Uniform Confidence Intervals

[FernandoRios]

Dear all
Before I got down the rabbit hole, is anyone aware of a utility that constructs Uniform Confidence Intervals, given a set of coefficients and Variance Covariance matrices?

Thank you
F

[Jared Greathouse]

Perhaps this might have part of the answer?

[FernandoRios]

Thank you Jared
Yes, I saw that. There is also the option from csdid. But both are command specific. Was wondering about something more general.
F

[Israel Garcia]

Hi Fernando,
I have been looking for the same thing the last days and I found this interesting approach. I do not know how general it gets but it seems pretty general.
Israel

[Leonardo Guizzetti]

Could you describe what exactly is a uniform confidence interval? I've never heard of this term before.

[FernandoRios]

THank you Israel, I didnt see that before. So, started writing my own code(with a small problem with some random numbers I hope to fix soon)

Leonardo Guizzetti It was also quite new to me when I started to work on DID (csdid in specific)
The uniform confidence intervals or uniform /simultaneous confidence bands are Confidence intervals that are valid for multiple testing
They are kind of an alternative to corrections for simultanous testing.

The procedure basically finds a new critical value so only alpha percent of any of the tested coefficients falls above that threshold. (this statistic is larger than the standard critical values)
(Hope i didnt mixed the idea too much)

[Leonardo Guizzetti]

Thank you for the explanation. I have occasionally used simultaneous CIs but I had not heard the other name.

[Israel Garcia]

FernandoRios I would be very thankful if you are able/willing to share your code. I follow your site and it is always very helpful