Coding Survey Data Questions for "Theme" Analysis

[JMGData]

Hi,

I am in the process of trying to present survey data for a school district but have run into an issue. The raw data looks something like the table below with Respondent Answers (Likert Scale) occupying each row and correspond to the columns that the questions are in. The issue is that I need to aggregate the responses (50% agree, 25% disagree, 25% strongly disagree) based on coded themes. For example Question 1 and Question 4 are about Instruction, Question 2 and Question 5 are about Behavior, Question3 and Question 6 are about Grading. What's the best way of preparing this data so that I can show results based on these themes and also use the demographics as a filter later on? Is there a way to "Theme" this data?

Thanks in advance for any help you can provide!

- J.

Respondent ID	Demographics	Question 1	Question 2	Question 3	Question 4	Question 5	Question 6
1	A	Agree	Disagree	Strongly Disagree	I don't know	Disagree	Agree
2	B	Disagree	Disagree	Agree	Agree	Strongly Agree	Agree
3	A	Strongly Disagree	Disagree	Disagree	Agree	Strongly Agree	Disagree
4	B	Strongle Agree	Agree	Agree	Agree	Strongly Disagree	Disagree

[amitchandak · Accepted solution]

@JMGData , select the first two columns (columns other then questions) and unpivot other columns, that should help

https://radacad.com/pivot-and-unpivot-with-power-bi