0
$\begingroup$

Wondering if my method here for coding qualitative research data is valid.

My raw data looks like this:

Contains apples Contains bananas Contains pears
Cart 1 x
Cart 2 x
Cart 3 x x
Cart 4 x x

I then added another column for shopping carts with more than 1 type of fruit:

Contains apples Contains bananas Contains pears More than 1
Cart 1 x
Cart 2 x
Cart 3 x x x
Cart 4 x x x

Total of all carts,

Contains apples Contains bananas Contains pears More than 1
Total 2 3 1 2

Expressed as a percentage, do I divide by 8 (the row sum) or by 4 (the number of carts)? I've found that if I divide by 4 to get "X% of carts contained bananas", the row sum is not 100%, due to the 'More than 1' column. Dividing by the row sum gives me percentages which sum to 100, but I'm not totally sure what they mean.

$\endgroup$
3
  • 1
    $\begingroup$ You are discussing how to code variables, but you haven't yet explained why. The answer depends on how you intend to use this coding for analysis and interpretation. Please explain. $\endgroup$
    – whuber
    Commented Aug 9, 2022 at 16:45
  • $\begingroup$ Why do you expect that your row sum should add to 100%? $\endgroup$
    – CFD
    Commented Aug 9, 2022 at 18:06
  • $\begingroup$ I guess because, if I have percentages, I am expecting that the whole is represented in the data somehow. The fact that I can't find a whole when dividing by the number of carts in my example makes me question it. What if the example I gave is only for customers on Monday, and I repeat this process for customers on Tuesday? I want to compare percentages for total of all carts between these days - which is correct? Divide by number of carts, or divide by row sum? $\endgroup$ Commented Aug 9, 2022 at 19:21

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.