Logit Regression Help

[#1]

I am trying to run a Logit Regression and it has to include a column called "validation" which is just one of two words (half the data points are word type, the other half are another) so it comes up on JMP as nominal. What do I do? Thanks. The data is attached, choice is the dependent and all of ther other variables should be the predictors (including validation column).  

[#2]

Your request is unclear:

• Do you want :validation to be a predictor, or
• Do you want to use the "Train" rows to create the model and the "Holdout" rows to validate the quality of the model?

[#3]

Run a Logit Regression with Choice as Dependent variable and all other variables as Predictors. Note that you have to either (1) Use Validation Column in running the model, or (2) if you don’t have the latest JMP, “hide & exclude” the Holdout sample before running your model.

[#4]

If you have JMP 14, here is the dialog. Note Choice is nominal. If not, follow the instructions, to select rows with Holdout and select Rows > Hide and Exclude. Then use Fit Model and it will look the same except for the Validation field.

[#5]

If I have JMP 14, do I put validation under P_Art as an x, right?

[#6]

No as Validation. I think a Validation is a JMP PRO feature.  So if  the Validation button does not appear, before, running the Logistic regression from the main menu select Rows >  Row Selection > Select Where ( same as CTRL + Shift + W ) and a dialog will appear. Select Validation and type Holdout, as shown below, then OK. Then select Rows, Hiide and Exclude. Then run your logistic regression. After Running if tiy save the probability model, the predictions for Holdout will also be computed.

[#7]

Run a Logit Regression with Choice as Dependent variable and all other variables as Predictors. Note that you have to either (1) Use Validation Column in running the model, or (2) if you don’t have the latest JMP, “hide & exclude” the Holdout sample before running your model.