Ramsey RESET test

[Daniel Gomez]

Hi,

How can I interpret these results from the Ramsey RESET test?
How can I know at which level can I reject the null hypothesis?

1. Prob > F = 0.0000

2. Prob > chi2 = 0.0171

3. Prob > chi2 = 0.9243


I know the Ramsey RESET test detects if potential variables have been omitted in the model.
The null hypothesis states that the model is correctly specified.

Thanks!!

[Carlo Lazzaro]

Daniel:
the issue here is how such results are conventionally reported in a working-paper/reasearch report/article.
One suggestion is to present the actual p-values with no more than two decimal digits (https://oxfordmedicine.com/view/10.1...-9780198599661).
That said:
1) p<0.0001
2) p<0.05
3) p=0.92 the F-test shows no evidence that the null can be rejected.

As an aside, it would be wonderful if RESET, like a superior entity that knows your dataset and its data generating process, could tell us what predictor(s)/interaction(s) to add to have our regression model correctly specified. Unfortunately, this is not the case, as -estat ovtest- or -linktest- simply warns us about the fact that the relationship between the regressand and one or more predictors is not linear but is rather quadratic or cubic.
I do believe that the power of sniffing out omitted variables attributed to RESET is a superb (but misleading) advertising stunt in favour of https://en.wikipedia.org/wiki/James_B._Ramsey, a towering Canadian econometrician.

[Daniel Gomez]

Carlo Lazzaro Thank you!