Aggregate StatisticsI have a rookie question regarding aggregate statistics in the Tabulate window.
I am tabulating the mean, min, max and range of two columns of continuous numerical data against a categorical variable (run time).
I click the Aggregate Statistics box to add a final "All" row to the data.
According to JMP help, the All row should represent the mean of each column in the table?
If I look at the last column, the All value is 5.64e-02 but if I use Excel to convert that column to an average the value is equal to 3.44e-02.
I assume that the tabulate function includes any outliers in the column data.
Am I missing something obvious?
Thanks in advance--jc
Basic Data Analysis and ModelingRe: Aggregate StatisticsWhere did you find that All represents Mean? According to this https://www.jmp.com/support/help/en/18.2/#page/jmp/add-statistics.shtml#
All
Aggregates summary information for categories of a grouping column.
So it can could be much more than just Mean (in your case Range most likely for last column)
Re: Aggregate StatisticsHi Jarmo;
Thanks for the quick reply. I was reading the examples described in the JMP community materials here.
https://community.jmp.com/t5/Learn-JMP-Events/Creating-Summary-Tables-Using-Tabulate/ev-p/810011
In their example, they report the All value as the mean of the column so i assumed it was always an aggregate mean.
Looking specifically at my data, the last column to the right is a Range value by Run time. The All value displayed in the last colummn cannot be an aggregate Range either. If I calculate the range in Excel, the range is equal to 1.05e-02 but the value displayed by All in JMP is 5.64e-02.
Re: Aggregate StatisticsIt is not the range of ranges, it is the range of the column in your data table. You can verify this for example with a Summary table

or with distribution by adding more summary statistics


Re: Aggregate StatisticsJMP’s “All” row in Aggregate Statistics often shows overall values, not group averages. For true group-based averages, use the Summary platform or calculate manually for accuracy.