Oprobit dependent variables only categorical?

[Oksana Shchybria]

Dear all,
I am not that familiar with ordered probit regression.
Please, tell me, should all dependent variables in oprobit be only categorical? Or I can use variables such as GDP, population with or without log?
Thanks in advance!

[Richard Williams]

In oprobit the DV should be categorical, where the categories are ordered, e.g. 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree.

I wonder if you meant independent instead of dependent though. If so, then sure, things like GDP could be independent variables.

If you need a brief overview of ordinal models, you can see

https://www3.nd.edu/~rwilliam/xsoc73994/Ologit01.pdf

or else

https://methods.sagepub.com/Foundati...ression-models

[Oksana Shchybria]

Dear Richard,

Thank you for your reply. You are right, I meant INdependent variables. Should all of them be categorical, or I can use GDP in logs or population as independent variables?

In oprobit the DV should be categorical, where the categories are ordered, e.g. 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree.

I wonder if you meant independent instead of dependent though. If so, then sure, things like GDP could be independent variables.

If you need a brief overview of ordinal models, you can see

https://www3.nd.edu/~rwilliam/xsoc73994/Ologit01.pdf

or else

https://methods.sagepub.com/Foundati...ression-models

[Richard Williams]

The regression model places no restrictions on the values that the independent variables take on. They may be continuous, interval level (net worth of a company), they may be only positive or zero (percent of vote a party received) or they may be dichotomous (dummy) variables (1 = male, 0 = female). oprobit estimates a generalized linear model, and the right hand side can contain the same sorts of variables that an OLS regression could have.

For more, you might look at

https://www3.nd.edu/~rwilliam/xsoc73994/L01.pdf

[Oksana Shchybria]

Dear Richard,

Thanks a lot for your detailed explanation and for your suggested readings, they are very helpful!

Best of luck to you!

The regression model places no restrictions on the values that the independent variables take on. They may be continuous, interval level (net worth of a company), they may be only positive or zero (percent of vote a party received) or they may be dichotomous (dummy) variables (1 = male, 0 = female). oprobit estimates a generalized linear model, and the right hand side can contain the same sorts of variables that an OLS regression could have.

For more, you might look at

https://www3.nd.edu/~rwilliam/xsoc73994/L01.pdf

[Richard Williams]

If you want a brief overview of ordinal models, see

https://www3.nd.edu/~rwilliam/xsoc73994/Ologit01.pdf

If you want a bit more depth, see

https://methods.sagepub.com/Foundati...ression-models

If you can't get it for free and you want it, email me directly.

There are, of course, numerous other good intros to ordinal regression, but the ones I've written are the ones I conveniently have links for!

[Anand Sunny]

The regression model places no restrictions on the values that the independent variables take on. They may be continuous, interval level (net worth of a company), they may be only positive or zero (percent of vote a party received) or they may be dichotomous (dummy) variables (1 = male, 0 = female). oprobit estimates a generalized linear model, and the right hand side can contain the same sorts of variables that an OLS regression could have.

For more, you might look at

https://www3.nd.edu/~rwilliam/xsoc73994/L01.pdf

Sir, I have to use ordered probit for an analysis in which the DV is ordered and Independent variables includes ordered, nominal and continuous variables. Do I have to create dummies for categorical independent variables (more than 2 levels). What is the logic behind prefixing variables with 'i.' in stata?