It seems like the problem can be solved using calculated field, but you will be doing an aggregated calculation.
say, you create a formula CalcX, with the formula SUM(Weight*ResponseCount)/Sum(ResponseCount).
then the visualization is created by putting the Question on Row and CalcX on Marks/Label
Hope i understand you correctly.
1 of 1 people found this helpful
I hope I'm not pointing out something you already know, but one thing that can be done is to highlight all the rows of the Response Numbers except for the top one, then right click and select 'Hide'. Then, right click on the Response Numbers pill and un-check 'Show Header.' Hope this helps!
The problem is that the weight is response dimension itself. Response * Count(Response) won't work, because of the "can't mix agg/non-agg calculations" error.
That's a neat trick, didn't know that, thanks! It's great for presentation, but would give problems when you have a hierarchy in place.
I did figure out a way; I used a set that contained all values of response, constructed the table. The set, unlike the response dimension, could be hidden. For the agg/non agg mix calculation, I used Min(response) * count(response). Since it was broken down by response itself, min(response) would always be equal to the response value.
To eliminate the duplicate average values, a simple calculated field that filtered out all but the max(response) within the pane. It still takes a little adjustments to properly visualize different charts on the fly (since response dimension can't be eliminated from the charting) but it's doable.
Thanks to the both of you for chiming in!