How can use Domos filter to remove response rates less than 15%?

Matt_D

Hello,

I have a data set containing all response ever recorded for several different surveys.

I would like to build a dashboard that allows the user to select a question and see the response rates for that question (accomplished this by creating a variable and using case statements).

The backend > grouped rows by survey to get a count of participants per survey > divided number of responses (case statement variable) by total count of participants per survey. (Tested with only 3 survey questions shown below).

The result in a table format is that I can see the survey and response rate for what ever question selected.

Now, I'd like to filter out surveys where the response rate for that question was less than 15% and get the final average for that.

When in table format, this isn't an issue and surveys with less than 15% response rate for that question are removed.

However, for the gauge, the filter doesn't seem to apply well to the data.

I'm wondering why this is happening and why the max is only .05?

I think it has something to do with the math on the backend. The average (after all response rates <.15 should be around 37.8%….

The gauge average without filtering is correct for the selected question (0.14%).

RobSomers

@Matt_DAlto It works in your table because there you are grouping by company. In your gauge card, you're not grouping by anything, so you've only got one number or "row", which is 0.14. It's just the overall percentage. So if you filter to >= 0.15 then nothing will pop up. I think you would need to create an ETL and do group by company to get totals and then get a response rate for each company. Then you could use the resulting dataset for the card.

RobSomers

@Matt_DAlto It works in your table because there you are grouping by company. In your gauge card, you're not grouping by anything, so you've only got one number or "row", which is 0.14. It's just the overall percentage. So if you filter to >= 0.15 then nothing will pop up. I think you would need to create an ETL and do group by company to get totals and then get a response rate for each company. Then you could use the resulting dataset for the card.

jessdoe

It looks like your average response rate is way less than .37 as most of the companies have null response to company strength. You can try implementing an ifnull(response rate, 1) to see if that gets you closer to where you want to be in your gauge card.