Analysing Multiple Response Survey Data with Alteryx and Tableau – part 2

To continue from part one of this series we are going to explore a way of analysing combinations of responses for multiple response survey questions.

For example, we might want to know what combinations of responses are most selected by respondents. In this example we are asking respondents to pick three options.

The survey question:
I have changed the question from part one slightly. We can also apply this type of analysis to an “all that applies” question, however putting together all the possible responses takes MUCH longer (especially if you have a lot of possible responses!).

Firstly we need to list all the possible combinations of responses. The question specifies the respondent to pick exactly three options. The respondent cannot proceed to the next question unless exactly three have been selected, so we don’t have to include single or two option response. Therefore the possible combinations are:

Pie, Bar, Line
Pie, Line, Bubble
Bar, Line, Bubble
Pie, Bar, Bubble

Next we need to calculate how many times our respondents have selected these combinations in their answers.

To do this I have gone back to Alteryx.

Re-using the module from part one, I have added four Formula tools to create four new fields: “pie, bar, line”; “pie, line, bubble”; “bar, line, bubble” and “pie, bar, bubble”. These match our four possible combinations. (note: you can use just one formula tool if you wish, and create four fields in a single tool; however I’ve added four so it’s easier to see what’s going on).

The formula for “pie, bar, line” is as follows:

IF [Q1 Pie] == “1”
AND [Q1 Bar] ==”1″
AND [Q1 Line] ==”1″
AND [Q1 Bubble] ==”0″
THEN 1 ELSE Null() ENDIF

The remaining calculations follow the same logic. The fields are then transformed in the same way we transformed the data in part one. The result looks like this:

Now if we connect Tableau up to our .tde output we can visualise which were the most popular combinations:

Again I’ve used the same method from part one for simplicity! But feel free to experiment with your visualisation.  If you come up with anything interesting please get in touch!

You can also download the data, the Alteryx module and the Tableau workbook.

London, UK