Survey Analysis - Top/Bottom Scoring Questions

[AJV]

Hello:

I am looking for some suggestions as to how to produce a table that shows the analysis of survey data.

Of a survey containing 50-60 survey questions, we need to draft a table that shows what the highest 3 scoring questions and lowest 3 scoring questions are for each population (ie. business unit, demographic, etc.). I'd envision this table to list each population in a row with columns then displaying the top and bottom three questions.

I'm wondering if anyone has any pointers or suggestions, or knows of an online resource that they could point me toward.

Thank you!

[Ashish_Mathur]

Hi,

Share some data and show the expected result.

[amitchandak]

@AJV , for this your question and score both need to be in rows.  The based on average Score and Rank you can find top and bottom 3

Score = average(Table[question Score])

High Rank = RANKX(all(Table[Question]),[Score])

Low Rank = RANKX(all(Table[Question]),[Score],,asc, dense)

Top 3 = CALCULATE([Sales],TOPN(3,all(Table[Question]),[Score ],DESC),VALUES(Table[Question]))

bottom 3 = CALCULATE([Sales],TOPN(3,all(Table[Question]),[Score ],ASC),VALUES(Table[Question]))

[Greg_Deckler]

@AJV So in general you are going to want everything to be unpivoted. So like:

ResponseID,QuestionID,QuestionText,Score

1,Question1,This is a question,10

1,Question2,This is another question,3

2,Question1,This is a question,10

2,Question2,This is another question,3

3,Question1,This is a question,10

3,Question2,This is another question,3

You do NOT want:

ResponseID,Question1Text,Question1Score,QuestionIDText,Question2Score,Question3Text,Question3Score...

Not sure how much more specific I can be given the information. @ me in replies.