Set analysis for creating groups in TTEST_SIG function

[#1]

Hi,

I want to create two groups from data given below to be used in the function :

TTest_sig ( [set_expression][ total [<fld {,fld}>]] group, value)

GEOGRAPHY	SALES
A	50
A	43
A	23
A	65
B	19
B	53
B	11
B	62
C	11
C	89

Currently I am using:

TTest_sig(If(GEOGRAPHY='A' or GEOGRAPHY='B',GEOGRAPHY),SALES)

But the problem is I have a GEOGRAPHY list box and whenever the user selects a geography from this list box the data gets filtered for that geography and Ttest function returns a null. How can I nullify this selection in Ttest function?

Thanks & Regards,

Amit

[MK_QSL]

Can you provide little more information about your requirements?

[#3]

Hi Manish,

Thanks for quick reply.I need to find the p-value using ttest between Geography A and B. So in my Ttest function these two geographies will be static i.e. the Geography list box should not impact this. Whereas I have few other list boxes and the selections made in these list boxes will filter the data. For e.g. I may have a Products listbox. So in case a product is selected the p value will be for Geographies A and B for selected product.