cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
JMP is taking Discovery online, April 16 and 18. Register today and join us for interactive sessions featuring popular presentation topics, networking, and discussions with the experts.
Choose Language Hide Translation Bar

Viewing "Weighted N"

Hello,

First of all, I should say that I am an Engineer with a very basic knowledge in statistics (perhaps thats why I am using JMP :) 

I have a data table with 90,000 rows and 12 columns. One of the columns has a pre-selected role as "Weight". The dataset I have has been sampled from a series of events with each event having its own "Weight". In our literature, it is common to publish Tables with "Unweighted N" and "Weighted N". Unweghted N is the actual sample size and Weighted N being always higher than the Unweighted N (please see attached file). In all my analyses, even if I include the Weight column, I could not calculate the "Weighted N". That is, the total number of rows always adds up to the actual number of rows (90,000). However, the Tabulate functions seems to include the Weights, but again would not display the "Weighted N". 

 

Appreciate any help with this.

 

Thanks,

Mike

3 REPLIES 3
txnelson
Super User

Re: Viewing "Weighted N"

I believe you want to use the Freq capability in the JMP Procedures.....below is an example with a dummy data table, the Weight column is specified in the Freq selection areaweighted.PNG

Jim

Re: Viewing "Weighted N"

Jim,

Ahh .. got it. Perhaps I got confused between Weight and Freq procedures in JMP. I did additional reading on the SAS documentation website where they distinguish between both the procedures (http://support.sas.com/kb/22/600.html). Thank you.

 

One more question: is there a way we could "Tabulate" the Weighted and Unweighted N on the same table using the Tabulate procedure?

Best regards,

Mike

ron_horne
Super User (Alumni)

Re: Viewing "Weighted N"

I would say that you would need to make two different tables, one weighted and the other not weighted and just present them together, "glueing" them side by side. If you inted to present this in an external program all together it would be very simple.

 

I would like to take this opportunity to ask the members of the group, in what situations would we get different numbers in a tabulate platform between "N" and "sum wgt" and perhaps it is time for JMP to have "N rows" in this platform and in any other case in which analysis is weighted or using a frequency.