Hi all,
I would run a repeated measures with Linear Mixed Models. In attachment you can find the following dataset. In my study, the dependent variable is jump performance (HOPPING VARIABLE). Each participant has four repeated measurements, resulting from a combination of two factors:
Time: pre and post intervention
Leg: dominant and non-dominant
If i use the residual methods, than Satterthwaite, I have this message:
|
Warnings
|
|
The final Hessian matrix is not positive definite although all convergence criteria are satisfied. The MIXED procedure continues despite this warning. Validity of subsequent results cannot be ascertained.
|
While if I use Satterthwaite methods, I have a great standard error in covariance parameters
|
Estimates of Covariance Parametersa
|
|
Parameter
|
Estimate
|
Std. Error
|
Wald Z
|
Sig.
|
95% Confidence Interval
|
|
Lower Bound
|
Upper Bound
|
|
Repeated Measures
|
UN (1,1)
|
.132
|
13725.211
|
.000
|
1.000
|
.000
|
.
|
|
UN (2,1)
|
.063
|
13725.211
|
.000
|
1.000
|
-26900.856
|
26900.982
|
|
UN (2,2)
|
.100
|
13725.211
|
.000
|
1.000
|
.000
|
.
|
|
UN (3,1)
|
.054
|
13725.211
|
.000
|
1.000
|
-26900.865
|
26900.972
|
|
UN (3,2)
|
.019
|
13725.211
|
.000
|
1.000
|
-26900.900
|
26900.937
|
|
UN (3,3)
|
.076
|
13725.211
|
.000
|
1.000
|
.000
|
.
|
|
UN (4,1)
|
.046
|
13725.211
|
.000
|
1.000
|
-26900.873
|
26900.964
|
|
UN (4,2)
|
.044
|
13725.211
|
.000
|
1.000
|
-26900.875
|
26900.962
|
|
UN (4,3)
|
.026
|
13725.211
|
.000
|
1.000
|
-26900.893
|
26900.944
|
|
UN (4,4)
|
.097
|
13725.211
|
.000
|
1.000
|
.000
|
.
|
|
CODICE
|
Variance
|
.099
|
13725.211
|
.000
|
1.000
|
.000
|
.
|
|
a. Dependent Variable: HOPPING .
|
Someone could help me, please?
Thanks all!
In attachment, you can find also the output of results
My goal is to evaluate whether a specific type of training intervention had an effect on jump performance, considering both leg and time as repeated factors.
I am not entirely sure how to correctly define the repeated structure in SPSS MIXED, particularly regarding the choice of covariance structure and whether the model is correctly specified for this within-subject design. I
