Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 
Not applicable

T-test in QlikView / QlikSense

Hi!

I would like to perform an independent two-sample T-Test in Qlik.

{\displaystyle t={\frac {{\bar {X}}_{1}-{\bar {X}}_{2}}{s_{p}{\sqrt {2/n}}}}}

{\displaystyle \ s_{p}={\sqrt {\frac {s_{X_{1}}^{2}+s_{X_{2}}^{2}}{2}}}}

(i.e. T equals (the difference between the meas of the samples/arrays) divided by (the squareroot from (the sum of the sample standard deviations divided by two) times the quareroot from (2 divided by the sample size)

I believe it is possible to hardcode this in set analysis, but I would like to try the actual aggregation. I have searched through this community and only find unanswered questions similar to mine since 2012.

Much obliged

//Nathalie

3 Replies
miguelbraga
Partner - Specialist III
Partner - Specialist III

The T expression might be something like this:

=

(

count(X1)/count(Total X1) - count(X2)/count(Total X2)

)

/

(

($vSp)* sqrt( 2 / count([sample size])

)

Where vSp is a variable made by this expression:

=sqrt

     (

     sqr(X1) + sqr(X2)

     /

     2

     )


Best regards,

Data Architect MB

Not applicable
Author

Hi Miguel and thank you.

I may have stated my question badly. Right now I have done exactly what you also proposed in your reply, i.e.done the match by hand in the app, but I would like to better understand how to use the built-in Qlik-functions for T-test and the help.qlik.com was not very informative. Have you used it?

Regards

Nathalie

MarkMV
Contributor III
Contributor III

Have you had any luck with using this functionality in an application? I cant seem to find much information on this and anything I try seems to just load forever without actually rendering anything.