CATPCA Cronbach’s alpha scores: interpretation

I have a question about how to interpret the Cronbach’s alpha scores of the separate dimensions in a multi-dimensional CATPCA model.

In the example included on the SPSS info page (https://www.ibm.com/support/knowledgecenter/en/SSLVMB_sub/spss/tutorials/catpca_guttman_dim.html) the alpha score for dimension 2 is .315. Yet with an Eigenvalue of 1.337 this dimension still adds explained variance to the two-dimensional model (total alpha score = .986).

How to interpret this? Does a low alpha score for dim 2 ( 1). So should users only look at Eigenvalues of the dimensions and the Total Cronbach’s alpha (or Eigenvalue/Explained variance) of the overall model? In my own current research using CATPCA I encounter similar output.

Related:

  • No Related Posts

15 nodes TSP problem in cplexlsqnonnegmilpex.m(Matlab)

I try to solve 15 nodes TSP with nonlinear objective function. I used cplexlsqnonnegmilpex in Matlab.

[cplexlsqnonnegmilpex][1] is not able to solve this problem with low gap value. And reaching solution takes a long time. (approximately 2 hours)

How can I solve this problem with low gap value within a short time?

I attached the code and the result [link text][2]of the problem at the end of the run to this box.

Thanks in advance.

[1]: /answers/storage/temp/18771-cplexlsqnonnegmilpex-problem.txt
[2]: /answers/storage/temp/18772-15noderesults.txt

Related:

  • No Related Posts

Custom constraint issue-ILOG

Hi all–I’m trying to create a custom constraint in ILOG following the extensions manual but it doesn’t seem to work. Specifically, as soon as I call *addVar* to link a variable to the constraint, the solver will terminate immediately. It doesn’t matter what the constraint actually does–the execute function can be empty and the result will be the same.

Anyone else encounter this? The following code demonstrates the issue and is mostly copied verbatim from the extensions manual (LEConstraint).

import ilog.cp.*;
import ilog.concert.*;

public class Runner {
public static void main(String[] args) {
try {
IloCP cp = new IloCP();

IloIntVar a = cp.intVar(2, 10);
IloIntVar y = cp.intVar(2, 10);
cp.add(new LEConstraint(cp, a, y));
cp.add(cp.minimize(a));
cp.solve();
} catch (IloException e) {
}
}
}

class LEConstraint extends IloCustomConstraint {
private IloIntVar _x;
private IloIntVar _y;

public LEConstraint(IloCP cp, IloIntVar x, IloIntVar y) throws IloException {
super(cp);
_x = x;
_y = y;
addVar(x);
addVar(y);
}

public void execute() {
if (isFixed(_x) && isFixed(_y))
if (getValue(_x) > getValue(_y))
setValue(_x, getValue(_x)+1);
}
}

Resulting in:

! Search terminated , no solution found.
! Number of branches : 0
! Number of fails : 0
! Total memory usage : 2.3 MB (2.3 MB CP Optimizer + 0.0 MB Concert)
! Time spent in solve : 0.01s (0.01s engine + 0.00s extraction)
! Search speed (br. / s) : 0

@Xavier Nodet

Related:

  • No Related Posts

Interaction contrast for within-subject variable in mixed design

I have a 2-variable experimental design.
D: 6 levels, between Ss
C: 2 levels, within Ss

I need to test the effect of C at D3.

Here is what I thought the correct syntax should be:
GLM C1 C2 BY D
/WSFACTOR=time 2 Polynomial
/METHOD=SSTYPE(3)
/CRITERIA=ALPHA(.05)
/WSDESIGN=time
/DESIGN=D
/LMATRIX “locate contrast at D3 ” D 0 0 1 0 0 0;
/MMATRIX “effect of C” all 1 -1.

I get this error: “This L matrix is not estimable. Hypothesis tests cannot be computed.”

Question: The LMATRIX code must be wrong. What should the correct LMATRIX line be?

Related:

Orthogonal Design – SPPS Statistics Base 24

Try to da a conjoint analysis with SPSS Statistics Base 24. To me more correct I like to create an orthogonal design. My workbook says it’s done by choose “data” and then move to “orthogonal design” but there is now way to select this. Is there a different way to create this orthogonal design or do I need to get an update? Please tell me how to do.

Related: