Obtain covariance matrix with regression results

Question

I have regression results as follows:

I also ran a variance-covariance matrix that output the following:

It's clear that squaring the standard errors provides the variance on the diagonal. With just this, however, can you calculate the .000027, or the covariance of a and b?

Do you want to know what to use the covariance for? – mdewey Mar 07 '17 at 09:45 — mdewey, Mar 07 '17 at 09:45

score 2 · Accepted Answer · edited Apr 13 '17 at 12:44

2

You can't. You've stated the result back-to-front. The SE is the square root of the main diagonal of the covariance matrix of parameter estimates. You need to start with the covariance matrix of parameter estimates, rather than start with the standard errors.

You can find the formula for the covariance matrix of parameter estimates for OLS here.

edited Apr 13 '17 at 12:44

Community

1

answered Mar 07 '17 at 05:36

Tim

3,255
14
24

Obtain covariance matrix with regression results

1 Answers1