proportion of variance explained

[Makrem Ben Youssef]

Hi Phil

I have some trouble with the proportion of variance explained. when it is a one factor model MIRT displays the propotion of variance explained , but when the model is three-dimensional MIRT just displays the factor correlations without information on the proportion of variance provided by each factor. Is there any syntax to obtain this information?

Thanks for your Help

Makrem

[Phil Chalmers]

Hi Makrem,

I'm not sure what you mean, I don't recall ever implementing a proportion explained variance output in mirt. Could you provide an example of what you mean?

Phil

--
You received this message because you are subscribed to the Google Groups "mirt-package" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mirt-package...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Makrem Ben Youssef]

Hi phil

this is the output from MIRT

SS loadings:  10.615 

Proportion Var:  0.342 

unless Proportion var dosen't refer to "percentage of variance explained" like it's the case for traditional exploratory factor analysis.

Thank you fot the clarification.

Makrem

[Phil Chalmers]

For EFA models the proportion explained doesn't make much sense when there is rotation involved (which there always should be). That's why only the SS explained is returned, and is the same for traditional linear factor analysis. E.g.

> library(mirt)

Loading required package: stats4

Loading required package: lattice

> mod <- mirt(Science, 1)

Iteration: 36, Log-Lik: -1608.870, Max-Change: 0.00010

> summary(mod)

           F1    h2

Comfort 0.522 0.273

Work    0.584 0.342

Future  0.803 0.645

Benefit 0.541 0.293

SS loadings:  1.552 

Proportion Var:  0.388 

Factor correlations: 

   F1

F1  1

> mod2 <- mirt(Science, 2)

Iteration: 313, Log-Lik: -1601.969, Max-Change: 0.00010

> summary(mod2)

Rotation:  oblimin 

Rotated factor loadings: 

            F1     F2    h2

Comfort  0.602 -0.031 0.382

Work    -0.057 -0.797 0.592

Future   0.330 -0.515 0.548

Benefit  0.723  0.024 0.506

Rotated SS loadings:  0.997 0.902 

Factor correlations: 

       F1     F2

F1  1.000 -0.511

F2 -0.511  1.000

> summary(mod2, rotate = 'varimax')

Rotation:  varimax 

Rotated factor loadings: 

           F1    F2    h2

Comfort 0.216 0.579 0.382

Work    0.760 0.121 0.592

Future  0.605 0.428 0.548

Benefit 0.200 0.683 0.506

Rotated SS loadings:  1.03 0.999 

Factor correlations: 

   F1 F2

F1  1  0

F2  0  1

> summary(mod2, rotate = 'none')

Unrotated factor loadings: 

           F1     F2    h2

Comfort 0.559  0.264 0.382

Work    0.629 -0.444 0.592

Future  0.731 -0.116 0.548

Benefit 0.620  0.349 0.506

SS loadings:  1.627 0.402 

Proportion Var:  0.407 0.1 

Factor correlations: 

   F1 F2

F1  1  0

F2  0  1

--

[Makrem Ben Youssef]

Hi Phil

Thank you for the clarification and the example. it's really helpful.

Makrem

Le lundi 24 août 2015 13:09:21 UTC+2, Makrem Ben Youssef a écrit :