I'm attempting to conduct a survey using proportionate stratified random sampling and I'm trying to figure out how to keep track of which group my participants are in without collecting identifiable information. I have four groups (City, Suburb, Rural, and Town) that I am trying to gain a representative sample from. However, the participants wouldn't already know which group they are in and I don't want them to have to report their school district (which is how I am determining their group).
So, my ideas are to create a question where participants would have to enter a code that I include in the recruitment email that would correspond to the given groups (i.e. 4001 = City, 4002 = Suburb, etc.). The code would be the same for all the participants that meet the criteria for City and so on. I could then use this code in my analysis to compare groups.
I've also thought about copying the survey to create a unique URL for each group and then applying an identifier in a new variable once I have concluded data collection. Then I would just aggregate all 4 datasets into a master dataset.
I've also looked into sending the recruitment emails via Qualtrics Distribution so I can upload participant contact information into 4 separate contact lists representing each group. However, I haven't been able to figure out how to track which contact list the submission came from.
I'm sure there are better ways of doing this and I would love to know your thoughts! How should I track group information without making the participants either look it up themselves or have to share identifiable information in the survey? Any thoughts are appreciated!