Dear all,
I have performed the following Mahalanobis distance command to test for multivariate outliers:
multnorm resi1 resi2 resi3
display invchi2(3, 0.999)
In the output, I have a table with a column for MD2. These values are important to me because by comparing them with the chi2 value obtained by the previous test I can identify the outliers. Checking the process run by SPSS, I noted that MD2 is ordered by id (observation) adding a column in the data editor.
My question is if we can do the same with stata and if yes, what is the command.
Thank you for your support.
Regards.
Maria
I have performed the following Mahalanobis distance command to test for multivariate outliers:
multnorm resi1 resi2 resi3
display invchi2(3, 0.999)
In the output, I have a table with a column for MD2. These values are important to me because by comparing them with the chi2 value obtained by the previous test I can identify the outliers. Checking the process run by SPSS, I noted that MD2 is ordered by id (observation) adding a column in the data editor.
My question is if we can do the same with stata and if yes, what is the command.
Thank you for your support.
Regards.
Maria
