Share your ideas for the JMP Scripting Unsession at Discovery Summit by September 17th. We hope to see you there!
Choose Language Hide Translation Bar

Show reference value of Nominal/Ordinal Columns in a report

Currentaly (in JMP 14.3), on reports of platforms like Fit Model, all the levels (unique values) of a Column is shown in the Parameter Estimates (and also in Indicator Parameterization Estimates) table, but the reference value does NOT appear on this report, which is confusing. Give an option to show the reference value in those tables.

Tracking Number:

Defect ID:

3 Comments
Community Manager

Hi @nao,

When parameterizing a model with categorical factors, the last parameter estimate is not independent (k levels for a factor have k-1 independent tests associated with them) and the default for most software is to report only the independent, truly parameterized terms of the model. However, there are times when it's useful to see that final, non-independent parameter. In JMP you do this by selecting "Expanded Estimates" from the Red Triangle > Estimates in Fit Model.

expanded estimates.png

Hope this helps,

@julian 

Level IV
@julian, thank you for your comment. I tried "Expanded Estimates" and found that it (1) does NOT work in combination with "Indicator Function Parameterization" and only shows JMP's rather unique coding scheme (https://www.jmp.com/support/help/14-2/nominal-factors.shtml), and (2) does NOT even exist in (Nominal/Ordinal) Logistic Fit report. I wish JMP will provide a similar table for "Indicator Function Parameterization" in (Nominal/Ordinal) Logistic Fit report. It will be even more handy if this can be called from a red triangle menu attached to the outline box of the (two kinds of) parameter estimate tables, or right-click menu of those tables.
Community Manager

Hi @nao,

You're right that Expanded Estimates will show the effect coded estimates, which are fairly common, but if that's not what you're after I can understand the frustration. The idea of having the option right on the parameter estimates table is a great one! That would be a very intuitive place to look for options relating to the parameterization of the model.  As for the logistic fit, you're right that there does not appear to be an option to show the expanded estimates in the same way.