topic Re: Correlation of estimates in Discussions
https://community.jmp.com/t5/Discussions/Correlation-of-estimates/m-p/2595#M2595
If your independent variables are correlated, then most likely the estimates of your model parameters will be correlated.<BR /><BR />It means that you can't interpret the individual estimates by themselves. The prediction model is still fine, just the estimates are not easily interpretable.<BR /><BR />If, for example, beta1 is estimated to be 3 and beta2 is estimated to be 1, then you canNOT say that a one unit change in x1 results in a 3 unit change in Y, because x1 and x2 are correlated, so when x1 changes by 1 unit, x2 is also changing.<BR /><BR />What can be done? Remove one (or more) of the highly correlated X's from the model, and adjust your mindset to use the prediction model without trying to interpret the coefficients.Mon, 08 Nov 2010 17:37:38 GMT2010-11-08T17:37:38ZCorrelation of estimates
https://community.jmp.com/t5/Discussions/Correlation-of-estimates/m-p/2594#M2594
What does it mean, if estimates of a least squares fit are correlated. What values are ok, what values are critical? What can be done, if the correlation is high?<BR />Thanks.Mon, 08 Nov 2010 17:19:12 GMThttps://community.jmp.com/t5/Discussions/Correlation-of-estimates/m-p/2594#M2594joberi2010-11-08T17:19:12ZRe: Correlation of estimates
https://community.jmp.com/t5/Discussions/Correlation-of-estimates/m-p/2595#M2595
If your independent variables are correlated, then most likely the estimates of your model parameters will be correlated.<BR /><BR />It means that you can't interpret the individual estimates by themselves. The prediction model is still fine, just the estimates are not easily interpretable.<BR /><BR />If, for example, beta1 is estimated to be 3 and beta2 is estimated to be 1, then you canNOT say that a one unit change in x1 results in a 3 unit change in Y, because x1 and x2 are correlated, so when x1 changes by 1 unit, x2 is also changing.<BR /><BR />What can be done? Remove one (or more) of the highly correlated X's from the model, and adjust your mindset to use the prediction model without trying to interpret the coefficients.Mon, 08 Nov 2010 17:37:38 GMThttps://community.jmp.com/t5/Discussions/Correlation-of-estimates/m-p/2595#M25952010-11-08T17:37:38ZRe: Correlation of estimates
https://community.jmp.com/t5/Discussions/Correlation-of-estimates/m-p/2596#M2596
Thanks a lot.<BR />Can you give me rough numbers, what high correlation means?Mon, 08 Nov 2010 17:53:38 GMThttps://community.jmp.com/t5/Discussions/Correlation-of-estimates/m-p/2596#M2596joberi2010-11-08T17:53:38ZRe: Correlation of estimates
https://community.jmp.com/t5/Discussions/Correlation-of-estimates/m-p/2597#M2597
No, it depends on the situation and the use you are planning to make of the numbers.Mon, 08 Nov 2010 18:53:25 GMThttps://community.jmp.com/t5/Discussions/Correlation-of-estimates/m-p/2597#M25972010-11-08T18:53:25ZRe: Correlation of estimates
https://community.jmp.com/t5/Discussions/Correlation-of-estimates/m-p/2598#M2598
<HTML><HEAD></HEAD><BODY><P> I have now a correlation of one estimate with the estimate of the intercept. Does it matter? How can this be explained? When does this happen? And can it be avoided?</P></BODY></HTML>Tue, 20 Mar 2012 17:25:53 GMThttps://community.jmp.com/t5/Discussions/Correlation-of-estimates/m-p/2598#M2598joberi2012-03-20T17:25:53Z