In the case of data blending, you have primary data source and secondary data source.
Typically should not use secondary data source as dimension because of these "*".
I change d the axis to primary data source and the number has been changed a little, but I believe this one is correct.
Thank you for your response Shinichiro,
I had tried using the Primary Data Source as well but had gotten the same answer as yours. Is there a reason why these numbers shouldn't match? My assumption was if you use a dimension, such as a date field, switching it to Discrete would solve the issue as far as stacking the data which inflated the numbers. When I used the secondary data source I at least had gotten the same numbers as the crosstab version, with the exception of ' * '.
I was also working on this yesterday. At first, I did not realized that SchAdherence have a [dated test] relating to [datea test] of GregInbResults, so I didn't post my solution since Shinichiro's answer was more simple that what I came with. But what I found is that there is a mismatch in GregInbResults between [The Week] and WEEK(datea test).
You can see this on Sheet11 in the attached.
My solution attempt was to make a correction by computing datea test(Adjusted) returning the Max WEEK(datea test) by [The Week].
This way you have a 1 to 1 relation. It get rid of the * on the axis but the average change a little bit (see sheet 12).
Also on the sheet Batch Code Adherence, you filter by [The Date] starting on Jan 1st, but if you look on Sheet 11, week 2 starts on Jan 4th, before that you are on week 53. So you don't filter the exact same range of dates on both sheet.
I was filtering out January 1-3 being that was a partial week and I was looking for the average over the weeks of 2016 so far so that is the reason for the removal of those dates.
Makes sense now, thanks so much!
Thank you for the following up Michel.