Page 1 of 1

Singular Matrix Problem

Posted: Thu Jul 04, 2019 6:44 pm
by Sinke
Hello,

When i want to do a multigroupe analysis of My Formative Model with Smart pls to seperate People in two groups Theres a Singular Matrix Problem. I read, that it can help to check the correlations of the indicators, But i cant find this Function. I only found the correlation Matrix of the variables But Not of Indicators.

Can someone help me with My Problem. I‘m really desperate because i‘m Not a statistical expert :(

Re: Singular Matrix Problem

Posted: Fri Jul 12, 2019 12:10 pm
by jmbecker
The "indicators" is another word for the items/manifest variables that you assign to a construct as measurement model. Thus, they are the variables in your dataset.

A common problem with singular matrix in MGA is fact that people include the grouping variable in their model (e.g., gender). This variable will have no (zero) variance after splitting the dataset into each group and thus cannot be used within the model anymore.

Re: Singular Matrix Problem

Posted: Sat Jul 13, 2019 11:12 am
by Sinke
I didnt include Group variables in My Model. After Dome Research i think now, that of My groupe samples Are too small (for example there Are only 23 female persons).
Maybe Thats too less for My Model.

Re: Singular Matrix Problem

Posted: Thu Jul 18, 2019 8:28 am
by lockylaw
Indeed that is a little too small. Otherwise, you can try lowering the subsamples and repeating bootstrapping until you get a result.

Re: Singular Matrix Problem

Posted: Thu Jul 18, 2019 3:21 pm
by Sinke
Thanks for your help, i will Try that :)

Re: Singular Matrix Problem

Posted: Mon Oct 07, 2019 8:08 am
by zuna
i have the same problems. I am newcomers to use the SmartPLS. but when I'm Run. the calculating said it is SINGULAR MATRIX PORBLEMs. I have 100 samples and 18 Indicators. how to solve the problem ? can everyone help me? thankyou

Re: Singular Matrix Problem

Posted: Thu Oct 10, 2019 7:44 am
by jmbecker
The problem message already gives some suggestion to resolve the problem. Here we discussed some more. Did you try any of these?