Forum Discussion
Satisfaction Rate Calculation In Power BI (Likert Scale)
I am fairly new BI users. I am working on students evaluation survay.
I have all this built in Tableau and now need to re-built my dashboards in Power BI. Is this even doable in power BI or are there different or less steps to take to get this result:
My Tableau Calcualted Field Steps in order to get me the Satisfaction Rate fields, in another word I had to take these steps to get the right count for my numerator and right count for denominator; therefore, as a result I could divide them and get the Satisfaction Rating %:
1. Combined ResponseID Question ID = [Response Unique ID]+"-"+str([Question ID])
2. Promoter = IF [Response] = "Strongly Agree" OR [Response] = "Somewhat Agree"
THEN [Combined Response Question ID]
ELSEIF [Response] ='Not Applicable' THEN [Combined Response Question ID]
END
3. Numerator= COUNTD ( IF [Category]="CA" THEN [Promoter ] END)
4. Denominator= COUNTD([Combined ResponseID Question ID])
5. Satisfaction Rate % = Numerator / Denominator
How Would I acheive this in Power BI, I have followed and tried multiple ways and calculations but nothing gives me the desired results. I have a sample of my Power BI file attached so if anyone want to look into this, you can see all my Power BI (DAX) calculations. I even re-did my table connections but nothings seems to give me the result I am looking for.
I really would appreciate it if anyone want to tackle this and help me out please.
https://drive.google.com/file/d/18udf1Ygl-PDC9Jba5-f8ymvzFcdW-w5S/view?usp=drive_link
Hi,
Finally, I'm getting the same values as you.
The final result is this:
Measures:
As you can see in the matrix for Arkansas you have 34 rows but you only consider 24. With your example I found that you only count if the column "Satisfied Socores" is different to "" (empty), therefore I used this criteria to do my denominator. In your first mensage you have a different criteria using "Category"="CA" but I didn't find any row with that information.
In my solution I'm using your first pbix and only makes some changes on the relations. The final model is this:I hope this could solve your problem.
13 Replies
Hi, I'm trying to help you but I can't understand what you really want to do in your calculations.
Please give me same pictures with the correct results and I will try to do it in power bi._AAndrade As I mentioned I have built these in Tableau so I have those screenshots and correct results. As it shows in the image. Mercy Hospital has 77% Satisfaction Rate and then when I click on Mercy Hosptial I will see the satisfaction rate for it's units as shown in the image students rated 5 different unit. So I need Satisfaction Rate as an overall for Hosptial(Agency) then by selecting I want to see it breaking down so I can see satisfaction rate for it's units. thank you for taking the time and helping, I have included my Power BI project link in Google drive and you can also see all my table connections as well as all the DAX formula's I have tried and still no luck to acheive this result.
1st Image shows Satisfaction Rate for Agency:
2nd Image shows once I click on Mercy Hospital is showin it's units satisfaction rate:
What I have in Power BI as a result which is not correct outcome
I don't have data with [Category]="CA". Let's concentrate in one Agency firt.
For Mercy Hospital we have this responses:First measure is distinct count of all responses and "New Promoter" it's my new measure than only counts if the responses is "Strongly Agree", "Somewhat Agree", or "Not Applicable".
How I should do to calculate the Satisfaction Rate % ?To calculate the Satisfaction Rate % we need to treat the new promoter as the numerator and divide it by SUM # of RESP (since # of RESP is the count of the respondents) Does that make sense? I do have # of RESP as a measurment and as a column. I believe my issue is in the numerator (Promoter) formula.
I changed my model so I will update the file on google doc as well this is the new model:
Hi,
Please check if any satisfaction rate is what you're looking for: