1) What determines the number of decimal places given in a report for summary statistics?Re: 1) What determines the number of decimal places given in a report for summary statistics?Do you have an example of what you are seeing that is counterintuitive? I am still pretty sure I am misunderstanding your question.
Usually (but not always) JMP uses a number format called "Best". As I understand it, it basically says "I have room to print N digits". If it has to use all N of those digits on the left side of the decimal point, then it does and it puts no decimal part on the right. If it has some digits left over, then it puts some decimal on the right.
So "Best" formatting with 9 digits -- which I think is the default in Distribution, as much as I know now that's not the platform you're talking about -- can show the following numbers:
0 (1 digits)
0.5 (2 digits)
5 (1 digit)
5.125 (4 digits)
5.12345678 (9 digits)
51.2345678 (9 digits)
51234.5678 (9 digits)
But it won't show 51234.56789 (because there's 10 digits in that) - it would instead round and display 51234.5679.
Re: 1) What determines the number of decimal places given in a report for summary statistics?Actually, I'm not talking about the distribution platform. My comment was a general comment regarding the reason for the number of decimal places used in different statistics.Re: 1) What determines the number of decimal places given in a report for summary statistics?Let's refine the question a bit more since it seems that I've not understood it. Are you getting different numbers of decimal places in the same analysis window, or getting different numbers of decimal places from one column of data to another or one file to another? Is this just in Distribution, or somewhere else in JMP?
Re: 1) What determines the number of decimal places given in a report for summary statistics?Actually, I was wondering why some statistics appear with a single decimal place while others can have as many as 7 decimal places. For example, I've since 7 decimal places for a nominal variable with 5 levels (1-5). That seems excessive and implies a level of percision that seems unlikely. So I was curious if JMP has any type of 'standard' they use to determine the number of decimal places for each statistic.
Re: 1) What determines the number of decimal places given in a report for summary statistics?I assume you're talking about the Distribution platform. I think by default JMP uses "Best" formatting with a field width of 9. This can be changed by double clicking on the column of numbers in the Summary Statistics panel. You could switch it from "Best" to "Fixed Dec", or just change the "Best" field width.
