Hello,

adding a squareterm of my independent variable I noticed an inverse curvlinear relationship.
I wanted to prove this with an utest.
The utest confirmed an inverse curvlinear realtionship, however it generates missing values. The number of missing values is higher than the obsevations included in my regression.
And above all in all the posts I read about utest here, I did not notice any missing values.
Is there any explanation for those missing values?


I ran the following regression:
regress nov scp scp_square fueausgaben size Gruppe1 Gruppe2 Gruppe4, vce(robust)

--------------------------------------------------------------------------------------
and afterwards the utest:


utest scp scp_square
(2,429 missing values generated)

Specification: f(x)=x^2
Extreme point: 2.630221

Test:
H1: Inverse U shape
vs. H0: Monotone or U shape

-------------------------------------------------
| Lower bound Upper bound
-----------------+-------------------------------
Interval | 0 8
Slope | .5074897 -1.036075
t-value | 4.291503 -3.378399
P>|t| | 9.44e-06 .0003739
-------------------------------------------------

Overall test of presence of a Inverse U shape:
t-value = 3.38
P>|t| = .000374


Another question: am I right that utest does not work after runing a Tobit regression?

Many Thanks in advance.