Test Probabilities in Distribution Platform...version 15 change?

[#1]

When testing probability distributions in the distribution platform (to do a one-way chi-square test), you used to be able to enter any number in the hypothesized probability column and JMP would recalculate to adjust the sum to equal one.  This was very helpful since if, for example, I had a ratio set of 9:3:3:1, I could enter 9, 3, 3, 1 and JMP would do the correct math.  Now, anytime I enter a number that's not between 0 and 1, I get an error message that these entries have to be between 0 and 1 even though it still suggests it can "rescale" the input probabilities.  Am I doing something wrong or is this a "feature" and not a "bug"?

Given that I use this to teach my students about chi-square testing, it is really not helpful to tell them that now they have to do the math.

[#2]

This change was made as a fix to another issue with Test Probabilities.  We will investigate other fixes and see if we can add this functionality back.

[#3 · Accepted solution]

In the meantime, the students won't really need to do that much math. In your example of 9:3:3:1, the total is 16. So in the text entry box for the probability type "=9/16", "=3/16", "=3/16", and "=1/16" (without the double-quotes). JMP will do the division for you.

[#4]

Thank you so much for this solution. It will work for my students...and me.

[#5]

Thanks for the quick response. It will be helpful to have this functionality back, but Dan's solution below will work well until then.

[#6 · Accepted solution]

Greetings - i can confirm that as of JMP 16 (which is currently available), the original functionality has been added back in. You are now able to again enter probabilities greater than 1 and JMP will rescale. 

[#7]

Excellent!  Thank you!  Both my students and I thank you.

[#8]

I am preparing my lessons on the chi-square test for this semester using JMP 16, and again, I can't tell you how relieved I am that the probabilities can be automatically rescaled with using the one-way chi-square test in the distribution platform.  Thank you again so much, and thank whatever team members or member did the actual recode to "correct this feature" from JMP 15.