How do I reverse the order of contrasts in the Stepwise platform?

[#1]

I have a variety of categorical factors that I am trying to fit in the Stepwise platform. For example, "Does the product have a FRAME, LINER, BAFFLE, INSULATION, etc.?"  These are listed in the stepwise menu as Frame{Y-N}, Liner (Y-N), Baffle{N-Y}, Insulation{N-Y}, etc.  Why are some contrasts listed as {Y-N} and others listed as {N-Y}?  This is really slows my analysis down as I build a model because I have to mentally reverse the sign of the coefficient when assessing some factors.  I have found no reason why JMP lists something one way versus the other, nor any way to change the default order of the contrast.  I have already tried changing the Value Order, switching the Design Role to Categorical,  changing the Target Level, and anything else I can think of.  Nothing makes a difference.  Is there a reason JMP uses one order versus another and more importantly, is there any way to change this order?

[#2]

Hi,

It looks like you went through the logical steps, but one thought comes to mind:

Is it possible that you use Labels (e.g., 0 = "N" and 1 = "Y") for some but not all columns? 

Also, you mentioned that you "changed the Design Role to Categorical," suggesting that the original data might be numerical.

Otherwise, I am not sure what else to try.

Best,

TS

[#3]

Thanks for your reply.  The data were born categorical as "Y" or "N" and all columns are consistent with respect to their design role, etc.  It may be that JMP has a peculiar method of assigning states.  So far, nothing I've tried works and I see no pattern as to whether a contrast would be listed as {Y-N} or {N-Y}.

[#4]

This is a pure guess.......maybe the order that the Y's and N's appear in the data table set the values.  Just for grins, try sorting one of the column in question that selects Y-N order and see if the order changes.

[#5]

Merci, Thierry.  Unfortunately, the order doesn't make any difference.  But thank you for your reply.