|
|
Intercept/slope correlations |
|
Message/Author |
|
JM posted on Sunday, October 04, 2009 - 5:01 pm
|
|
|
I'm running a parallel proces LGM. How do I obtain/locate the correlation coefficient between the two intercepts and slopes in the output along with corresponding p-value? |
|
|
Following example 6.13 in the UG, you should have the following in your output: s2 on i1 s1 on i2 i1 with i2 s1 with s2 If you don't have these, you might have misspecified your model. |
|
JM posted on Monday, October 12, 2009 - 1:49 pm
|
|
|
Amir, Thanks, I have specified that, but which output value corresponds to the r value and associated p-statistic? I've seen reports citing the correlation coefficient and p-value, but I'm not sure where that is w/n the output. Thanks! James. |
|
|
James, The output of example 6.13 will give you the following: S1 ON I2 0.329 0.025 13.197 0.000 S2 ON I1 0.602 0.029 20.653 0.000 I2 WITH I1 0.025 0.057 0.443 0.658 S2 WITH S1 0.081 0.019 4.360 0.000 The values in the first column provides you with the regression coefficients (s2 on i1 and s1 on i2) and covariances (i2 with i1 and s2 with s1) while the last column provides the corresponding p-values. I hope that this is what you're looking for. Sincerely, Amir |
|
Back to top |
|
|