topic Re: Can I add one Ordinal and one Nominal (Columns) to a Multiple Linear Regression model? in Discussions

Can I add one Ordinal and one Nominal (Columns) to a Multiple Linear Regression model?

Denis — Fri, 09 Jun 2023 00:27:27 GMT

First I don't know if this question is logical to you, but i am trying to analyse a datasheet and there is one Nominal columns and one Ordinal. I have done one Multiple Linear Regression with only one Nominal columns, but i dont know if things change if i add another Ordinal columns?

Re: Can I add one Ordinal and one Nominal (Columns) to a Multiple Linear Regression model?

Mark_Bailey — Mon, 11 Jan 2021 18:46:32 GMT

I assume that you mean that you have a nominal data column and an ordinal data column and you want to use both of them as effects when launching Fit Least Squares. Yes, you can. Select Analyze > Fit Model. Select the column with the response data and click Y. Select the nominal and ordinal columns and click Add. You are now ready to Run the regression model.

Re: Can I add one Ordinal and one Nominal (Columns) to a Multiple Linear Regression model?

mlo1 — Mon, 11 Jan 2021 19:07:40 GMT

Did you check out the BigClass example file. There is a "Fit Model" with different data Modeling type.

Re: Can I add one Ordinal and one Nominal (Columns) to a Multiple Linear Regression model?

Quan711 — Thu, 06 Feb 2025 03:49:47 GMT

May I ask a follow-up question? In this case, how should one interpret the parameter estimates? According to https://www.jmp.com/support/help/en/18.1/index.shtml#page/jmp/ordinal-factors.shtml#, the intercept when fitting with ordinal factor is the first level of the ordinal effect, while based on https://www.jmp.com/support/help/en/18.1/index.shtml#page/jmp/nominal-factors.shtml, the intercept when fitting with nominal factor is the average of all levels of the nominal effect. Say if we have both nominal and ordinal factor in one model, how should we interpret the intercept and each estimate?

Thanks!