Hello !!!

Since simple multiple linear regression doesn't take into account the within-subject model, nor the fact that my dependent variable is an ordinal variable, I need to control for individual heterogeinity. For this, it's possible to run a regression model with clustered standard errors to take into account the pattern within the subject.

Can anyone explain how to run a regression model with clustered standard errors on ibm spss?

Explanation of my survey:

112 participants indicated on a scale of 1 to 7 their level of agreement with different statements after reading a scenario. 7 Likert scale (from strongly disagree to strongly agree)

The statements measured my dependent variables: work motivation, bonus satisfaction, collaboration and help. There were a total of 7 different scenarios. I also control for gender, age, and status ( employed, self-employed, student, unemployed, retired, other).

My aim is to see how the different scenarios affect the dependent variables.

My thesis supervisor advised me to take into account clustered standard errors in my regression model, but I have no idea how to do this on spss. I can't find the right test and command to do this.

Could someone help me?

Thanks in advance,

Best regards