Complex Survey Construction Help Request! Bear with me here...

Question

Hi all,I am working on my dissertation in clinical psychology. Since COVID-19 hit, I have been forced to move my project completely online. I am running into several snags trying to set up an automated set of surveys for a treatment study. I have the design envisioned so perhaps I will start with that, and then explain what I have tried and where I am running into issues. Essentially, it is an 8-week treatment study. Participants will access the study by going directly to a Qualtrics link to the consent form, or emailing me for the Qualtrics link for the consent form. After participants consent to the study, they will complete a set of pre-treatment measures. They will then be randomized into one of two groups. Each day for 8 weeks, participants will complete a condition-specific exercise and answer a couple questions. After 8 weeks, participants will complete an Post-Treatment survey.I envision enrollment to be rolling, so if someone were to start today, they would finish in exactly 8 weeks. If my second participants started a week from today, they would finish in 9 weeks (1 week from now + 8 weeks of study).It is possible for this to be completed automated through Qualtrics? In my mind I would have three surveys that I would later merge based on ID numbers:First survey: They would sign Consent form, complete pre-measures, get a randomized ID number (must not be repeatable), and provide an email address to receive the other surveys. I used the random number generator in Qualtrics, but understand from other users that it does not offer unique numbers. I am hoping I can keep the numbers simple to reduce memory and entering errors for participants. The second survey would be received daily in the AM in their email. I am thinking that I would have a random number selected between 1 and 200 (these are arbitrary) in the first survey to function as their ID number so I could merge all the data later. If their number is 1-100, they complete Group A stuff. If their number is 101-200 they do Group B stuff. This would satisfy them being randomized to one of the two treatment groups. I had set this up successfully with the Display Logic tool. Additionally, they would receive this email daily for 8 weeks from the time they first consented. So it is more of a set number of sends (8 weeks x 7 days = 56 sends) rather than a set start and end date, if that makes sense. I am able to use the Action command in the first survey above to send the second survey based on the email address they provide in the first survey, but can only do it once. I am having difficulties setting up the second survey to send 56 times (i.e., for 8 weeks). The third survey will basically only consist of the post assessment surveys. It should be sent exactly 8 weeks after the first day (so on the 57th day).So basically, I am struggling with how to get a random number generator to provide simple, unique IDs in the first survey (I was reading about using a Web Service ???). For the second Survey, I need it sent by email daily for 56 days in a row. And the final third survey sent on the 57th day. This is to ensure that people can begin the study as they sign up, instead of waiting for sign ups and doing a cohort start (this is to reduce attrition). Thanks for any help! I am hoping it can be automated through Qualtrics, but my skills are too basic for that at this time. Thank you!!!Take care,Eric

MatthewM · Answer

Hi ETifft,I don't think it's possible for a single beginning user to completely automate this process within Qualtrics. Especially with the rolling start dates for participants, it would likely require outside systems and programming expertise.Assuming that external support is not an option for a dissertation, you'll have to manage some degree of the project manually, mainly survey 2.My suggestions are:For the unique random ID issue, either use a 10 to 12 digit number as Qualtrics suggests (and verify they are unique on a regular basis) or manually assign a unique ID upon enrollment (though the latter may inhibit your ability to use Actions to send out the first email for survey 2 and the email for survey 3 56 days later).Use Actions to create a living Contact list that grows with each new enrollment.Use an Authenticator, linked to your living Contact list, for respondents to access survey 2 after the first time. Email address is the easiest unique identifier to use, but if IRB does not allow that, then use the ID number assigned in survey 1. I know you want to send out emails with unique links to participants every day but that's going to become very unmanageable in the Qualtrics mailer to do that individually for 56 days, even with only 100 to 200 participants. You might consider sending generic "reminder" emails to all participants, with the general link to survey 2, through your own email instead.Hope this helps, and good luck!

Sign up

Login to the Community

Scanning file for viruses.

This file cannot be downloaded