Hi there,
I have survey data where each respondents answers correspond to a row in my data. The columns are the questions.
A question that can have only one answer has each of the possible answers coded as a number.
For example:
1 - stronly agree
2 - agree
3 - disagree
4 - strongly disagree
k.A. - respondent didn't answer the question
blank - respondent wasn't asked that specific question
I can create a pivot table from two questions where only one answer is allowed and count the responses. So far so good.
My trouble is with multiple choice questions.
In my research I found there are two ways to enter multiple choice questions. Either you create a code for each possible combination or you set up a column for each choice and mark the ones chosen. I went with the second choice.
For example:
Q2: Which fruits did you eat today?
Answers:
Q2-01 - apples
Q2-02 - oranges
Q2-03 - bananas
Q2-04 - other
So Q2-01 through Q2-04 would be the column headers. The values as of now are "x" when someone checked any of the first 3 options and the value they entered under "other" - for example "strawberries". It is possible to change the values to numeric ones if that makes things easier (1-yes, 2-no).
I'm not very experienced with pivot tables and am having trouble when trying to combine a multiple choice question with a "normal" question. Essentially I only want to use the cells where something is filled out in the table and disregard the blanks. I'm not sure if pivot table is even the best way to go here to determine if there is a connection between the answers to one question and the answers to another.
If anyone has some suggestions on reading material or tutorials I would appreciate it very much!
Thanks a bunch,
Denise
I have survey data where each respondents answers correspond to a row in my data. The columns are the questions.
A question that can have only one answer has each of the possible answers coded as a number.
For example:
1 - stronly agree
2 - agree
3 - disagree
4 - strongly disagree
k.A. - respondent didn't answer the question
blank - respondent wasn't asked that specific question
I can create a pivot table from two questions where only one answer is allowed and count the responses. So far so good.
My trouble is with multiple choice questions.
In my research I found there are two ways to enter multiple choice questions. Either you create a code for each possible combination or you set up a column for each choice and mark the ones chosen. I went with the second choice.
For example:
Q2: Which fruits did you eat today?
Answers:
Q2-01 - apples
Q2-02 - oranges
Q2-03 - bananas
Q2-04 - other
So Q2-01 through Q2-04 would be the column headers. The values as of now are "x" when someone checked any of the first 3 options and the value they entered under "other" - for example "strawberries". It is possible to change the values to numeric ones if that makes things easier (1-yes, 2-no).
I'm not very experienced with pivot tables and am having trouble when trying to combine a multiple choice question with a "normal" question. Essentially I only want to use the cells where something is filled out in the table and disregard the blanks. I'm not sure if pivot table is even the best way to go here to determine if there is a connection between the answers to one question and the answers to another.
If anyone has some suggestions on reading material or tutorials I would appreciate it very much!
Thanks a bunch,
Denise
