Dear friends, why the test results using ttable and pstest are not consistent.


Code:
pstest $xlist Y  , t(treated)
The results is



Mean t-test V(T)/
Variable Treated Control %bias t p>t V(C)

x1 .42697 .38202 9.0 0.61 0.544 .
x2 30.427 31.517 -13.1 -0.87 0.387 0.84
y -.17585 -.39273 27.7 1.73 0.085 0.60*

* if variance ratio outside [0.66; 1.52]


Ps R2 LR chi2 p>chi2 MeanBias MedBias B R %Var

0.030 7.20 0.616 11.1 10.8 40.3* 0.51 17

* if B>25%, R outside [0.5; 2]

If I use

Code:
ttest x1 Y, by(treated)  
ttest x2 Y, by(treated)
The result is changed,


Two-sample t test with equal variances

Group Obs Mean Std. Err. Std. Dev. [95% Conf. Interval]

0 1,368 32.70029 .192066 7.103845 32.32352 33.07707
1 64 29.29688 .8038473 6.430778 27.69051 30.90324

combined 1,432 32.54818 .1878338 7.107966 32.17973 32.91664

diff 3.403417 .9048948 1.628354 5.178481

diff = mean(0) - mean(1) t = 3.7611
Ho: diff = 0 degrees of freedom = 1430

Ha: diff < 0 Ha: diff != 0 Ha: diff > 0
Pr(T < t) = 0.9999 Pr(T > t) = 0.0002 Pr(T > t) = 0.0001
Or simply,
[/QUOTE]
Two-sample t test with equal variances

Variables G1(0) Mean1 G2(1) Mean2 MeanDiff

x1 2631 34.269 89 30.427 3.843***
x2 2631 1248.832 89 989.303 259.529***

[/QUOTE]

The latter test shows that there was no parallel trend.

Thanks a lot.