Anonymous test retest reliability | XM Community
Solved

Anonymous test retest reliability

  • 28 January 2019
  • 5 replies
  • 163 views
M
Hello,
For my survey I would like participants to fill in the survey twice, with an interval of a week, to determine the questionnaires internal consistency. However, I want their data to be anynomous. Is there some way to ask people their e-mail address in such a way that when they fill it in, it becomes a random code or something? And then when a week passes, a link to the survey is emailed to the participants and when they fill the survey in again, the first response becomes linked to the second response (for analysis).
Thanks in advance.
icon

Best answer by TomG 28 January 2019, 15:58

View original

5 replies

TomG
Rank
Userlevel 7
Badge +27
In order to email them again, you are going to have to save their email address somewhere. You could separate it into a different survey by redirecting at the end of your survey and only sending the ResponseID from the first survey to the survey you redirect to (as a different name like ResponseID1). Then add a contact trigger to the survey you redirect to to save just the email and ResponseID1 in the contact list. That way the answers to the first survey are stored separately from their email address.

For the follow-up survey in one week, turn on anonymize response so the email isn't saved with the survey data and include your ResponseID1 field from the first survey as an embedded data field.
M
How do I send the response ID's from one survey to another? And how do I include the response ID's from the first survey as an embedded data field?
TomG
Rank
Userlevel 7
Badge +27
Add the embedded data field ResponseID to the beginning of your survey flow and don't assign a value. Then pipe it into your redirect url as ${e://Field/ResponseID}.
SamS
Rank
Badge +4
Hi, @michielwiggers! You might want to check out our new support page on a similar topic!
L
Badge

Hi there, I'm having a similar situation and am not particularly tech-savvy.
I need to establish an average amount of time to take the online survey AND need to establish a Reliability Coefficient for the survey. In my naïve brain, I see this as a test-retest but tried to check how I'd do this by setting up a dummy survey, taking it, and then retaking it as a new survey, getting a new link through the data & analysis tab. I found my first answers on the "new" test but wanted it to be blank so that it would truly be a "retest" not a revision.
Ideally, I'd like to be able to set it up that an anonymous respondent to my first link would be able to get a 2nd, blank test after a 2 week break to meet what I understand as "best practices" for this process.
Help??? I welcome all comments, and TIA!


Leave a Reply


Terms & Conditions