------------------------------------------------------------
Variable: ed (sd=2.9464271)
Odds comparing |
Alternative 1 |
to Alternative 2 | b z P>|z| e^b e^bStdX
------------------+---------------------------------------------
BlueCol -Menial | -0.09942 -0.972 0.331 0.9054 0.7461
Craft -BlueCol | 0.19324 2.494 0.013 1.2132 1.7671
WhiteCol-Craft | 0.25934 2.773 0.006 1.2961 2.1471
Prof -WhiteCol | 0.42569 4.616 0.000 1.5307 3.5053
----------------------------------------------------------------
Variable: exper (sd=13.959364)
Odds comparing |
Alternative 1 |
to Alternative 2 | b z P>|z| e^b e^bStdX
------------------+---------------------------------------------
BlueCol -Menial | 0.00472 0.271 0.786 1.0047 1.0681
Craft -BlueCol | 0.02296 1.829 0.067 1.0232 1.3779
WhiteCol-Craft | 0.00691 0.495 0.621 1.0069 1.1013
Prof -WhiteCol | 0.00106 0.073 0.941 1.0011 1.0148
----------------------------------------------------------------
added matrices:
e(b_raw) : 1 x 12 (b)
e(b_se) : 1 x 12 (se)
e(b_z) : 1 x 12 (z)
e(b_p) : 1 x 12 (P>|z|)
e(b_fact) : 1 x 12 (e^b)
e(b_facts) : 1 x 12 (e^bStdX)
e(b_sdx) : 1 x 12 (SDofX)
. esttab , cell("b_raw b_fact b_facts b_sdx") varwidth(14)
------------------------------------------------------------------
(1)
occ
b_raw b_fact b_facts b_sdx
------------------------------------------------------------------
BlueCol-Menial
white 1.236504 3.443553 1.407476 .2764227
ed -.0994247 .9053581 .7460612 2.946427
exper .0047212 1.004732 1.068126 13.95936
------------------------------------------------------------------
Craft-BlueCol
white -.7641602 .4657249 .8095869 .2764227
ed .1932401 1.213174 1.76715 2.946427
exper .0229626 1.023228 1.377875 13.95936
------------------------------------------------------------------
WhiteCol-Craft
white 1.099042 3.001288 1.354998 .2764227
ed .2593423 1.296077 2.147132 2.946427
exper .0069121 1.006936 1.101296 13.95936
------------------------------------------------------------------
Prof-WhiteCol
white .2029212 1.224976 1.057695 .2764227
ed .4256943 1.530653 3.505304 2.946427
exper .001055 1.001056 1.014837 13.95936
------------------------------------------------------------------
N 337
------------------------------------------------------------------
. spex nomocc2
(1982 General Social Survey)
. quietly mlogit occ white ed exper, nolog
. estadd mlogtest, wald lr set(white exper)
**** Likelihood-ratio tests for independent variables (N=337)
Ho: All coefficients associated with given variable(s) are 0.
| chi2 df P>chi2
-------------+-------------------------
white | 8.095 4 0.088
ed | 156.937 4 0.000
exper | 8.561 4 0.073
-------------+-------------------------
set_1: | 16.452 8 0.036
white |
exper |
---------------------------------------
**** Wald tests for independent variables (N=337)
Ho: All coefficients associated with given variable(s) are 0.
| chi2 df P>chi2
-------------+-------------------------
white | 8.149 4 0.086
ed | 84.968 4 0.000
exper | 7.995 4 0.092
-------------+-------------------------
set_1: | 15.773 8 0.046
white |
exper |
---------------------------------------
added scalars:
e(wald_set1_chi2) = 15.773146
e(wald_set1_df) = 8
e(wald_set1_p) = .0457446
e(lrtest_set1_chi2) = 16.451934
e(lrtest_set1_df) = 8
e(lrtest_set1_p) = .03634985
added matrices:
e(wald) : 3 x 3 (chi2, df, p)
e(lrtest) : 3 x 3 (chi2, df, p)
. estout, cell("wald[chi2] wald[df] wald[p]") ///
> stat(wald_set1_chi2 wald_set1_df wald_set1_p, ///
> layout("@ @ @") label("white&exper")) ///
> mlabel(none)
---------------------------------------------------
chi2 df p
---------------------------------------------------
white 8.149203 4 .0862631
ed 84.96817 4 1.54e-17
exper 7.994939 4 .0917638
---------------------------------------------------
white&exper 15.77315 8 .0457446
---------------------------------------------------
. estout, cell("lrtest[chi2] lrtest[df] lrtest[p]") ///
> stat(lrtest_set1_chi2 lrtest_set1_df lrtest_set1_p, ///
> layout("@ @ @") label("white&exper")) ///
> mlabel(none)
---------------------------------------------------
chi2 df p
---------------------------------------------------
white 8.095408 4 .0881451
ed 156.9372 4 6.63e-33
exper 8.560953 4 .073061
---------------------------------------------------
white&exper 16.45193 8 .0363498
---------------------------------------------------
. estout, cell(" wald[p](label(P>Wald) fmt(4)) lrtest[p](label(P>LR))") ///
> stat(wald_set1_p lrtest_set1_p, layout("@ @") label("white&exper")) ///
> mlabel(none)
--------------------------------------
P>Wald P>LR
--------------------------------------
white 0.0863 0.0881
ed 0.0000 0.0000
exper 0.0918 0.0731
--------------------------------------
white&exper 0.0457 0.0363
--------------------------------------
. spex nomocc2
(1982 General Social Survey)
. quietly mlogit occ white ed exper, nolog
. estadd prchange
mlogit: Changes in Probabilities for occ
white
Avg|Chg| Menial BlueCol Craft WhiteCol Prof
0->1 .11623582 -.13085523 .04981799 -.15973434 .07971004 .1610615
ed
Avg|Chg| Menial BlueCol Craft WhiteCol
Min->Max .39242268 -.13017954 -.70077323 -.15010394 .02425591
-+1/2 .05855425 -.02559762 -.06831616 -.05247185 .01250795
-+sd/2 .1640657 -.07129153 -.19310513 -.14576758 .03064777
MargEfct .05894859 -.02579097 -.06870635 -.05287415 .01282041
Prof
Min->Max .95680079
-+1/2 .13387768
-+sd/2 .37951647
MargEfct .13455107
exper
Avg|Chg| Menial BlueCol Craft WhiteCol
Min->Max .12193559 -.11536534 -.18947365 .03115708 .09478889
-+1/2 .00233425 -.00226997 -.00356567 .00105992 .0016944
-+sd/2 .03253578 -.03167491 -.04966453 .01479983 .02360725
MargEfct .00233427 -.00226997 -.00356571 .00105992 .00169442
Prof
Min->Max .17889298
-+1/2 .00308132
-+sd/2 .04293236
MargEfct .00308134
Menial BlueCol Craft WhiteCol Prof
Pr(y|x) .09426806 .18419114 .29411051 .16112968 .26630062
white ed exper
x= .916914 13.095 20.5015
sd_x= .276423 2.94643 13.9594
added scalars:
e(predval1) = .09426806
e(predval2) = .18419114
e(predval3) = .29411051
e(predval4) = .16112968
e(predval5) = .26630062
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 6 x 18 (main, Min->Max, 0->1, -+1/2, -+sd/2, Marg
> Efct)
e(pattern) : 1 x 18
e(X) : 4 x 3 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
. eststo mlogit
. quietly mprobit occ white ed exper, nolog
. estadd prchange
mprobit: Changes in Probabilities for occ
white
Avg|Chg| Menial BlueCol Craft WhiteCol Prof
0->1 .11142595 -.13099539 .03923495 -.14756948 .07652746 .16280247
ed
Avg|Chg| Menial BlueCol Craft WhiteCol
Min->Max .39433155 -.13480758 -.63618958 -.21483173 .03136472
-+1/2 .05591886 -.02555373 -.06636748 -.04787594 .00948766
-+sd/2 .1577241 -.0714798 -.18783711 -.13499337 .02528359
Prof
Min->Max .95446413
-+1/2 .13030948
-+sd/2 .36902665
exper
Avg|Chg| Menial BlueCol Craft WhiteCol
Min->Max .123281 -.11218768 -.19601482 .01516485 .07135144
-+1/2 .00229492 -.00215125 -.00358605 .00069633 .00136708
-+sd/2 .03197956 -.02998788 -.04996105 .00971583 .01904152
Prof
Min->Max .2216862
-+1/2 .00367388
-+sd/2 .05119154
Menial BlueCol Craft WhiteCol Prof
Pr(y|x) .09325961 .18944861 .27852002 .15457167 .2842001
white ed exper
x= .916914 13.095 20.5015
sd_x= .276423 2.94643 13.9594
added scalars:
e(predval1) = .09325961
e(predval2) = .18944861
e(predval3) = .27852002
e(predval4) = .15457167
e(predval5) = .2842001
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 5 x 18 (main, Min->Max, 0->1, -+1/2, -+sd/2)
e(pattern) : 1 x 18
e(X) : 4 x 3 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
. eststo mprobit
. esttab mlogit, main(dc) nostar not unstack compress
----------------------------------------------------------------------
(1)
occ
Avg|Chg| Menial BlueCol Craft WhiteCol Prof
----------------------------------------------------------------------
white 0.116 -0.131 0.0498 -0.160 0.0797 0.161
ed 0.164 -0.0713 -0.193 -0.146 0.0306 0.380
exper 0.0325 -0.0317 -0.0497 0.0148 0.0236 0.0429
----------------------------------------------------------------------
N 337
----------------------------------------------------------------------
dc coefficients
. esttab mprobit, main(dc) nostar not unstack compress
----------------------------------------------------------------------
(1)
occ
Avg|Chg| Menial BlueCol Craft WhiteCol Prof
----------------------------------------------------------------------
white 0.111 -0.131 0.0392 -0.148 0.0765 0.163
ed 0.158 -0.0715 -0.188 -0.135 0.0253 0.369
exper 0.0320 -0.0300 -0.0500 0.00972 0.0190 0.0512
----------------------------------------------------------------------
N 337
----------------------------------------------------------------------
dc coefficients
. esttab, main(dc) nostar not mtitles
--------------------------------------
(1) (2)
mlogit mprobit
--------------------------------------
Avg|Chg|
white 0.116 0.111
ed 0.164 0.158
exper 0.0325 0.0320
--------------------------------------
Menial
white -0.131 -0.131
ed -0.0713 -0.0715
exper -0.0317 -0.0300
--------------------------------------
BlueCol
white 0.0498 0.0392
ed -0.193 -0.188
exper -0.0497 -0.0500
--------------------------------------
Craft
white -0.160 -0.148
ed -0.146 -0.135
exper 0.0148 0.00972
--------------------------------------
WhiteCol
white 0.0797 0.0765
ed 0.0306 0.0253
exper 0.0236 0.0190
--------------------------------------
Prof
white 0.161 0.163
ed 0.380 0.369
exper 0.0429 0.0512
--------------------------------------
N 337 337
--------------------------------------
dc coefficients
. eststo clear
. spex nomocc2
(1982 General Social Survey)
. quietly mlogit occ white ed exper, nolog
. estadd prchange, outcome(3)
mlogit: Changes in Probabilities for occ
Outcome: 3 (Craft)
Min->Max 0->1 -+1/2 -+sd/2 MargEfct
white -.15973434 -.15973434 -.17549489 -.05025825 -.18235627
ed -.15010394 .01737602 -.05247185 -.14576758 -.05287415
exper .03115708 .001993 .00105992 .01479983 .00105992
Menial BlueCol Craft WhiteCol Prof
Pr(y|x) .09426806 .18419114 .29411051 .16112968 .26630062
white ed exper
x= .916914 13.095 20.5015
sd_x= .276423 2.94643 13.9594
added scalars:
e(predval) = .29411051
e(outcome) = 3
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 6 x 3 (main, Min->Max, 0->1, -+1/2, -+sd/2, Marg
> Efct)
e(pattern) : 1 x 3
e(X) : 4 x 3 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
. eststo mlogit
. quietly mprobit occ white ed exper, nolog
. estadd prchange, outcome(3)
mprobit: Changes in Probabilities for occ
Outcome: 3 (Craft)
Min->Max 0->1 -+1/2 -+sd/2
white -.14756948 -.14756948 -.15760018 -.04463357
ed -.21483173 .01855293 -.04787594 -.13499337
exper .01516485 .00139609 .00069633 .00971583
Menial BlueCol Craft WhiteCol Prof
Pr(y|x) .09325961 .18944861 .27852002 .15457167 .2842001
white ed exper
x= .916914 13.095 20.5015
sd_x= .276423 2.94643 13.9594
added scalars:
e(predval) = .27852002
e(outcome) = 3
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 5 x 3 (main, Min->Max, 0->1, -+1/2, -+sd/2)
e(pattern) : 1 x 3
e(X) : 4 x 3 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
. eststo mprobit
. esttab, aux(dc) wide nopar stats(predval outcome) keep(Craft:) mtitles
----------------------------------------------------------------------
(1) (2)
mlogit mprobit
----------------------------------------------------------------------
Craft
white -1.302* -0.160 -0.890 -0.148
ed -0.685*** -0.146 -0.472*** -0.135
exper -0.00797 0.0148 -0.00778 0.00972
_cons 10.43*** 7.140***
----------------------------------------------------------------------
predval 0.294 0.279
outcome 3 3
----------------------------------------------------------------------
dc in second column
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex nomocc2
(1982 General Social Survey)
. quietly mlogit occ white ed exper, nolog
. estadd prchange, split
mlogit: Changes in Probabilities for occ
white
Avg|Chg| Menial BlueCol Craft WhiteCol Prof
0->1 .11623582 -.13085523 .04981799 -.15973434 .07971004 .1610615
ed
Avg|Chg| Menial BlueCol Craft WhiteCol
Min->Max .39242268 -.13017954 -.70077323 -.15010394 .02425591
-+1/2 .05855425 -.02559762 -.06831616 -.05247185 .01250795
-+sd/2 .1640657 -.07129153 -.19310513 -.14576758 .03064777
MargEfct .05894859 -.02579097 -.06870635 -.05287415 .01282041
Prof
Min->Max .95680079
-+1/2 .13387768
-+sd/2 .37951647
MargEfct .13455107
exper
Avg|Chg| Menial BlueCol Craft WhiteCol
Min->Max .12193559 -.11536534 -.18947365 .03115708 .09478889
-+1/2 .00233425 -.00226997 -.00356567 .00105992 .0016944
-+sd/2 .03253578 -.03167491 -.04966453 .01479983 .02360725
MargEfct .00233427 -.00226997 -.00356571 .00105992 .00169442
Prof
Min->Max .17889298
-+1/2 .00308132
-+sd/2 .04293236
MargEfct .00308134
Menial BlueCol Craft WhiteCol Prof
Pr(y|x) .09426806 .18419114 .29411051 .16112968 .26630062
white ed exper
x= .916914 13.095 20.5015
sd_x= .276423 2.94643 13.9594
added scalars:
e(predval) = .09426806
e(outcome) = 1
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 6 x 3 (main, Min->Max, 0->1, -+1/2, -+sd/2, Marg
> Efct)
e(pattern) : 1 x 3
e(X) : 4 x 3 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
results for outcome 1 stored as mlogit_1
results for outcome 2 stored as mlogit_2
results for outcome 3 stored as mlogit_3
results for outcome 4 stored as mlogit_4
results for outcome 5 stored as mlogit_5
. esttab, main(dc) nostar not scalars(predval outcome) noobs mtitles
-----------------------------------------------------------------------------
(1) (2) (3) (4) (5)
Menial BlueCol Craft WhiteCol Prof
-----------------------------------------------------------------------------
white -0.131 0.0498 -0.160 0.0797 0.161
ed -0.0713 -0.193 -0.146 0.0306 0.380
exper -0.0317 -0.0497 0.0148 0.0236 0.0429
-----------------------------------------------------------------------------
predval 0.0943 0.184 0.294 0.161 0.266
outcome 1 2 3 4 5
-----------------------------------------------------------------------------
dc coefficients
. eststo clear
. spex nomocc2
(1982 General Social Survey)
. quietly mprobit occ white ed exper, nolog
. estadd prchange, split
mprobit: Changes in Probabilities for occ
white
Avg|Chg| Menial BlueCol Craft WhiteCol Prof
0->1 .11142595 -.13099539 .03923495 -.14756948 .07652746 .16280247
ed
Avg|Chg| Menial BlueCol Craft WhiteCol
Min->Max .39433155 -.13480758 -.63618958 -.21483173 .03136472
-+1/2 .05591886 -.02555373 -.06636748 -.04787594 .00948766
-+sd/2 .1577241 -.0714798 -.18783711 -.13499337 .02528359
Prof
Min->Max .95446413
-+1/2 .13030948
-+sd/2 .36902665
exper
Avg|Chg| Menial BlueCol Craft WhiteCol
Min->Max .123281 -.11218768 -.19601482 .01516485 .07135144
-+1/2 .00229492 -.00215125 -.00358605 .00069633 .00136708
-+sd/2 .03197956 -.02998788 -.04996105 .00971583 .01904152
Prof
Min->Max .2216862
-+1/2 .00367388
-+sd/2 .05119154
Menial BlueCol Craft WhiteCol Prof
Pr(y|x) .09325961 .18944861 .27852002 .15457167 .2842001
white ed exper
x= .916914 13.095 20.5015
sd_x= .276423 2.94643 13.9594
added scalars:
e(predval) = .09325961
e(outcome) = 1
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 5 x 3 (main, Min->Max, 0->1, -+1/2, -+sd/2)
e(pattern) : 1 x 3
e(X) : 4 x 3 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
results for outcome 1 stored as mprobit_1
results for outcome 2 stored as mprobit_2
results for outcome 3 stored as mprobit_3
results for outcome 4 stored as mprobit_4
results for outcome 5 stored as mprobit_5
. esttab, main(dc) nostar not scalars(predval outcome) noobs mtitles
-----------------------------------------------------------------------------
(1) (2) (3) (4) (5)
Menial BlueCol Craft WhiteCol Prof
-----------------------------------------------------------------------------
white -0.131 0.0392 -0.148 0.0765 0.163
ed -0.0715 -0.188 -0.135 0.0253 0.369
exper -0.0300 -0.0500 0.00972 0.0190 0.0512
-----------------------------------------------------------------------------
predval 0.0933 0.189 0.279 0.155 0.284
outcome 1 2 3 4 5
-----------------------------------------------------------------------------
dc coefficients
. eststo clear
. spex nomocc2
(1982 General Social Survey)
. quietly mlogit occ white ed exper, nolog
. levelsof ed, local(edlevels)
3 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
. foreach l of local edlevels {
2. quietly estadd prvalue, x(ed=`l' white=0) label(`l')
3. }
. estadd prvalue post NonWhite
scalars:
e(N) = 337
macros:
e(depvar) : "occ"
e(cmd) : "estadd_prvalue"
e(model) : "mlogit"
e(properties) : "b"
matrices:
e(b) : 1 x 80 (predictions)
e(se) : 1 x 80 (standard errors)
e(LB) : 1 x 80 (lower CI bounds)
e(UB) : 1 x 80 (upper CI bounds)
e(Category) : 1 x 80 (outcome values)
e(X) : 3 x 16 (white, ed, exper)
results stored as NonWhite
. foreach l of local edlevels {
2. quietly estadd prvalue, x(ed=`l' white=1) label(`l')
3. }
. estadd prvalue post White
scalars:
e(N) = 337
macros:
e(depvar) : "occ"
e(cmd) : "estadd_prvalue"
e(model) : "mlogit"
e(properties) : "b"
matrices:
e(b) : 1 x 160 (predictions)
e(se) : 1 x 160 (standard errors)
e(LB) : 1 x 160 (lower CI bounds)
e(UB) : 1 x 160 (upper CI bounds)
e(Category) : 1 x 160 (outcome values)
e(X) : 3 x 32 (white, ed, exper)
results stored as White
. esttab NonWhite White, b(4) se nostar wide ///
> keep(Menial:) mtitles eqlabels(none) noobs
----------------------------------------------------------------
(1) (2)
NonWhite White
----------------------------------------------------------------
3 0.2847 (0.2013) 0.1216 (0.0917)
6 0.2987 (0.1578) 0.1384 (0.0680)
7 0.2988 (0.1440) 0.1417 (0.0585)
8 0.2963 (0.1312) 0.1431 (0.0487)
9 0.2906 (0.1198) 0.1417 (0.0392)
10 0.2814 (0.1100) 0.1366 (0.0308)
11 0.2675 (0.1021) 0.1265 (0.0245)
12 0.2476 (0.0956) 0.1104 (0.0212)
13 0.2199 (0.0895) 0.0883 (0.0195)
14 0.1832 (0.0821) 0.0632 (0.0175)
15 0.1393 (0.0714) 0.0401 (0.0142)
16 0.0944 (0.0566) 0.0228 (0.0102)
17 0.0569 (0.0399) 0.0120 (0.0066)
18 0.0310 (0.0250) 0.0060 (0.0039)
19 0.0158 (0.0143) 0.0029 (0.0022)
20 0.0077 (0.0077) 0.0014 (0.0012)
----------------------------------------------------------------
Standard errors in parentheses
. eststo clear
. spex nomocc2
(1982 General Social Survey)
. quietly mprobit occ white ed exper, nolog
. levelsof ed, local(edlevels)
3 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
. foreach l of local edlevels {
2. quietly estadd prvalue, x(ed=`l' white=0) label(`l')
3. }
. estadd prvalue post NonWhite
scalars:
e(N) = 337
macros:
e(depvar) : "occ"
e(cmd) : "estadd_prvalue"
e(model) : "mprobit"
e(properties) : "b"
matrices:
e(b) : 1 x 80 (predictions)
e(se) : 1 x 80 (standard errors)
e(LB) : 1 x 80 (lower CI bounds)
e(UB) : 1 x 80 (upper CI bounds)
e(Category) : 1 x 80 (outcome values)
e(X) : 3 x 16 (white, ed, exper)
results stored as NonWhite
. foreach l of local edlevels {
2. quietly estadd prvalue, x(ed=`l' white=1) label(`l')
3. }
. estadd prvalue post White
scalars:
e(N) = 337
macros:
e(depvar) : "occ"
e(cmd) : "estadd_prvalue"
e(model) : "mprobit"
e(properties) : "b"
matrices:
e(b) : 1 x 160 (predictions)
e(se) : 1 x 160 (standard errors)
e(LB) : 1 x 160 (lower CI bounds)
e(UB) : 1 x 160 (upper CI bounds)
e(Category) : 1 x 160 (outcome values)
e(X) : 3 x 32 (white, ed, exper)
results stored as White
. esttab NonWhite White, b(4) nostar not ///
> keep(Menial:) mtitles eqlabels(none) noobs
--------------------------------------
(1) (2)
NonWhite White
--------------------------------------
3 0.2446 0.1274
6 0.2617 0.1421
7 0.2654 0.1450
8 0.2676 0.1462
9 0.2679 0.1446
10 0.2652 0.1389
11 0.2577 0.1275
12 0.2429 0.1098
13 0.2185 0.0869
14 0.1842 0.0623
15 0.1429 0.0398
16 0.1006 0.0225
17 0.0636 0.0112
18 0.0357 0.0049
19 0.0178 0.0019
20 0.0078 0.0006
--------------------------------------
. eststo clear
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly slogit warm yr89 male white age ed prst, nolog
. estadd fitstat
Measures of Fit for slogit of warm
Log-Lik Full Model: -2845.595 D(2282): 5691.189
Wald X2(6): 185.448 Prob > X2: 0.000
AIC: 2.492 AIC*n: 5713.189
BIC: -11966.051
BIC used by Stata: 5776.303 AIC used by Stata: 5713.189
added scalars:
e(dev) = 5691.1894
e(dev_df) = 2282
e(lrx2) = 185.44833
e(lrx2_df) = 6
e(lrx2_p) = 2.361e-37
e(aic0) = 2.4915785
e(aic_n) = 5713.1894
e(bic0) = -11966.051
e(statabic) = 5776.3032
e(stataaic) = 5713.1894
e(n_rhs) = 8
e(n_parm) = 11
. estadd listcoef
slogit (N=2293): Factor Change in Odds
Odds of: 4SA vs 1SD
----------------------------------------------------------------------
warm | b z P>|z| e^b e^bStdX SDofX
-------------+--------------------------------------------------------
yr89 | 0.94405 6.179 0.000 2.5704 1.5878 0.4897
male | -1.25606 -8.274 0.000 0.2848 0.5344 0.4989
white | -0.63901 -2.973 0.003 0.5278 0.8104 0.3290
age | -0.03841 -8.541 0.000 0.9623 0.5249 16.7790
ed | 0.10933 3.737 0.000 1.1155 1.4128 3.1608
prst | 0.01148 1.983 0.047 1.0115 1.1810 14.4923
-------------+--------------------------------------------------------
phi1_1 | 1.00000 . .
phi1_2 | 0.74885 13.840 0.000
phi1_3 | 0.31837 6.397 0.000
-------------+--------------------------------------------------------
theta1 | -1.06006 -2.577 0.010
theta2 | 0.13237 0.423 0.672
theta3 | 0.62730 4.399 0.000
----------------------------------------------------------------------
added matrices:
e(b_fact) : 1 x 6 (e^b)
e(b_facts) : 1 x 6 (e^bStdX)
e(b_sdx) : 1 x 6 (SDofX)
. esttab, cell("b b_fact b_facts") scalars(aic0 bic0) ///
> eqlabels(none)
---------------------------------------------------
(1)
warm
b b_fact b_facts
---------------------------------------------------
yr89 .9440522 2.570376 1.587752
male -1.256064 .2847727 .5343958
white -.6390139 .5278126 .8103988
age -.0384116 .9623168 .52492
ed .1093335 1.115534 1.412815
prst .0114819 1.011548 1.181044
phi1_1 1
phi1_2 .7488452
phi1_3 .3183653
theta1 -1.060064
theta2 .1323735
theta3 .6272993
---------------------------------------------------
N 2293
aic0 2.492
bic0 -11966.1
---------------------------------------------------
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly slogit warm yr89 male white age ed prst, nolog
. estadd prchange male age prst
slogit: Changes in Probabilities for warm
male
Avg|Chg| 1SD 2D 3A 4SA
0->1 .09093904 .07668686 .1051912 -.08178753 -.10009058
age
Avg|Chg| 1SD 2D 3A 4SA
Min->Max .19115428 .17745935 .2048492 -.18460661 -.19770195
-+1/2 .00280473 .002321 .00328845 -.00249711 -.00311236
-+sd/2 .04692763 .03891201 .05494326 -.04169974 -.05215551
prst
Avg|Chg| 1SD 2D 3A 4SA
Min->Max .05812308 -.04720601 -.06904018 .0497874 .06645873
-+1/2 .00083838 -.00069379 -.000983 .0007464 .00093034
-+sd/2 .01214788 -.01005406 -.0142417 .01081413 .01348165
1SD 2D 3A 4SA
Pr(y|x) .11714774 .32349858 .39201239 .16734134
yr89 male white age ed prst
x= .398604 .464893 .876581 44.9355 12.2181 39.5853
sd_x= .489718 .498875 .328989 16.779 3.16083 14.4923
added scalars:
e(predval1) = .11714774
e(predval2) = .32349858
e(predval3) = .39201239
e(predval4) = .16734134
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 5 x 15 (main, Min->Max, 0->1, -+1/2, -+sd/2)
e(pattern) : 1 x 15
e(X) : 4 x 6 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
. esttab, main(dc) unstack nostar not
-----------------------------------------------------------------------------
(1)
warm
Avg|Chg| 1SD 2D 3A 4SA
-----------------------------------------------------------------------------
male 0.0909 0.0767 0.105 -0.0818 -0.100
age 0.0469 0.0389 0.0549 -0.0417 -0.0522
prst 0.0121 -0.0101 -0.0142 0.0108 0.0135
-----------------------------------------------------------------------------
N 2293
-----------------------------------------------------------------------------
dc coefficients
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly slogit warm yr89 male white age ed prst, nolog
. estadd prchange age ed prst, outcome(2)
slogit: Changes in Probabilities for warm
Outcome: 2 (2D)
Min->Max 0->1 -+1/2 -+sd/2
age .2048492 .00294244 .00328845 .05494326
ed -.17157121 -.00616026 -.00935909 -.0295499
prst -.06904018 -.00090283 -.000983 -.0142417
1SD 2D 3A 4SA
Pr(y|x) .11714774 .32349858 .39201239 .16734134
yr89 male white age ed prst
x= .398604 .464893 .876581 44.9355 12.2181 39.5853
sd_x= .489718 .498875 .328989 16.779 3.16083 14.4923
added scalars:
e(predval) = .32349858
e(outcome) = 2
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 5 x 3 (main, Min->Max, 0->1, -+1/2, -+sd/2)
e(pattern) : 1 x 3
e(X) : 4 x 6 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
. esttab, cell("dc[Min->Max] dc[-+1/2] dc[-+sd/2]") stats(predval outcome)
---------------------------------------------------
(1)
warm
Min->Max -+1/2 -+sd/2
---------------------------------------------------
age .2048492 .0032884 .0549433
ed -.1715712 -.0093591 -.0295499
prst -.0690402 -.000983 -.0142417
---------------------------------------------------
predval .3234986
outcome 2
---------------------------------------------------
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly slogit warm yr89 male white age ed prst, nolog
. estadd prchange male age prst, split
slogit: Changes in Probabilities for warm
male
Avg|Chg| 1SD 2D 3A 4SA
0->1 .09093904 .07668686 .1051912 -.08178753 -.10009058
age
Avg|Chg| 1SD 2D 3A 4SA
Min->Max .19115428 .17745935 .2048492 -.18460661 -.19770195
-+1/2 .00280473 .002321 .00328845 -.00249711 -.00311236
-+sd/2 .04692763 .03891201 .05494326 -.04169974 -.05215551
prst
Avg|Chg| 1SD 2D 3A 4SA
Min->Max .05812308 -.04720601 -.06904018 .0497874 .06645873
-+1/2 .00083838 -.00069379 -.000983 .0007464 .00093034
-+sd/2 .01214788 -.01005406 -.0142417 .01081413 .01348165
1SD 2D 3A 4SA
Pr(y|x) .11714774 .32349858 .39201239 .16734134
yr89 male white age ed prst
x= .398604 .464893 .876581 44.9355 12.2181 39.5853
sd_x= .489718 .498875 .328989 16.779 3.16083 14.4923
added scalars:
e(predval) = .11714774
e(outcome) = 1
e(delta) = 1
e(centered) = 1
added matrices:
e(dc) : 5 x 3 (main, Min->Max, 0->1, -+1/2, -+sd/2)
e(pattern) : 1 x 3
e(X) : 4 x 6 (X, SD, Min, Max)
first row in e(dc) contains:
01 change for binary variables
sd change for continuous variables
results for outcome 1 stored as slogit_1
results for outcome 2 stored as slogit_2
results for outcome 3 stored as slogit_3
results for outcome 4 stored as slogit_4
. esttab, main(dc) nostar not stats(predval outcome) ///
> mtitles nonumbers
----------------------------------------------------------------
1SD 2D 3A 4SA
----------------------------------------------------------------
male 0.0767 0.105 -0.0818 -0.100
age 0.0389 0.0549 -0.0417 -0.0522
prst -0.0101 -0.0142 0.0108 0.0135
----------------------------------------------------------------
predval 0.117 0.323 0.392 0.167
outcome 1 2 3 4
----------------------------------------------------------------
dc coefficients
. eststo clear
. spex ordwarm2
(77 & 89 General Social Survey)
. quietly slogit warm yr89 male white age ed prst, nolog
. estadd prvalue, x(yr89=0 male=1 prst=20 age=64 ed=16) ///
> brief label(type1)
slogit: Predictions for warm
Pr(y=1SD|x): 0.2341
Pr(y=2D|x): 0.4333
Pr(y=3A|x): 0.2646
Pr(y=4SA|x): 0.0680
added matrices:
e(_estadd_prvalue) : 1 x 24
e(_estadd_prvalue_x) : 1 x 6
. estadd prvalue, x(yr89=1 male=0 prst=80 age=30 ed=24) ///
> brief label(type2)
slogit: Predictions for warm
Pr(y=1SD|x): 0.0112
Pr(y=2D|x): 0.0737
Pr(y=3A|x): 0.3993
Pr(y=4SA|x): 0.5158
updated matrices:
e(_estadd_prvalue) : 2 x 24
e(_estadd_prvalue_x) : 2 x 6
. estadd prvalue, x(yr89=0) brief label(type3)
slogit: Predictions for warm
Pr(y=1SD|x): 0.1412
Pr(y=2D|x): 0.3548
Pr(y=3A|x): 0.3656
Pr(y=4SA|x): 0.1384
updated matrices:
e(_estadd_prvalue) : 3 x 24
e(_estadd_prvalue_x) : 3 x 6
. estadd prvalue, x(yr89=1) brief label(type4)
slogit: Predictions for warm
Pr(y=1SD|x): 0.0860
Pr(y=2D|x): 0.2738
Pr(y=3A|x): 0.4236
Pr(y=4SA|x): 0.2167
updated matrices:
e(_estadd_prvalue) : 4 x 24
e(_estadd_prvalue_x) : 4 x 6
. estadd prvalue post
scalars:
e(N) = 2293
macros:
e(depvar) : "warm"
e(cmd) : "estadd_prvalue"
e(model) : "slogit"
e(properties) : "b"
matrices:
e(b) : 1 x 16 (predictions)
e(se) : 1 x 16 (standard errors)
e(LB) : 1 x 16 (lower CI bounds)
e(UB) : 1 x 16 (upper CI bounds)
e(Category) : 1 x 16 (outcome values)
e(X) : 6 x 4 (yr89, male, white, age, ed, prst)
. esttab, nostar not unstack ///
> coeflabels(type1 "old working class men 1977" ///
> type2 "young prestigious women 1989" ///
> type3 "average individual 1977" ///
> type4 "average individual 1989") ///
> varwidth(28) compress
--------------------------------------------------------------------
(1)
warm
1SD 2D 3A 4SA
--------------------------------------------------------------------
old working class men 1977 0.234 0.433 0.265 0.0680
young prestigious women 1989 0.0112 0.0737 0.399 0.516
average individual 1977 0.141 0.355 0.366 0.138
average individual 1989 0.0860 0.274 0.424 0.217
--------------------------------------------------------------------
N 2293
--------------------------------------------------------------------
. spex travel2
(Greene & Hensher 1997 data on travel mode choice)
. quietly clogit choice train bus time invc, group(id) nolog
. estadd fitstat
Measures of Fit for clogit of choice
Log-Lik Intercept Only: -166.989 Log-Lik Full Model: -80.961
D(148): 161.922 LR(4): 172.056
Prob > LR: 0.000
McFadden's R2: 0.515 McFadden's Adj R2: 0.491
ML (Cox-Snell) R2: 0.678 Cragg-Uhler(Nagelkerke) R2: 0.762
Count R2: 0.875
AIC: 1.118 AIC*n: 169.922
BIC: -581.612 BIC': -151.960
BIC used by Stata: 186.412 AIC used by Stata: 169.922
added scalars:
e(dev) = 161.92227
e(dev_df) = 148
e(lrx2) = 172.05587
e(lrx2_df) = 4
e(lrx2_p) = 3.786e-36
e(r2_mf) = .51517105
e(r2_mfadj) = .49121738
e(r2_ml) = .67759491
e(r2_cu) = .76229428
e(r2_ct) = .875
e(aic0) = 1.1179097
e(aic_n) = 169.92227
e(bic0) = -581.61205
e(bic_p) = -151.96034
e(statabic) = 186.41224
e(stataaic) = 169.92227
e(n_rhs) = 3
e(n_parm) = 4
. estadd listcoef
clogit (N=456): Factor Change in Odds
Odds of: 1 vs 0
--------------------------------------------------
choice | b z P>|z| e^b
-------------+------------------------------------
train | 2.67124 5.895 0.000 14.4579
bus | 1.47233 3.674 0.000 4.3594
time | -0.01915 -7.812 0.000 0.9810
invc | -0.04817 -4.030 0.000 0.9530
--------------------------------------------------
added matrices:
e(b_fact) : 1 x 4 (e^b)
. estadd listcoef, percent
clogit (N=456): Percentage Change in Odds
Odds of: 1 vs 0
--------------------------------------------------
choice | b z P>|z| %
-------------+------------------------------------
train | 2.67124 5.895 0.000 1345.8
bus | 1.47233 3.674 0.000 335.9
time | -0.01915 -7.812 0.000 -1.9
invc | -0.04817 -4.030 0.000 -4.7
--------------------------------------------------
added matrices:
e(b_pct) : 1 x 4 (%)
. esttab, cell("b b_fact b_pct") scalars(r2_mf r2_mfadj r2_ml r2_cu)
---------------------------------------------------
(1)
choice
b b_fact b_pct
---------------------------------------------------
choice
train 2.671238 14.45786 1345.786
bus 1.472335 4.359401 335.9401
time -.0191453 .9810368 -1.896319
invc -.0481658 .9529758 -4.702424
---------------------------------------------------
N 456
r2_mf 0.515
r2_mfadj 0.491
r2_ml 0.678
r2_cu 0.762
---------------------------------------------------
. spex travel2
(Greene & Hensher 1997 data on travel mode choice)
. quietly clogit choice train bus time invc, group(id) nolog
. quietly asprvalue, x(time=643.4 674.6 578.3) rest(asmean) ///
> cat(train bus) base(car) save
. estadd asprvalue, x(time=653.4 674.6 578.3) rest(asmean) ///
> cat(train bus) base(car) label(time train + 10 min) brief diff
clogit: Predictions for choice
Current Saved Diff
train .38845369 .43478274 -.04632905
bus .16446434 .15200497 .01245937
car .44708198 .41321227 .03386971
added matrices:
e(_estadd_asprval) : 1 x 3
. estadd asprvalue, x(time=643.4 684.6 578.3) rest(asmean) ///
> cat(train bus) base(car) label(time bus + 10 min) brief diff
clogit: Predictions for choice
Current Saved Diff
train .44661152 .43478274 .01182878
bus .12893429 .15200497 -.02307068
car .42445418 .41321227 .01124191
updated matrices:
e(_estadd_asprval) : 2 x 3
. estadd asprvalue, x(time=643.4 674.6 588.3) rest(asmean) ///
> cat(train bus) base(car) label(time car + 10 min) brief diff
clogit: Predictions for choice
Current Saved Diff
train .46851528 .43478274 .03373253
bus .16379826 .15200497 .01179329
car .36768648 .41321227 -.04552579
updated matrices:
e(_estadd_asprval) : 3 x 3
. estadd asprvalue post
scalars:
e(N) = 456
macros:
e(depvar) : "choice"
e(cmd) : "estadd_asprvalue"
e(model) : "clogit"
e(properties) : "b"
matrices:
e(b) : 1 x 9 (predictions)
. esttab, unstack not nostar varwidth(20)
-----------------------------------------------------------
(1)
choice
train bus car
-----------------------------------------------------------
time train + 10 min -0.0463 0.0125 0.0339
time bus + 10 min 0.0118 -0.0231 0.0112
time car + 10 min 0.0337 0.0118 -0.0455
-----------------------------------------------------------
N 456
-----------------------------------------------------------
. spex travel2
(Greene & Hensher 1997 data on travel mode choice)
. gen busXhinc = bus*hinc
. gen trainXhinc = train*hinc
. gen busXpsize = bus*psize
. gen trainXpsize = train*psize
. quietly clogit choice busXhinc busXpsize bus trainXhinc trainXpsize train ///
> time invc, group(id) nolog
. quietly asprvalue, x(psize=1) rest(asmean) base(car) save
. estadd asprvalue, x(psize=2) rest(asmean) base(car) label(_cons) brief diff
clogit: Predictions for choice
Current Saved Diff
bus .13919763 .21251462 -.07331699
train .44040644 .40365174 .0367547
car .42039591 .38383365 .03656226
added matrices:
e(_estadd_asprval) : 1 x 3
. estadd asprvalue post
scalars:
e(N) = 456
macros:
e(depvar) : "choice"
e(cmd) : "estadd_asprvalue"
e(model) : "clogit"
e(properties) : "b"
matrices:
e(b) : 1 x 3 (predictions)
. esttab, b not nostar eqlabels(none) ///
> mtitle("psize=2 - psize=1") modelw(20)
---------------------------------
(1)
psize=2 - psize=1
---------------------------------
bus -0.0733
train 0.0368
car 0.0366
---------------------------------
N 456
---------------------------------
. spex travel2
(Greene & Hensher 1997 data on travel mode choice)
. quietly asmprobit choice time invc, case(id) alternatives(mode) nolog
. estadd asprvalue, label(at means)
asmprobit: Predictions for choice
prob
Train .76511556
Bus .09945779
Car .13547295
alternative-specific variables
Train Bus Car
time 632.10965 632.10965 632.10965
invc 33.951754 33.951754 33.951754
added matrices:
e(_estadd_asprval) : 1 x 3
e(_estadd_asprval_asv) : 1 x 6
e(_estadd_asprval_csv) : 1 x 2
. estadd asprvalue, rest(asmean) label(at asmeans)
asmprobit: Predictions for choice
prob
Train .42618456
Bus .12483268
Car .44901165
alternative-specific variables
Train Bus Car
time 643.44079 674.61842 578.26974
invc 48.618421 33.144737 20.092105
updated matrices:
e(_estadd_asprval) : 2 x 3
e(_estadd_asprval_asv) : 2 x 6
e(_estadd_asprval_csv) : 2 x 2
. estadd asprvalue post, swap
scalars:
e(N) = 456
macros:
e(depvar) : "choice"
e(cmd) : "estadd_asprvalue"
e(model) : "asmprobit"
e(properties) : "b"
matrices:
e(b) : 1 x 6 (predictions)
e(asv) : 2 x 6 (time, invc)
e(csv) : 2 x 2 (hinc, psize)
. esttab, unstack not nostar nomtitle nonumber
--------------------------------------
at means at asmeans
--------------------------------------
Train 0.765 0.426
Bus 0.0995 0.125
Car 0.135 0.449
--------------------------------------
N 456
--------------------------------------
. spex wlsrnk
(1992 Wisconsin Longitudinal Study data on job values)
. label variable value1 "est"
. label variable value2 "var"
. label variable value3 "aut"
. label variable value4 "sec"
. case2alt, casevars(fem hn) rank(value) case(id) alt(hashi haslo) gen(rank)
(note: variable _altnum used since altnum() not specified)
ranks indicated by: rank
case identifier: id
case-specific interactions: est* var* aut* sec*
alternative-specific variables: hashi haslo
. rologit rank estXfem estXhn est varXfem varXhn var ///
> autXfem autXhn aut hashi haslo, group(id) reverse nolog
Rank-ordered logistic regression Number of obs = 12904
Group variable: id Number of groups = 3226
Ties handled via the exactm method Obs per group: min = 4
avg = 4.00
max = 4
LR chi2(11) = 1947.39
Log likelihood = -6127.559 Prob > chi2 = 0.0000
------------------------------------------------------------------------------
rank | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
estXfem | -.1497926 .0783025 -1.91 0.056 -.3032626 .0036774
estXhn | .1375338 .0394282 3.49 0.000 .060256 .2148116
est | -1.017202 .054985 -18.50 0.000 -1.12497 -.9094331
varXfem | -.1640212 .0728306 -2.25 0.024 -.3067666 -.0212759
varXhn | .2590404 .0370942 6.98 0.000 .1863372 .3317437
var | .528224 .0525894 10.04 0.000 .4251506 .6312974
autXfem | -.1401769 .0718325 -1.95 0.051 -.280966 .0006123
autXhn | .2133866 .0361647 5.90 0.000 .142505 .2842682
aut | -.1516741 .0510374 -2.97 0.003 -.2517055 -.0516427
hashi | .1780449 .0374744 4.75 0.000 .1045965 .2514934
haslo | -.2064148 .0425829 -4.85 0.000 -.2898758 -.1229539
------------------------------------------------------------------------------
. estadd fitstat
Measures of Fit for rologit of rank
Log-Lik Full Model: -6127.559 D(12893): 12255.119
LR(11): 1947.390
Prob > LR: 0.000
AIC: 0.951 AIC*n: 12277.119
BIC: -109780.899 BIC': -1843.272
BIC used by Stata: 12359.237 AIC used by Stata: 12277.119
added scalars:
e(dev) = 12255.119
e(dev_df) = 12893
e(lrx2) = 1947.3899
e(lrx2_df) = 11
e(lrx2_p) = 0
e(aic0) = .95141963
e(aic_n) = 12277.119
e(bic0) = -109780.9
e(bic_p) = -1843.2717
e(statabic) = 12359.237
e(stataaic) = 12277.119
e(n_rhs) = 10
e(n_parm) = 11
. estadd listcoef
rologit (N=12904): Factor Change in Odds
Odds of: ranked ahead vs ranked behind
--------------------------------------------------
rank | b z P>|z| e^b
-------------+------------------------------------
estXfem | -0.14979 -1.913 0.056 0.8609
estXhn | 0.13753 3.488 0.000 1.1474
est | -1.01720 -18.500 0.000 0.3616
varXfem | -0.16402 -2.252 0.024 0.8487
varXhn | 0.25904 6.983 0.000 1.2957
var | 0.52822 10.044 0.000 1.6959
autXfem | -0.14018 -1.951 0.051 0.8692
autXhn | 0.21339 5.900 0.000 1.2379
aut | -0.15167 -2.972 0.003 0.8593
hashi | 0.17804 4.751 0.000 1.1949
haslo | -0.20641 -4.847 0.000 0.8135
--------------------------------------------------
added matrices:
e(b_fact) : 1 x 11 (e^b)
. estadd listcoef, percent replace
rologit (N=12904): Percentage Change in Odds
Odds of: ranked ahead vs ranked behind
--------------------------------------------------
rank | b z P>|z| %
-------------+------------------------------------
estXfem | -0.14979 -1.913 0.056 -13.9
estXhn | 0.13753 3.488 0.000 14.7
est | -1.01720 -18.500 0.000 -63.8
varXfem | -0.16402 -2.252 0.024 -15.1
varXhn | 0.25904 6.983 0.000 29.6
var | 0.52822 10.044 0.000 69.6
autXfem | -0.14018 -1.951 0.051 -13.1
autXhn | 0.21339 5.900 0.000 23.8
aut | -0.15167 -2.972 0.003 -14.1
hashi | 0.17804 4.751 0.000 19.5
haslo | -0.20641 -4.847 0.000 -18.7
--------------------------------------------------
added matrices:
e(b_pct) : 1 x 11 (%)
. esttab, cell("b b_fact b_pct") scalars(aic0 aic_n bic0 bic_p)
---------------------------------------------------
(1)
rank
b b_fact b_pct
---------------------------------------------------
estXfem -.1497926 .8608865 -13.91135
estXhn .1375338 1.14744 14.74405
est -1.017202 .3616054 -63.83946
varXfem -.1640212 .848724 -15.1276
varXhn .2590404 1.295686 29.56862
var .528224 1.695918 69.59177
autXfem -.1401769 .8692045 -13.07955
autXhn .2133866 1.237863 23.78631
aut -.1516741 .8592683 -14.07317
hashi .1780449 1.194879 19.4879
haslo -.2064148 .8134955 -18.65045
---------------------------------------------------
N 12904
aic0 0.951
aic_n 12277.1
bic0 -109780.9
bic_p -1843.3
---------------------------------------------------
. spex wlsrnk
(1992 Wisconsin Longitudinal Study data on job values)
. label variable value1 "est"
. label variable value2 "var"
. label variable value3 "aut"
. label variable value4 "sec"
. case2alt, casevars(fem hn) rank(value) case(id) alt(hashi haslo) gen(rank)
(note: variable _altnum used since altnum() not specified)
ranks indicated by: rank
case identifier: id
case-specific interactions: est* var* aut* sec*
alternative-specific variables: hashi haslo
. quietly rologit rank estXfem estXhn est varXfem varXhn var ///
> autXfem autXhn aut hashi haslo, group(id) reverse nolog
. estadd asprvalue, x(fem=1 hashi=0 haslo=0) base(sec) label(fem=1) brief save
rologit: Predictions for rank
prob
est .08876531
var .41536212
aut .21456315
sec .28130943
added matrices:
e(_estadd_asprval) : 1 x 4
e(_estadd_asprval_asv) : 1 x 8
e(_estadd_asprval_csv) : 1 x 2
. estadd asprvalue, x(fem=0 hashi=0 haslo=0) base(sec) label(fem=0) brief
rologit: Predictions for rank
prob
est .09200718
var .43670157
aut .2202711
sec .25102016
updated matrices:
e(_estadd_asprval) : 2 x 4
e(_estadd_asprval_asv) : 2 x 8
e(_estadd_asprval_csv) : 2 x 2
. estadd asprvalue, x(fem=0 hashi=0 haslo=0) base(sec) label(diff) brief diff
rologit: Predictions for rank
Current Saved Diff
est .09200718 .08876531 .00324187
var .43670157 .41536212 .02133945
aut .2202711 .21456315 .00570795
sec .25102016 .28130943 -.03028926
updated matrices:
e(_estadd_asprval) : 3 x 4
. estadd asprvalue post, swap
scalars:
e(N) = 12904
macros:
e(depvar) : "rank"
e(cmd) : "estadd_asprvalue"
e(model) : "rologit"
e(properties) : "b"
matrices:
e(b) : 1 x 12 (predictions)
e(asv) : 2 x 8 (hashi, haslo)
e(csv) : 2 x 2 (fem, hn)
. esttab, not nostar unstack
---------------------------------------------------
(1)
rank
fem=1 fem=0 diff
---------------------------------------------------
est 0.0888 0.0920 0.00324
var 0.415 0.437 0.0213
aut 0.215 0.220 0.00571
sec 0.281 0.251 -0.0303
---------------------------------------------------
N 12904
---------------------------------------------------
. spex couart2
(Academic Biochemists / S Long)
. eststo poisson: quietly poisson art fem mar kid5 phd ment, nolog
. eststo nbreg: quietly nbreg art fem mar kid5 phd ment, nolog
. estadd fitstat : *
. esttab, wide scalars(r2_mf r2_mfadj aic0 aic_n) mtitles
----------------------------------------------------------------------
(1) (2)
poisson nbreg
----------------------------------------------------------------------
art
fem -0.225*** (-4.11) -0.216** (-2.98)
mar 0.155* (2.53) 0.150 (1.83)
kid5 -0.185*** (-4.61) -0.176*** (-3.32)
phd 0.0128 (0.49) 0.0153 (0.42)
ment 0.0255*** (12.73) 0.0291*** (8.38)
_cons 0.305** (2.96) 0.256 (1.85)
----------------------------------------------------------------------
lnalpha
_cons -0.817*** (-6.81)
----------------------------------------------------------------------
N 915 915
r2_mf 0.0525 0.0304
r2_mfadj 0.0491 0.0261
aic0 3.622 3.427
aic_n 3314.1 3135.9
----------------------------------------------------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. eststo poisson: quietly poisson art fem mar kid5 phd ment, nolog
. eststo nbreg: quietly nbreg art fem mar kid5 phd ment, nolog
. estadd listcoef fem ment: *
. estadd listcoef fem ment, percent nosd : *
. esttab, cell("b_facts b_pcts") keep(fem ment) mtitles
----------------------------------------------------------------
(1) (2)
poisson nbreg
b_facts b_pcts b_facts b_pcts
----------------------------------------------------------------
fem .8940439 -10.59561 .8976965 -10.23035
ment 1.274107 27.41066 1.317603 31.76034
----------------------------------------------------------------
N 915 915
----------------------------------------------------------------
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. eststo poisson: quietly poisson art fem mar kid5 phd ment, nolog
. eststo nbreg: quietly nbreg art fem mar kid5 phd ment, nolog
. estadd prchange: *
. esttab, aux(dc) nopar wide mtitles
----------------------------------------------------------------------
(1) (2)
poisson nbreg
----------------------------------------------------------------------
art
fem -0.225*** -0.359 -0.216** -0.344
mar 0.155* 0.244 0.150 0.235
kid5 -0.185*** -0.228 -0.176*** -0.216
phd 0.0128 0.0203 0.0153 0.0241
ment 0.0255*** 0.391 0.0291*** 0.443
_cons 0.305** 0.256
----------------------------------------------------------------------
lnalpha
_cons -0.817***
----------------------------------------------------------------------
N 915 915
----------------------------------------------------------------------
dc in second column
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. eststo poisson0: quietly poisson art fem mar kid5 phd ment, nolog
. eststo nbreg0: quietly nbreg art fem mar kid5 phd ment, nolog
. estadd prchange, outcome(0) : *0
. eststo poisson1: quietly poisson art fem mar kid5 phd ment, nolog
. eststo nbreg1: quietly nbreg art fem mar kid5 phd ment, nolog
. estadd prchange, outcome(1) : *1
. esttab, main(dc) nostar not scalars(outcome predval) mtitles
----------------------------------------------------------------
(1) (2) (3) (4)
poisson0 nbreg0 poisson1 nbreg1
----------------------------------------------------------------
fem 0.0725 0.0606 0.0431 0.0209
mar -0.0506 -0.0424 -0.0289 -0.0141
kid5 0.0455 0.0377 0.0277 0.0133
phd -0.00406 -0.00420 -0.00248 -0.00148
ment -0.0777 -0.0769 -0.0473 -0.0270
----------------------------------------------------------------
N 915 915 915 915
outcome 0 0 1 1
predval 0.200 0.298 0.322 0.279
----------------------------------------------------------------
dc coefficients
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. eststo poisson: quietly poisson art fem mar kid5 phd ment, nolog
. estadd prvalue
poisson: Predictions for art
Confidence intervals by delta method
95% Conf. Interval
Rate: 1.6101 [ 1.5265, 1.6937]
Pr(y=0|x): 0.1999 [ 0.1832, 0.2166]
Pr(y=1|x): 0.3218 [ 0.3116, 0.3320]
Pr(y=2|x): 0.2591 [ 0.2538, 0.2643]
Pr(y=3|x): 0.1390 [ 0.1290, 0.1491]
Pr(y=4|x): 0.0560 [ 0.0490, 0.0629]
Pr(y=5|x): 0.0180 [ 0.0149, 0.0212]
Pr(y=6|x): 0.0048 [ 0.0037, 0.0059]
Pr(y=7|x): 0.0011 [ 0.0008, 0.0014]
Pr(y=8|x): 0.0002 [ 0.0001, 0.0003]
Pr(y=9|x): 0.0000 [ 0.0000, 0.0001]
fem mar kid5 phd ment
x= .46010929 .66229508 .49508197 3.1031093 8.7672131
added matrices:
e(_estadd_prvalue) : 1 x 66
e(_estadd_prvalue_x) : 1 x 5
. eststo nbreg: quietly nbreg art fem mar kid5 phd ment, nolog
. estadd prvalue
nbreg: Predictions for art
Confidence intervals by delta method
95% Conf. Interval
Rate: 1.602 [ 1.4936, 1.7104]
Pr(y=0|x): 0.2978 [ 0.2788, 0.3167]
Pr(y=1|x): 0.2794 [ 0.2727, 0.2860]
Pr(y=2|x): 0.1889 [ 0.1859, 0.1919]
Pr(y=3|x): 0.1113 [ 0.1051, 0.1174]
Pr(y=4|x): 0.0607 [ 0.0549, 0.0664]
Pr(y=5|x): 0.0315 [ 0.0273, 0.0357]
Pr(y=6|x): 0.0158 [ 0.0130, 0.0186]
Pr(y=7|x): 0.0077 [ 0.0061, 0.0094]
Pr(y=8|x): 0.0037 [ 0.0028, 0.0046]
Pr(y=9|x): 0.0018 [ 0.0012, 0.0023]
fem mar kid5 phd ment
x= .46010929 .66229508 .49508197 3.1031093 8.7672131
added matrices:
e(_estadd_prvalue) : 1 x 66
e(_estadd_prvalue_x) : 1 x 5
. estadd prvalue post, swap: *
. esttab, b(4) nostar ci wide compress ///
> mtitles eqlabels(none)
----------------------------------------------------------------
(1) (2)
poisson nbreg
----------------------------------------------------------------
mu 1.6101 [1.5265,1.6937] 1.6020 [1.4936,1.7104]
0 0.1999 [0.1832,0.2166] 0.2978 [0.2788,0.3167]
1 0.3218 [0.3116,0.3320] 0.2794 [0.2727,0.2860]
2 0.2591 [0.2538,0.2643] 0.1889 [0.1859,0.1919]
3 0.1390 [0.1290,0.1491] 0.1113 [0.1051,0.1174]
4 0.0560 [0.0490,0.0629] 0.0607 [0.0549,0.0664]
5 0.0180 [0.0149,0.0212] 0.0315 [0.0273,0.0357]
6 0.0048 [0.0037,0.0059] 0.0158 [0.0130,0.0186]
7 0.0011 [0.0008,0.0014] 0.0077 [0.0061,0.0094]
8 0.0002 [0.0001,0.0003] 0.0037 [0.0028,0.0046]
9 0.0000 [0.0000,0.0001] 0.0018 [0.0012,0.0023]
----------------------------------------------------------------
N 915 915
----------------------------------------------------------------
95% confidence intervals in brackets
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. drop if art==0 // artificially truncated the data
(275 observations deleted)
. eststo ztp: quietly ztp art fem mar kid5 phd ment, nolog
. eststo ztnb: quietly ztnb art fem mar kid5 phd ment, nolog
. estadd fitstat : *
. esttab, wide scalars(r2_mf r2_mfadj aic0 aic_n) mtitles
----------------------------------------------------------------------
(1) (2)
ztp ztnb
----------------------------------------------------------------------
art
fem -0.229*** (-3.51) -0.245* (-2.52)
mar 0.0965 (1.32) 0.103 (0.95)
kid5 -0.142** (-2.93) -0.153* (-2.12)
phd -0.0127 (-0.41) -0.00293 (-0.06)
ment 0.0187*** (8.22) 0.0237*** (5.54)
_cons 0.671*** (5.48) 0.355 (1.80)
----------------------------------------------------------------------
lnalpha
_cons -0.603** (-2.68)
----------------------------------------------------------------------
N 640 640
r2_mf 0.0436 0.0212
r2_mfadj 0.0383 0.0146
aic0 3.394 3.232
aic_n 2172.1 2068.6
----------------------------------------------------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. drop if art==0 // artificially truncated the data
(275 observations deleted)
. eststo ztp: quietly ztp art fem mar kid5 phd ment, nolog
. eststo ztnb: quietly ztnb art fem mar kid5 phd ment, nolog
. estadd listcoef fem ment: *
. estadd listcoef fem ment, percent nosd : *
. esttab, cell("b_facts b_pcts") keep(fem ment) mtitles
----------------------------------------------------------------
(1) (2)
ztp ztnb
b_facts b_pcts b_facts b_pcts
----------------------------------------------------------------
fem .8926405 -10.73595 .8855335 -11.44665
ment 1.213629 21.36292 1.277855 27.78551
----------------------------------------------------------------
N 640 640
----------------------------------------------------------------
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. drop if art==0 // artificially truncated the data
(275 observations deleted)
. eststo ztp: quietly ztp art fem mar kid5 phd ment, nolog
. eststo ztnb: quietly ztnb art fem mar kid5 phd ment, nolog
. estadd prchange: *
. esttab, aux(dc) nopar wide mtitles
----------------------------------------------------------------------
(1) (2)
ztp ztnb
----------------------------------------------------------------------
art
fem -0.229*** -0.463 -0.245* -0.389
mar 0.0965 0.195 0.103 0.164
kid5 -0.142** -0.216 -0.153* -0.183
phd -0.0127 -0.0257 -0.00293 -0.00466
ment 0.0187*** 0.398 0.0237*** 0.396
_cons 0.671*** 0.355
----------------------------------------------------------------------
lnalpha
_cons -0.603**
----------------------------------------------------------------------
N 640 640
----------------------------------------------------------------------
dc in second column
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. drop if art==0 // artificially truncated the data
(275 observations deleted)
. eststo ztp0: quietly ztp art fem mar kid5 phd ment, nolog
. eststo ztnb0: quietly ztnb art fem mar kid5 phd ment, nolog
. estadd prchange, outcome(0) : *0
. eststo ztp1: quietly ztp art fem mar kid5 phd ment, nolog
. eststo ztnb1: quietly ztnb art fem mar kid5 phd ment, nolog
. estadd prchange, outcome(1) : *1
. esttab, main(dc) nostar not scalars(outcome predval) mtitles
----------------------------------------------------------------
(1) (2) (3) (4)
ztp0 ztnb0 ztp1 ztnb1
----------------------------------------------------------------
fem 0.0610 0.0662 0.0622 0.0207
mar -0.0259 -0.0281 -0.0263 -0.00874
kid5 0.0278 0.0307 0.0292 0.00993
phd 0.00331 0.000781 0.00348 0.000253
ment -0.0510 -0.0661 -0.0534 -0.0213
----------------------------------------------------------------
N 640 640 640 640
outcome 0 0 1 1
predval 0.129 0.315 0.264 0.270
----------------------------------------------------------------
dc coefficients
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. drop if art==0 // artificially truncated the data
(275 observations deleted)
. eststo ztp: quietly ztp art fem mar kid5 phd ment, nolog
. estadd prvalue
ztp: Predictions for art
Uncond Cond
Rate: 2.0507 2.3534
Pr(y=0|x): 0.1286 .
Pr(y=1|x): 0.2638 0.3028
Pr(y=2|x): 0.2705 0.3104
Pr(y=3|x): 0.1849 0.2122
Pr(y=4|x): 0.0948 0.1088
Pr(y=5|x): 0.0389 0.0446
Pr(y=6|x): 0.0133 0.0152
Pr(y=7|x): 0.0039 0.0045
Pr(y=8|x): 0.0010 0.0011
Pr(y=9|x): 0.0002 0.0003
fem mar kid5 phd ment
x= .440625 .671875 .471875 3.1539765 10.14375
added matrices:
e(_estadd_prvalue) : 1 x 77
e(_estadd_prvalue_x) : 1 x 5
. eststo ztnb: quietly ztnb art fem mar kid5 phd ment, nolog
. estadd prvalue
ztnb: Predictions for art
Uncond Cond
Rate: 1.6097 2.3505
Pr(y=0|x): 0.3152 .
Pr(y=1|x): 0.2698 0.3940
Pr(y=2|x): 0.1787 0.2609
Pr(y=3|x): 0.1067 0.1559
Pr(y=4|x): 0.0603 0.0881
Pr(y=5|x): 0.0329 0.0481
Pr(y=6|x): 0.0175 0.0256
Pr(y=7|x): 0.0092 0.0134
Pr(y=8|x): 0.0047 0.0069
Pr(y=9|x): 0.0024 0.0035
fem mar kid5 phd ment
x= .440625 .671875 .471875 3.1539765 10.14375
added matrices:
e(_estadd_prvalue) : 1 x 77
e(_estadd_prvalue_x) : 1 x 5
. estadd prvalue post, swap: *
. esttab, b(4) nostar not mtitles eqlabels(none)
--------------------------------------
(1) (2)
ztp ztnb
--------------------------------------
mu 2.0507 1.6097
0 0.1286 0.3152
1 0.2638 0.2698
2 0.2705 0.1787
3 0.1849 0.1067
4 0.0948 0.0603
5 0.0389 0.0329
6 0.0133 0.0175
7 0.0039 0.0092
8 0.0010 0.0047
9 0.0002 0.0024
--------------------------------------
N 640 640
--------------------------------------
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. eststo zip: quietly zip art fem mar kid5 phd ment, ///
> inf(fem mar kid5 phd ment) nolog
. eststo zinb: quietly zinb art fem mar kid5 phd ment, ///
> inf(fem mar kid5 phd ment) nolog
. estadd fitstat : *
. esttab, wide scalars(r2_mf r2_mfadj aic0 aic_n) mtitles
----------------------------------------------------------------------
(1) (2)
zip zinb
----------------------------------------------------------------------
art
fem -0.209*** (-3.30) -0.196** (-2.59)
mar 0.104 (1.46) 0.0976 (1.16)
kid5 -0.143** (-3.02) -0.152** (-2.80)
phd -0.00617 (-0.20) -0.000700 (-0.02)
ment 0.0181*** (7.89) 0.0248*** (7.10)
_cons 0.641*** (5.28) 0.417** (2.90)
----------------------------------------------------------------------
inflate
fem 0.110 (0.39) 0.636 (0.75)
mar -0.354 (-1.11) -1.499 (-1.60)
kid5 0.217 (1.10) 0.628 (1.42)
phd 0.00127 (0.01) -0.0377 (-0.12)
ment -0.134** (-2.96) -0.882** (-2.79)
_cons -0.577 (-1.13) -0.192 (-0.14)
----------------------------------------------------------------------
lnalpha
_cons -0.976*** (-7.21)
----------------------------------------------------------------------
N 915 915
r2_mf 0.0444 0.0372
r2_mfadj 0.0373 0.0292
aic0 3.534 3.416
aic_n 3233.5 3126.0
----------------------------------------------------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. eststo zip: quietly zip art fem mar kid5 phd ment, ///
> inf(fem mar kid5 phd ment) nolog
. eststo zinb: quietly zinb art fem mar kid5 phd ment, ///
> inf(fem mar kid5 phd ment) nolog
. estadd listcoef fem ment : *
. estadd listcoef fem ment, percent nosd : *
. esttab, cell("b_facts b_pcts") keep(fem ment) mtitles
----------------------------------------------------------------
(1) (2)
zip zinb
b_facts b_pcts b_facts b_pcts
----------------------------------------------------------------
art
fem .9009586 -9.904139 .9071068 -9.289321
ment 1.187247 18.72472 1.264998 26.49976
----------------------------------------------------------------
inflate
fem 1.056254 5.625355 1.373176 37.31758
ment .2802993 -71.97007 .0002323 -99.97677
----------------------------------------------------------------
N 915 915
----------------------------------------------------------------
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. eststo zip: quietly zip art fem mar kid5 phd ment, ///
> inf(fem mar kid5 phd ment) nolog
. eststo zinb: quietly zinb art fem mar kid5 phd ment, ///
> inf(fem mar kid5 phd ment) nolog
. estadd prchange: *
. esttab, aux(dc) nopar wide mtitles
----------------------------------------------------------------------
(1) (2)
zip zinb
----------------------------------------------------------------------
art
fem -0.209*** -0.380 -0.196** -0.331
mar 0.104 0.258 0.0976 0.164
kid5 -0.143** -0.226 -0.152** -0.198
phd -0.00617 -0.0106 -0.000700 -0.00116
ment 0.0181*** 0.594 0.0248*** 0.422
_cons 0.641*** 0.417**
----------------------------------------------------------------------
inflate
fem 0.110 0.636
mar -0.354 -1.499
kid5 0.217 0.628
phd 0.00127 -0.0377
ment -0.134** -0.882**
_cons -0.577 -0.192
----------------------------------------------------------------------
lnalpha
_cons -0.976***
----------------------------------------------------------------------
N 915 915
----------------------------------------------------------------------
dc in second column
* p<0.05, ** p<0.01, *** p<0.001
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. eststo zip0: quietly zip art fem mar kid5 phd ment, ///
> inf(fem mar kid5 phd ment) nolog
. eststo zinb0: quietly zinb art fem mar kid5 phd ment, ///
> inf(fem mar kid5 phd ment) nolog
. estadd prchange, outcome(0) : *0
. eststo zip1: quietly zip art fem mar kid5 phd ment, ///
> inf(fem mar kid5 phd ment) nolog
. eststo zinb1: quietly zinb art fem mar kid5 phd ment, ///
> inf(fem mar kid5 phd ment) nolog
. estadd prchange, outcome(1) : *1
. esttab, main(dc) nostar not scalars(outcome predval) mtitles
----------------------------------------------------------------
(1) (2) (3) (4)
zip0 zinb0 zip1 zinb1
----------------------------------------------------------------
art
fem 0.0609 0.0547 0.0439 0.0229
mar -0.0624 -0.0277 -0.0110 -0.0112
kid5 0.0429 0.0324 0.0198 0.0138
phd 0.00156 0.000186 0.00136 0.0000842
ment -0.173 -0.0752 0.00357 -0.0238
----------------------------------------------------------------
N 915 915 915 915
outcome 0 0 1 1
predval 0.258 0.269 0.236 0.278
----------------------------------------------------------------
dc coefficients
. eststo clear
. spex couart2
(Academic Biochemists / S Long)
. eststo zip: quietly zip art fem mar kid5 phd ment, ///
> inf(fem mar kid5 phd ment) nolog
. estadd prvalue
zip: Predictions for art
Expected y: 1.7032
Pr(Always0|z): 0.1388
Pr(y=0|x,z): 0.2580
Pr(y=1|x): 0.2357
Pr(y=2|x): 0.2331
Pr(y=3|x): 0.1536
Pr(y=4|x): 0.0760
Pr(y=5|x): 0.0300
Pr(y=6|x): 0.0099
Pr(y=7|x): 0.0028
Pr(y=8|x): 0.0007
Pr(y=9|x): 0.0002
x values for count equation
fem mar kid5 phd ment
x= .46010929 .66229508 .49508197 3.1031093 8.7672131
z values for binary equation
fem mar kid5 phd ment
z= .46010929 .66229508 .49508197 3.1031093 8.7672131
added matrices:
e(_estadd_prvalue) : 1 x 72
e(_estadd_prvalue_x) : 1 x 5
e(_estadd_prvalue_x2) : 1 x 5
. eststo zinb: quietly zinb art fem mar kid5 phd ment, ///
> inf(fem mar kid5 phd ment) nolog
. estadd prvalue
zinb: Predictions for art
Expected y: 1.701
Pr(Always0|z): 0.0002
Pr(y=0|x,z): 0.2687
Pr(y=1|x): 0.2784
Pr(y=2|x): 0.1987
Pr(y=3|x): 0.1204
Pr(y=4|x): 0.0665
Pr(y=5|x): 0.0346
Pr(y=6|x): 0.0172
Pr(y=7|x): 0.0083
Pr(y=8|x): 0.0039
Pr(y=9|x): 0.0018
x values for count equation
fem mar kid5 phd ment
x= .46010929 .66229508 .49508197 3.1031093 8.7672131
z values for binary equation
fem mar kid5 phd ment
z= .46010929 .66229508 .49508197 3.1031093 8.7672131
added matrices:
e(_estadd_prvalue) : 1 x 72
e(_estadd_prvalue_x) : 1 x 5
e(_estadd_prvalue_x2) : 1 x 5
. estadd prvalue post, swap: *
. esttab, b(4) nostar not wide compress ///
> mtitles eqlabels(none)
------------------------------
(1) (2)
zip zinb
------------------------------
Ey 1.7032 1.7010
All0 0.1388 0.0002
0|xy 0.2580 0.2687
1|x 0.2357 0.2784
2|x 0.2331 0.1987
3|x 0.1536 0.1204
4|x 0.0760 0.0665
5|x 0.0300 0.0346
6|x 0.0099 0.0172
7|x 0.0028 0.0083
8|x 0.0007 0.0039
9|x 0.0002 0.0018
------------------------------
N 915 915
------------------------------
. eststo clear
. spex regjob2
(Academic Biochemists / S Long)
. quietly regress job fem phd ment fel art cit
. estadd fitstat
Measures of Fit for regress of job
Log-Lik Intercept Only: -567.512 Log-Lik Full Model: -519.397
D(401): 1038.793 LR(6): 96.230
Prob > LR: 0.000
R2: 0.210 Adjusted R2: 0.198
AIC: 2.580 AIC*n: 1052.793
BIC: -1371.725 BIC': -60.162
BIC used by Stata: 1080.872 AIC used by Stata: 1052.793
added scalars:
e(dev) = 1038.7933
e(dev_df) = 401
e(lrx2) = 96.229915
e(lrx2_df) = 6
e(lrx2_p) = 1.533e-18
e(r2_adj) = .19828803
e(aic0) = 2.5803757
e(aic_n) = 1052.7933
e(bic0) = -1371.7248
e(bic_p) = -60.162312
e(statabic) = 1080.8722
e(stataaic) = 1052.7933
e(n_rhs) = 6
e(n_parm) = 7
. estadd listcoef
regress (N=408): Unstandardized and Standardized Estimates
Observed SD: .97360294
SD of Error: .8717482
-------------------------------------------------------------------------------
job | b t P>|t| bStdX bStdY bStdXY SDofX
-------------+-----------------------------------------------------------------
fem | -0.13919 -1.543 0.124 -0.0680 -0.1430 -0.0698 0.4883
phd | 0.27268 5.529 0.000 0.2601 0.2801 0.2671 0.9538
ment | 0.00119 1.692 0.091 0.0778 0.0012 0.0799 65.5299
fel | 0.23414 2.469 0.014 0.1139 0.2405 0.1170 0.4866
art | 0.02280 0.789 0.430 0.0514 0.0234 0.0528 2.2561
cit | 0.00448 2.275 0.023 0.1481 0.0046 0.1521 33.0599
-------------------------------------------------------------------------------
added matrices:
e(b_xs) : 1 x 6 (bStdX)
e(b_ys) : 1 x 6 (bStdY)
e(b_std) : 1 x 6 (bStdXY)
e(b_sdx) : 1 x 6 (SDofX)
. esttab, aux(b_std) wide scalars(aic0 aic_n bic0 bic_p)
-----------------------------------------
(1)
job
-----------------------------------------
fem -0.139 (-0.0698)
phd 0.273*** (0.267)
ment 0.00119 (0.0799)
fel 0.234* (0.117)
art 0.0228 (0.0528)
cit 0.00448* (0.152)
_cons 1.067***
-----------------------------------------
N 408
aic0 2.580
aic_n 1052.8
bic0 -1371.7
bic_p -60.16
-----------------------------------------
b_std in parentheses
* p<0.05, ** p<0.01, *** p<0.001
. spex regjob2
(Academic Biochemists / S Long)
. quietly regress job fem phd ment fel art cit
. estadd prvalue, x(ment=min) label(ment=min)
regress: Predictions for job
95% Conf. Interval
Predicted y: 2.1795 [ 2.0743, 2.2846]
fem phd ment fel art cit
x= .38970588 3.2005637 0 .61764706 2.2769608 21.715686
added matrices:
e(_estadd_prvalue) : 1 x 6
e(_estadd_prvalue_x) : 1 x 6
. estadd prvalue, x(ment=mean) label(ment=mean)
regress: Predictions for job
95% Conf. Interval
Predicted y: 2.2334 [ 2.1488, 2.318]
fem phd ment fel art cit
x= .38970588 3.2005637 45.470584 .61764706 2.2769608 21.715686
updated matrices:
e(_estadd_prvalue) : 2 x 6
e(_estadd_prvalue_x) : 2 x 6
. estadd prvalue, x(ment=max) label(ment=max)
regress: Predictions for job
95% Conf. Interval
Predicted y: 2.8108 [ 2.1369, 3.4847]
fem phd ment fel art cit
x= .38970588 3.2005637 531.99988 .61764706 2.2769608 21.715686
updated matrices:
e(_estadd_prvalue) : 3 x 6
e(_estadd_prvalue_x) : 3 x 6
. estadd prvalue post
scalars:
e(N) = 408
macros:
e(depvar) : "job"
e(cmd) : "estadd_prvalue"
e(model) : "regress"
e(properties) : "b"
matrices:
e(b) : 1 x 3 (predictions)
e(se) : 1 x 3 (standard errors)
e(LB) : 1 x 3 (lower CI bounds)
e(UB) : 1 x 3 (upper CI bounds)
e(Category) : 1 x 3 (outcome values)
e(X) : 6 x 3 (fem, phd, ment, fel, art, cit)
. esttab, ci nostar wide eqlabels(none)
------------------------------------------------
(1)
job
------------------------------------------------
ment=min 2.179 [2.074,2.285]
ment=mean 2.233 [2.149,2.318]
ment=max 2.811 [2.137,3.485]
------------------------------------------------
N 408
------------------------------------------------
95% confidence intervals in brackets
. spex tobjob2
(Academic Biochemists / S Long)
. quietly tobit jobcen fem phd ment fel art cit, ll(1) nolog
. estadd fitstat
Measures of Fit for tobit of jobcen
Log-Lik Intercept Only: -604.850 Log-Lik Full Model: -560.252
D(400): 1120.504 LR(6): 89.195
Prob > LR: 0.000
McFadden's R2: 0.074 McFadden's Adj R2: 0.061
ML (Cox-Snell) R2: 0.196 Cragg-Uhler(Nagelkerke) R2: 0.207
McKelvey & Zavoina's R2: 0.205
Variance of y*: 1.488 Variance of error: 1.182
AIC: 2.786 AIC*n: 1136.504
BIC: -1284.003 BIC': -53.128
BIC used by Stata: 1168.594 AIC used by Stata: 1136.504
added scalars:
e(dev) = 1120.5042
e(dev_df) = 400
e(lrx2) = 89.195123
e(lrx2_df) = 6
e(lrx2_p) = 4.452e-17
e(r2_mf) = .0737333
e(r2_mfadj) = .06050687
e(r2_ml) = .19636934
e(r2_cu) = .20704523
e(r2_mz) = .20535921
e(v_ystar) = 1.4875706
e(v_error) = 1.1820843
e(aic0) = 2.7855494
e(aic_n) = 1136.5042
e(bic0) = -1284.0027
e(bic_p) = -53.12752
e(statabic) = 1168.5943
e(stataaic) = 1136.5042
e(n_rhs) = 6
e(n_parm) = 8
. estadd listcoef
tobit (N=408): Unstandardized and Standardized Estimates
Observed SD: .97360294
Latent SD: 1.21966
SD of Error: 1.087237
-------------------------------------------------------------------------------
jobcen | b t P>|t| bStdX bStdY bStdXY SDofX
-------------+-----------------------------------------------------------------
fem | -0.23685 -2.032 0.043 -0.1156 -0.1942 -0.0948 0.4883
phd | 0.32258 5.047 0.000 0.3077 0.2645 0.2523 0.9538
ment | 0.00134 1.514 0.131 0.0880 0.0011 0.0722 65.5299
fel | 0.32527 2.656 0.008 0.1583 0.2667 0.1298 0.4866
art | 0.03391 0.929 0.353 0.0765 0.0278 0.0627 2.2561
cit | 0.00509 2.057 0.040 0.1683 0.0042 0.1380 33.0599
-------------------------------------------------------------------------------
added matrices:
e(b_xs) : 1 x 6 (bStdX)
e(b_ys) : 1 x 6 (bStdY)
e(b_std) : 1 x 6 (bStdXY)
e(b_sdx) : 1 x 6 (SDofX)
. esttab, aux(b_std) wide scalars(r2_mfadj r2_ml r2_cu r2_mz)
-----------------------------------------
(1)
jobcen
-----------------------------------------
model
fem -0.237* (-0.0948)
phd 0.323*** (0.252)
ment 0.00134 (0.0722)
fel 0.325** (0.130)
art 0.0339 (0.0627)
cit 0.00509* (0.138)
_cons 0.685**
-----------------------------------------
sigma
_cons 1.087***
-----------------------------------------
N 408
r2_mfadj 0.0605
r2_ml 0.196
r2_cu 0.207
r2_mz 0.205
-----------------------------------------
b_std in parentheses
* p<0.05, ** p<0.01, *** p<0.001
. spex tobjob2
(Academic Biochemists / S Long)
. gen cens = -(jobcen<=1)
. quietly cnreg jobcen fem phd ment fel art cit, censored(cens) nolog
. estadd fitstat
Measures of Fit for cnreg of jobcen
Log-Lik Intercept Only: -604.850 Log-Lik Full Model: -560.252
D(400): 1120.504 LR(6): 89.195
Prob > LR: 0.000
McFadden's R2: 0.074 McFadden's Adj R2: 0.061
ML (Cox-Snell) R2: 0.196 Cragg-Uhler(Nagelkerke) R2: 0.207
McKelvey & Zavoina's R2: 0.205
Variance of y*: 1.488 Variance of error: 1.182
AIC: 2.786 AIC*n: 1136.504
BIC: -1284.003 BIC': -53.128
BIC used by Stata: 1168.594 AIC used by Stata: 1136.504
added scalars:
e(dev) = 1120.5042
e(dev_df) = 400
e(lrx2) = 89.195123
e(lrx2_df) = 6
e(lrx2_p) = 4.452e-17
e(r2_mf) = .0737333
e(r2_mfadj) = .06050687
e(r2_ml) = .19636934
e(r2_cu) = .20704523
e(r2_mz) = .20535921
e(v_ystar) = 1.4875706
e(v_error) = 1.1820843
e(aic0) = 2.7855494
e(aic_n) = 1136.5042
e(bic0) = -1284.0027
e(bic_p) = -53.12752
e(statabic) = 1168.5943
e(stataaic) = 1136.5042
e(n_rhs) = 6
e(n_parm) = 8
. estadd listcoef
cnreg (N=408): Unstandardized and Standardized Estimates
Observed SD: .97360294
Latent SD: 1.21966
SD of Error: 1.087237
-------------------------------------------------------------------------------
jobcen | b t P>|t| bStdX bStdY bStdXY SDofX
-------------+-----------------------------------------------------------------
fem | -0.23685 -2.032 0.043 -0.1156 -0.1942 -0.0948 0.4883
phd | 0.32258 5.047 0.000 0.3077 0.2645 0.2523 0.9538
ment | 0.00134 1.514 0.131 0.0880 0.0011 0.0722 65.5299
fel | 0.32527 2.656 0.008 0.1583 0.2667 0.1298 0.4866
art | 0.03391 0.929 0.353 0.0765 0.0278 0.0627 2.2561
cit | 0.00509 2.057 0.040 0.1683 0.0042 0.1380 33.0599
-------------------------------------------------------------------------------
added matrices:
e(b_xs) : 1 x 6 (bStdX)
e(b_ys) : 1 x 6 (bStdY)
e(b_std) : 1 x 6 (bStdXY)
e(b_sdx) : 1 x 6 (SDofX)
. esttab, aux(b_std) wide scalars(r2_mfadj r2_ml r2_cu r2_mz)
-----------------------------------------
(1)
jobcen
-----------------------------------------
model
fem -0.237* (-0.0948)
phd 0.323*** (0.252)
ment 0.00134 (0.0722)
fel 0.325** (0.130)
art 0.0339 (0.0627)
cit 0.00509* (0.138)
_cons 0.685**
-----------------------------------------
sigma
_cons 1.087***
-----------------------------------------
N 408
r2_mfadj 0.0605
r2_ml 0.196
r2_cu 0.207
r2_mz 0.205
-----------------------------------------
b_std in parentheses
* p<0.05, ** p<0.01, *** p<0.001
. spex tobjob2
(Academic Biochemists / S Long)
. gen jobcen0 = jobcen if jobcen>1
(99 missing values generated)
. intreg jobcen0 jobcen fem phd ment fel art cit, nolog
Interval regression Number of obs = 408
LR chi2(6) = 89.20
Log likelihood = -560.25209 Prob > chi2 = 0.0000
------------------------------------------------------------------------------
| Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
fem | -.2368486 .1165852 -2.03 0.042 -.4653513 -.0083458
phd | .3225846 .0639229 5.05 0.000 .197298 .4478712
ment | .0013436 .0008876 1.51 0.130 -.000396 .0030832
fel | .3252657 .1224575 2.66 0.008 .0852534 .565278
art | .0339053 .0365017 0.93 0.353 -.0376367 .1054474
cit | .00509 .0024752 2.06 0.040 .0002388 .0099412
_cons | .6854061 .2182717 3.14 0.002 .2576014 1.113211
-------------+----------------------------------------------------------------
/lnsigma | .0836397 .0428043 1.95 0.051 -.0002552 .1675346
-------------+----------------------------------------------------------------
sigma | 1.087237 .0465384 .9997449 1.182386
------------------------------------------------------------------------------
Observation summary: 99 left-censored observations
309 uncensored observations
0 right-censored observations
0 interval observations
. estadd fitstat
Measures of Fit for intreg of jobcen0 jobcen
Log-Lik Intercept Only: -604.850 Log-Lik Full Model: -560.252
D(400): 1120.504 LR(6): 89.195
Prob > LR: 0.000
McFadden's R2: 0.074 McFadden's Adj R2: 0.061
ML (Cox-Snell) R2: 0.196 Cragg-Uhler(Nagelkerke) R2: 0.207
McKelvey & Zavoina's R2: 0.160
Variance of y*: 1.408 Variance of error: 1.182
AIC: 2.786 AIC*n: 1136.504
BIC: -1284.003 BIC': -53.128
BIC used by Stata: 1168.594 AIC used by Stata: 1136.504
added scalars:
e(dev) = 1120.5042
e(dev_df) = 400
e(lrx2) = 89.195124
e(lrx2_df) = 6
e(lrx2_p) = 4.452e-17
e(r2_mf) = .0737333
e(r2_mfadj) = .06050687
e(r2_ml) = .19636935
e(r2_cu) = .20704524
e(r2_mz) = .16016798
e(v_ystar) = 1.4075249
e(v_error) = 1.1820845
e(aic0) = 2.7855494
e(aic_n) = 1136.5042
e(bic0) = -1284.0027
e(bic_p) = -53.127521
e(statabic) = 1168.5943
e(stataaic) = 1136.5042
e(n_rhs) = 6
e(n_parm) = 8
. estadd listcoef
intreg (N=408): Unstandardized and Standardized Estimates
LHS vars: jobcen0 jobcen
Observed SD: .77904266
Latent SD: .48211618
SD of Error: .08363969
-------------------------------------------------------------------------------
| b t P>|t| bStdX bStdY bStdXY SDofX
-------------+-----------------------------------------------------------------
fem | -0.23685 -2.032 0.043 -0.1156 -0.4913 -0.2399 0.4883
phd | 0.32258 5.046 0.000 0.3077 0.6691 0.6382 0.9538
ment | 0.00134 1.514 0.131 0.0880 0.0028 0.1826 65.5299
fel | 0.32527 2.656 0.008 0.1583 0.6747 0.3283 0.4866
art | 0.03391 0.929 0.354 0.0765 0.0703 0.1587 2.2561
cit | 0.00509 2.056 0.040 0.1683 0.0106 0.3490 33.0599
-------------------------------------------------------------------------------
added matrices:
e(b_xs) : 1 x 6 (bStdX)
e(b_ys) : 1 x 6 (bStdY)
e(b_std) : 1 x 6 (bStdXY)
e(b_sdx) : 1 x 6 (SDofX)
. esttab, aux(b_std) wide scalars(r2_mfadj r2_ml r2_cu r2_mz)
-----------------------------------------
(1)
jobcen0
-----------------------------------------
model
fem -0.237* (-0.240)
phd 0.323*** (0.638)
ment 0.00134 (0.183)
fel 0.325** (0.328)
art 0.0339 (0.159)
cit 0.00509* (0.349)
_cons 0.685**
-----------------------------------------
lnsigma
_cons 0.0836
-----------------------------------------
N 408
r2_mfadj 0.0605
r2_ml 0.196
r2_cu 0.207
r2_mz 0.160
-----------------------------------------
b_std in parentheses
* p<0.05, ** p<0.01, *** p<0.001
. spex tobjob2
(Academic Biochemists / S Long)
. eststo tobit: quietly tobit jobcen fem phd ment fel art cit, ll(1) nolog
. gen cens = -(jobcen<=1)
. eststo cnreg: quietly cnreg jobcen fem phd ment fel art cit, censored(cens) n
> olog
. gen jobcen0 = jobcen if jobcen>1
(99 missing values generated)
. eststo intreg: quietly intreg jobcen0 jobcen fem phd ment fel art cit, nolog
. estadd prvalue, x(ment=min) label(ment=min) : *
. estadd prvalue, x(ment=mean) label(ment=mean) : *
. estadd prvalue, x(ment=max) label(ment=max) : *
. estadd prvalue post : *
. esttab, se nostar eqlabels(none) mtitles
---------------------------------------------------
(1) (2) (3)
tobit cnreg intreg
---------------------------------------------------
ment=min 2.014 2.014 2.014
(0.0695) (0.0695) (0.0695)
ment=mean 2.075 2.075 2.075
(0.0563) (0.0563) (0.0563)
ment=max 2.729 2.729 2.729
(0.435) (0.435) (0.435)
---------------------------------------------------
N 408 408 408
---------------------------------------------------
Standard errors in parentheses
. eststo clear
The three models are formally equivalent in this case and, therefore, yield identical predictions.