Creating Demographic Breakdown Chart

[TKendrick20]

Attached is a picture which I will reference.

I want to be able to select a Demographic category in the first listbox and have it show up on the x-axis, and then be able to select a second Demographic category which will stack the bar chart based on the categories in that selection. Is there some way to do this? I assume it would involve the use of variables, but I am not well versed in that area.

[MK_QSL]

PCreate a bar chart

Calculated Dimension

Aggr(Class(Count(Response_Text),5),Demograpic)

Expression

Count(Response_Text)

UPDATE:
Please check enclosed file for better understanding of Class function.

[TKendrick20]

Not really what I need...

[#4]

Post some sample data; it shouldn't be too hard.

[TKendrick20]

Data is stored like so:

[#6]

Take a look at this and see if it is what you had in mind. I wasn't clear on what you meant with the second list box though.

[rwunderlich]

I modified Rebecca's example a bit. I had to load the data using GENERIC to get it to work, making the Demographic categories separate columns, It was the only way I could link the two dimensions together -- to have them link by RespondentID.

-Rob

[TKendrick20]

Thanks! That's just the functionality I was looking for. However, would storing the data in these separate tables for each demographic potentially eat up a lot of memory if I have a few thousand respondents?

[rwunderlich · Accepted solution]

Storing the demographics in separate tables vs a single table would not be any significant difference in memory. And you could support many millions of respondents easily as you don't have many Response_Text values and QV only stores unique values.

However, just for neatness of the model, you may want to join the demographics into a single table after the Generic Load. I've added some code to this updated attached example showing how to do that.

-Rob

http://masterssummit.com

http://robwunderlich.com

[TKendrick20]

Thank you very much. This has been extremely helpful.