Reputation: 55
I want to run a study that recruits Mechanical Turk workers to take a survey on Qualtrics. But the survey itself is pretty long, and I want to make sure I get a good number of respondents from hard-to-find groups on MTurk (such as Latino men) without having to pay for thousands and thousands of respondents. So my idea is this: I'll first recruit many people to take a short demographic survey where they answer a few questions about themselves. Then, I'll load all those survey responses into R and, using MTurkR, I'll select a subset of those people to recontact (they'll be my sample for the real survey) and assign them some qualification, then send them a message asking them to sign up for my second HIT.
Okay, so then hopefully a bunch of people from my selected sample will go on to take my second (longer) survey. The challenge is that I'd like some of the questions in this second survey to include demographic data from the first survey. So if you said in the first survey that your religion is Catholic, I'd like there to be a question on the second survey that goes "As a Catholic...". But I don't see an obvious way to pull respondents' answers from the first survey into the second survey. Ideally, I'd have them put in their MTurk worker ID in both surveys, and that ID would serve as a link between surveys. But practically, how do I do this?
Upvotes: 2
Views: 520
Reputation: 5029
After you select the subset of people you want, save the recontact data to a CSV then upload it into a Qualtrics panel.
At the beginning of your Qualtrics survey flow, define the recontact fields as embedded variables that will be populated from the panel. Then you can use them in your survey as piped text (e.g. ${e://Field/religion}) or in logic (if embedded variable religion = Catholic).
In Qualtrics, create an email invite, then invite the people in your panel to take the survey. They will receive an email invite with a link that ties them to their panel data.
Upvotes: 0