topic Re: In JMP, how to adjust the encoding of categorical variables in a regression model? in Discussions

In JMP, how do you adjust the parameterization of categorical variables in a regression model?

lyz6 — Wed, 14 Feb 2018 15:39:55 GMT

In JMP, how to adjust the encoding of categorical variables in a regression model?

Re: In JMP, how to adjust the encoding of categorical variables in a regression model?

Peter_Bartell — Thu, 23 Jun 2016 11:52:03 GMT

lyz6: Can you perhaps provide an example of what you mean by adjusting encoding of categorical variables in a regression model? It's not clear to me what you are trying to accomplish.

Re: In JMP, how to adjust the encoding of categorical variables in a regression model?

shoffmeister — Thu, 23 Jun 2016 12:36:18 GMT

As far as I know, if you want to use dummy-coding instead of effect-coding in your models you will have to create the dummy-variables manually.

In most recent versions of JMP you can do so by "Cols=>Utilities=>Make Indicator Columns". Then use the dummy-variables as factors in your model (of course you should only use all but one).

Hope that helps.

Re: In JMP, how to adjust the encoding of categorical variables in a regression model?

ron_horne — Fri, 28 Oct 2016 13:27:57 GMT

hi lyz6,

try this:

Comparison the results of Standard Least Squares in Fit model with the results of SAS Proc Reg

and then this:

JMP estimates vs. manual calculations

Re: In JMP, how to adjust the encoding of categorical variables in a regression model?

jiancao — Wed, 19 Oct 2016 03:11:04 GMT

Creating dummy variables is one approach. But if your model is standard least squares, there is an option called Indicator Parameterization Estimates that does just that.