Hello,
I am looking at whether there is a threshold upon which the total value traded of all stocks has a negative effect on economic growth. I know that a U Test is a good option to study this. I have done the following regression and the test. However, I am not entirely sure what the output is telling me. Specifically, 90% Fieller interval is showing a min value of -0.039. I have read that Fieller interval shows the intervals where the variable of interest has a significant/insignificant and positive/negative effect on the dependent variable. Nonetheless, in my dataset I do not have a value -0.039 for value traded. Thus, what does Fieller interval actually show?
Can somebody help me with the interpretation? Lastly, what does the extreme point mean, and should one add min and max options after the comma? And if yes what values should be included in min or max?
reg gr linitial Vtraded Vtraded2 log_trade log_govsize log_school log_infl , robust
Linear regression Number of obs = 37
F(7, 29) = 16.29
Prob > F = 0.0000
R-squared = 0.7100
Root MSE = .66812
------------------------------------------------------------------------------
| Robust
gr | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
linitial | -1.334921 .3200277 -4.17 0.000 -1.989451 -.6803911
Vtraded | 1.205823 .7188942 1.68 0.104 -.2644808 2.676126
Vtraded2 | -.5845261 .2418092 -2.42 0.022 -1.079081 -.0899708
log_trade | .7635798 .1626319 4.70 0.000 .4309603 1.096199
log_govsize | -1.014386 .7306139 -1.39 0.176 -2.508659 .4798871
log_school | 2.888284 .5590793 5.17 0.000 1.744839 4.03173
log_infl | -.1681535 .1208932 -1.39 0.175 -.4154079 .0791009
_cons | 8.059888 4.008975 2.01 0.054 -.1393863 16.25916
------------------------------------------------------------------------------
utest Vtraded Vtraded2 , fieller level(90)
Specification: f(x)=x^2
Extreme point: 1.031453
Test:
H1: Inverse U shape
vs. H0: Monotone or U shape
-------------------------------------------------
| Lower bound Upper bound
-----------------+-------------------------------
Interval | .0124025 2.59707
Slope | 1.191324 -1.830288
t-value | 1.670285 -2.920065
P>|t| | .0528116 .0033531
-------------------------------------------------
Overall test of presence of a Inverse U shape:
t-value = 1.67
P>|t| = .0528
90% Fieller interval for extreme point: [-.03980864; 1.3818198]
Many thanks!
Related Posts with U test for Inverted U-shape
where does collect store its matrices?Stata 17.0 has the collect command. I wonder where does it store its intermediate and final output. …
Splitting messy string varsHello, I am hoping to split up these string vars and generate binary new vars based on whether they …
Problems with a loop to generate new variablesHello Stata Community. I'm working in my master thesis with a data base about middle class, their v…
Help with mergingHello, I have been working with a data set that uses NHANES data from 2003-2012. I wanted to add dat…
MaxDiff (Best-Worst ranking) Analysis in StataHi all, long time lurker here deciding to seek direct support from the experts here as I didn't find…
Subscribe to:
Post Comments (Atom)
0 Response to U test for Inverted U-shape
Post a Comment