A few versions of the same project... | XM Community
Solved

A few versions of the same project...

  • 28 May 2018
  • 4 replies
  • 221 views

Hello everyone,
Hopfully you will understand what I am trying to do and know how to do it but any assistance will be greatly appreciated.

I want to run a survey for my MA research project, it is an online survey about crime severity. I am playing around with some vignettes ans asking respondents to rate the severity of the crime described and their suggestion of punishment. I want to make a few versions of the same project and have Qualtricks distribute them randomly...

Does that make sense? Is it possible?
In case it isn't, what can i do? other than give each participant a link to a different project?

thanks in advance
icon

Best answer by KatC 29 May 2018, 17:47

View original

4 replies

Userlevel 7
Badge +27
You would probably better off doing the randomization within a single survey. For example, randomize the vignettes so you show different vignettes to different respondents.
thank you!
but i need each survey to contain a different combination of the vignettes so that no 2 participants receive the same combination
Userlevel 4
Badge +2
Similar to what Tom said, this may be easier by using one survey. You can make a separate block for each of the vignettes. Then in the survey flow section of qualtrics you can randomize the different blocks so that there's randomization. Further, you can set limits to frequency. I'd suggest to check out this link on randomizing in Qualtrics to get more details on how this may work for you.

To answer your original question, you can create different surveys for the vignettes, and then assign people to take the different survey links. However, you would then need to combine survey data responses and it could lead to a host of other issues especially since you'd have to create several different survey links.
Userlevel 7
Badge +19
If you really want 3 separate surveys you can build them all out. Then I would randomly assign a group in Excel to each participant, 1-3. But, this is going to be really hard to run analysis on (which is what I think the other respondents meant by saying "it would be easier to randomly show blocks").

The process of getting respondents a random set of vignettes via both methods is simple. But randomly displaying them, rather than randomly sending them, is going to be much simpler on you when you're trying to conduct the analysis.

Leave a Reply