------------------------------------------------------------------------------------------------
name: <unnamed>
log: C:\Documents and Settings\Michael Rosenfeld\My Documents\newer web pages\soc_meth_p
> roj3\fall_2010_s381_logs\class14.log
log type: text
opened on: 9 Nov 2010, 12:03:35
. use "C:\Documents and Settings\Michael Rosenfeld\Desktop\cps_mar_2000_new.dta", clear
. regress incwage male age age_sq lawyers if occ1990==178|occ1990==125
Source | SS df MS Number of obs = 447
-------------+------------------------------ F( 4, 442) = 5.96
Model | 1.0759e+11 4 2.6898e+10 Prob > F = 0.0001
Residual | 1.9958e+12 442 4.5153e+09 R-squared = 0.0512
-------------+------------------------------ Adj R-squared = 0.0426
Total | 2.1034e+12 446 4.7160e+09 Root MSE = 67196
------------------------------------------------------------------------------
incwage | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
male | 17090.75 7088.051 2.41 0.016 3160.276 31021.22
age | 5026.88 1434.899 3.50 0.001 2206.808 7846.953
age_sq | -46.29393 14.42811 -3.21 0.001 -74.65014 -17.93771
lawyers | 36634.82 27921.33 1.31 0.190 -18240.23 91509.88
_cons | -99244.38 45668.63 -2.17 0.030 -188999 -9489.743
------------------------------------------------------------------------------
. jacknife: regress incwage male age age_sq lawyers if occ1990==178|occ1990==125
(running regress on estimation sample)
Jackknife replications (447)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5
.................................................. 50
.................................................. 100
.................................................. 150
.................................................. 200
.................................................. 250
.................................................. 300
.................................................. 350
.................................................. 400
...............................................
Linear regression Number of obs = 447
Replications = 447
F( 4, 446) = 10.48
Prob > F = 0.0000
R-squared = 0.0512
Adj R-squared = 0.0426
Root MSE = 6.72e+04
------------------------------------------------------------------------------
| Jackknife
incwage | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
male | 17090.75 5845.159 2.92 0.004 5603.271 28578.22
age | 5026.88 1112.785 4.52 0.000 2839.927 7213.834
age_sq | -46.29393 11.50344 -4.02 0.000 -68.9016 -23.68625
lawyers | 36634.82 7824.133 4.68 0.000 21258.08 52011.57
_cons | -99244.38 29046.14 -3.42 0.001 -156328.7 -42160.09
------------------------------------------------------------------------------
. regress incwage male age age_sq lawyers if occ1990==178|occ1990==125, robust
Linear regression Number of obs = 447
F( 4, 442) = 11.49
Prob > F = 0.0000
R-squared = 0.0512
Root MSE = 67196
------------------------------------------------------------------------------
| Robust
incwage | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
male | 17090.75 5835 2.93 0.004 5622.954 28558.54
age | 5026.88 1098.171 4.58 0.000 2868.595 7185.165
age_sq | -46.29393 11.30526 -4.09 0.000 -68.51267 -24.07518
lawyers | 36634.82 7021.79 5.22 0.000 22834.58 50435.07
_cons | -99244.38 28613.76 -3.47 0.001 -155480.3 -43008.45
------------------------------------------------------------------------------
. jackknife: regress incwage male age age_sq lawyers if occ1990==178|occ1990==125
(running regress on estimation sample)
Jackknife replications (447)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5
.................................................. 50
.................................................. 100
.................................................. 150
.................................................. 200
.................................................. 250
.................................................. 300
.................................................. 350
.................................................. 400
...............................................
Linear regression Number of obs = 447
Replications = 447
F( 4, 446) = 10.48
Prob > F = 0.0000
R-squared = 0.0512
Adj R-squared = 0.0426
Root MSE = 6.72e+04
------------------------------------------------------------------------------
| Jackknife
incwage | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
male | 17090.75 5845.159 2.92 0.004 5603.271 28578.22
age | 5026.88 1112.785 4.52 0.000 2839.927 7213.834
age_sq | -46.29393 11.50344 -4.02 0.000 -68.9016 -23.68625
lawyers | 36634.82 7824.133 4.68 0.000 21258.08 52011.57
_cons | -99244.38 29046.14 -3.42 0.001 -156328.7 -42160.09
------------------------------------------------------------------------------
. regress incwage male age age_sq lawyers if occ1990==178|occ1990==125, vce(jackknife)
(running regress on estimation sample)
Jackknife replications (447)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5
.................................................. 50
.................................................. 100
.................................................. 150
.................................................. 200
.................................................. 250
.................................................. 300
.................................................. 350
.................................................. 400
...............................................
Linear regression Number of obs = 447
Replications = 447
F( 4, 446) = 10.48
Prob > F = 0.0000
R-squared = 0.0512
Adj R-squared = 0.0426
Root MSE = 6.72e+04
------------------------------------------------------------------------------
| Jackknife
incwage | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
male | 17090.75 5845.159 2.92 0.004 5603.271 28578.22
age | 5026.88 1112.785 4.52 0.000 2839.927 7213.834
age_sq | -46.29393 11.50344 -4.02 0.000 -68.9016 -23.68625
lawyers | 36634.82 7824.133 4.68 0.000 21258.08 52011.57
_cons | -99244.38 29046.14 -3.42 0.001 -156328.7 -42160.09
------------------------------------------------------------------------------
*class started here…
. ttest incwage if occ1990==178|oc1990==125, by (occ1990)
oc1990 not found
r(111);
. ttest incwage if occ1990==178|occ1990==125, by (occ1990)
Two-sample t test with equal variances
------------------------------------------------------------------------------
Group | Obs Mean Std. Err. Std. Dev. [95% Conf. Interval]
---------+--------------------------------------------------------------------
Sociolog | 6 41508.33 2842.722 6963.219 34200.88 48815.78
Lawyers | 441 74044.33 3287.284 69032.96 67583.6 80505.06
---------+--------------------------------------------------------------------
combined | 447 73607.6 3248.139 68673.38 67224.04 79991.16
---------+--------------------------------------------------------------------
diff | -32535.99 28215.44 -87988.05 22916.07
------------------------------------------------------------------------------
diff = mean(Sociolog) - mean(Lawyers) t = -1.1531
Ho: diff = 0 degrees of freedom = 445
Ha: diff < 0 Ha: diff != 0 Ha: diff > 0
Pr(T < t) = 0.1247 Pr(|T| > |t|) = 0.2495 Pr(T > t) = 0.8753
. ttest incwage if occ1990==178|occ1990==125, by (occ1990) unequal
Two-sample t test with unequal variances
------------------------------------------------------------------------------
Group | Obs Mean Std. Err. Std. Dev. [95% Conf. Interval]
---------+--------------------------------------------------------------------
Sociolog | 6 41508.33 2842.722 6963.219 34200.88 48815.78
Lawyers | 441 74044.33 3287.284 69032.96 67583.6 80505.06
---------+--------------------------------------------------------------------
combined | 447 73607.6 3248.139 68673.38 67224.04 79991.16
---------+--------------------------------------------------------------------
diff | -32535.99 4345.953 -41456.75 -23615.24
------------------------------------------------------------------------------
diff = mean(Sociolog) - mean(Lawyers) t = -7.4865
Ho: diff = 0 Satterthwaite's degrees of freedom = 26.7692
Ha: diff < 0 Ha: diff != 0 Ha: diff > 0
Pr(T < t) = 0.0000 Pr(|T| > |t|) = 0.0000 Pr(T > t) = 1.0000
* If you will recall HW2, there was a big difference between equal variance and unequal variance t-tests as far as the lawyer-sociologist difference was concerned.
. ttest incwage if occ1990==178|occ1990==125, by (occ1990)
Two-sample t test with equal variances
------------------------------------------------------------------------------
Group | Obs Mean Std. Err. Std. Dev. [95% Conf. Interval]
---------+--------------------------------------------------------------------
Sociolog | 6 41508.33 2842.722 6963.219 34200.88 48815.78
Lawyers | 441 74044.33 3287.284 69032.96 67583.6 80505.06
---------+--------------------------------------------------------------------
combined | 447 73607.6 3248.139 68673.38 67224.04 79991.16
---------+--------------------------------------------------------------------
diff | -32535.99 28215.44 -87988.05 22916.07
------------------------------------------------------------------------------
diff = mean(Sociolog) - mean(Lawyers) t = -1.1531
Ho: diff = 0 degrees of freedom = 445
Ha: diff < 0 Ha: diff != 0 Ha: diff > 0
Pr(T < t) = 0.1247 Pr(|T| > |t|) = 0.2495 Pr(T > t) = 0.8753
. ttest incwage if occ1990==178|occ1990==125, by (occ1990) unequal
Two-sample t test with unequal variances
------------------------------------------------------------------------------
Group | Obs Mean Std. Err. Std. Dev. [95% Conf. Interval]
---------+--------------------------------------------------------------------
Sociolog | 6 41508.33 2842.722 6963.219 34200.88 48815.78
Lawyers | 441 74044.33 3287.284 69032.96 67583.6 80505.06
---------+--------------------------------------------------------------------
combined | 447 73607.6 3248.139 68673.38 67224.04 79991.16
---------+--------------------------------------------------------------------
diff | -32535.99 4345.953 -41456.75 -23615.24
------------------------------------------------------------------------------
diff = mean(Sociolog) - mean(Lawyers) t = -7.4865
Ho: diff = 0 Satterthwaite's degrees of freedom = 26.7692
Ha: diff < 0 Ha: diff != 0 Ha: diff > 0
Pr(T < t) = 0.0000 Pr(|T| > |t|) = 0.0000 Pr(T > t) = 1.0000
. regress incwage age age_sq lawyers if occ1990==178|occ1990==125
Source | SS df MS Number of obs = 447
-------------+------------------------------ F( 3, 443) = 5.94
Model | 8.1342e+10 3 2.7114e+10 Prob > F = 0.0006
Residual | 2.0220e+12 443 4.5644e+09 R-squared = 0.0387
-------------+------------------------------ Adj R-squared = 0.0322
Total | 2.1034e+12 446 4.7160e+09 Root MSE = 67560
------------------------------------------------------------------------------
incwage | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | 5192.546 1441.019 3.60 0.000 2360.463 8024.63
age_sq | -46.5991 14.50573 -3.21 0.001 -75.10769 -18.09051
lawyers | 43885.33 27909.35 1.57 0.117 -10965.85 98736.5
_cons | -101263 45908.38 -2.21 0.028 -191488.2 -11037.71
------------------------------------------------------------------------------
* Regress gives us the equal variance t-test (here slightly different because we have added some additional controls).
. jackknife: regress incwage age age_sq lawyers if occ1990==178|occ1990==125
(running regress on estimation sample)
Jackknife replications (447)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5
.................................................. 50
.................................................. 100
.................................................. 150
.................................................. 200
.................................................. 250
.................................................. 300
.................................................. 350
.................................................. 400
...............................................
Linear regression Number of obs = 447
Replications = 447
F( 3, 446) = 17.49
Prob > F = 0.0000
R-squared = 0.0387
Adj R-squared = 0.0322
Root MSE = 6.76e+04
------------------------------------------------------------------------------
| Jackknife
incwage | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | 5192.546 1121.748 4.63 0.000 2987.978 7397.114
age_sq | -46.5991 11.50142 -4.05 0.000 -69.2028 -23.9954
lawyers | 43885.33 6164.988 7.12 0.000 31769.29 56001.36
_cons | -101263 28817.54 -3.51 0.000 -157898 -44627.94
------------------------------------------------------------------------------
* The jackknife runs the same regression N times, dropping one point each time. The coefficients are unchanged, but the standard errors of the coefficients are just the standard deviation of the N trials for each beta. The bigger the N, the more time this process will take.
. regress incwage age age_sq lawyers if occ1990==178|occ1990==125
Source | SS df MS Number of obs = 447
-------------+------------------------------ F( 3, 443) = 5.94
Model | 8.1342e+10 3 2.7114e+10 Prob > F = 0.0006
Residual | 2.0220e+12 443 4.5644e+09 R-squared = 0.0387
-------------+------------------------------ Adj R-squared = 0.0322
Total | 2.1034e+12 446 4.7160e+09 Root MSE = 67560
------------------------------------------------------------------------------
incwage | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | 5192.546 1441.019 3.60 0.000 2360.463 8024.63
age_sq | -46.5991 14.50573 -3.21 0.001 -75.10769 -18.09051
lawyers | 43885.33 27909.35 1.57 0.117 -10965.85 98736.5
_cons | -101263 45908.38 -2.21 0.028 -191488.2 -11037.71
------------------------------------------------------------------------------
. regress incwage age age_sq lawyers if occ1990==178|occ1990==125,vce(jackknife)
(running regress on estimation sample)
Jackknife replications (447)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5
.................................................. 50
.................................................. 100
.................................................. 150
.................................................. 200
..........................................--Break--
r(1);
* The above is just another syntax for the same jackknife request, so I interrupted it rather than wait for the output.
. regress incwage age age_sq lawyers if occ1990==178|occ1990==125,vce(robust)
Linear regression Number of obs = 447
F( 3, 443) = 19.63
Prob > F = 0.0000
R-squared = 0.0387
Root MSE = 67560
------------------------------------------------------------------------------
| Robust
incwage | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | 5192.546 1107.785 4.69 0.000 3015.38 7369.713
age_sq | -46.5991 11.30753 -4.12 0.000 -68.82216 -24.37604
lawyers | 43885.33 5794.468 7.57 0.000 32497.27 55273.39
_cons | -101263 28523.55 -3.55 0.000 -157321.3 -45204.69
------------------------------------------------------------------------------
* Robust regression is a one-step alteration of the standard errors which is described in detail in the Stata manual, and which in this case yields results similar to the jackknife, and similar to the unequal variance regression which most students thought was preferable to the equal variance assumption in this case…
. log close
name: <unnamed>
log: C:\Documents and Settings\Michael Rosenfeld\My Documents\newer web pages\so
> c_meth_proj3\fall_2010_s381_logs\class14.log
log type: text
closed on: 9 Nov 2010, 16:00:38
----------------------------------------------------------------------------------------