Author Topic: R square in latent regregression  (Read 579 times)

gonzalezdepaz

  • Newbie
  • *
  • Posts: 10
    • View Profile
R square in latent regregression
« on: September 24, 2012, 07:44:04 AM »
 
After run an analysis and produce the estimated coefficients of the regression model from an 24 item set and four dimensions, I’m stuck in trying to obtain the R-square values for each of the dimensions of the model.  I wonder if you could send me the formulae, or any advice to compute these determination coefficients in ConQuest or any collateral software as it is in the Conquest manual on chapters of latent regression (both unidimensional and multidimensional).

 

Thanks for any advice,

Luis
 

ConQuest

  • Global Moderator
  • Newbie
  • *****
  • Posts: 24
    • View Profile
    • Email
Re: R square in latent regregression
« Reply #1 on: September 24, 2012, 11:52:53 PM »
Luis, to get the R-squared do a run with regressors and one without.  In each case you'll get an estimate of the (conditional) variance of the latent variable.  Let's call V, the variance without regressors and VR the variance with regressors.  The R-squared = 1-VR/V=(V-VR)/V

ie the proporation of total variance that is explained by the regressors.

gonzalezdepaz

  • Newbie
  • *
  • Posts: 10
    • View Profile
Re: R square in latent regregression
« Reply #2 on: September 25, 2012, 06:28:03 PM »
Excellent!!

Thanks!!

Luis

gonzalezdepaz

  • Newbie
  • *
  • Posts: 10
    • View Profile
Re: R square in latent regregression
« Reply #3 on: October 05, 2012, 05:27:54 AM »
After running a Regression analysis I can see the coefficients on the table, but the summary states:

Regression: not applicable.

What does it mean?

Thanks

Luis


ConQuest

  • Global Moderator
  • Newbie
  • *****
  • Posts: 24
    • View Profile
    • Email
Re: R square in latent regregression
« Reply #4 on: October 07, 2012, 08:56:53 PM »
luis, the formula is in the previous post