(originally posted to StatsiQ but I think this is the better location.)
Hello all, I'm new to Qualtrics. I am looking for advice about how to use bucketing or other related data grouping methods so that I can group responses to narrow down my analysis. Let me give a simple example. My survey asks respondents to identify their gender by checking all that apply among the following options: woman, man, transgender, non-binary, other, prefer not to say. I would like to group these into Cis Woman, Cis Man, and Other (due to very few responses from non-cisgendered people). I am concerned that bucketing will not work because the bucketed variable will only assign them to their first selected answer. This will not be a problem for someone who checked "man": this response will properly be bucketed into "man." But for a trans man who checked "man" AND "transgender", will there be a problem? Will he also get grouped with "man", even if he should be grouped into "other"?
Do experienced users have suggestions for other ways to go about grouping responses? I have several other responses that might be grouped in a similar way, and manually tagging them all seems inefficient and clunky. Thank you for your help.
If someone can have multiple categories (as in your example), and the data is already collected, then I would suggest you use filtering & tagging to apply the tags you want to use. Bucketing won't be a fit if people could fall into more than one bucket, or if you need a combination of factors to apply to put people into the right categories.
If you have NOT yet collected the data, then I would probably use a combination of branch logic and embedded data to set their categories through the survey flow. (So if in Q1 you ask them to self identify, after that block of questions you could use their responses to "flow" them into setting a "Gender Identity" embedded data field based on their answer or combination of answers).
Thank you, I appreciate your detailed response! We have collected data from our first round of respondents, so we will have to filter and tag. But for future responses, it definitely sounds worth it to go in and tweak the embedded data and branch logic firs.t
Leave a Reply
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.