Integrating several scenarios to one survey

Question

Hello,

for my research model am using a 2x2x2 design, this means, there are 8 different scenarios to be surveyed. What differentiates the scenarios from one another is their starting page (showing a picture which should be different in each scenario). In general, the questions are the same in each scenario, but there are  also some minor variations between the scenarios. Each respondent should be randomly assigned to one scenario with every scenario being displayed to the same number of respondents.

Is there a way of integrating all 8 scenarios to one survey? I've read something about Maxdiff, but as I am not very familiar to Sawtooth yet, this did not really help me out. I would highly appreciate any hint on how to deal with this issue.

Thank you in advance.
Best

Answer 1

There are a number of ways you could handle this.  I had a similar one recently with 4 different scenarios.  I just set a quota for the 4 different scenarios and set them to be randomly assigned (which will ultimately keep them equal although you may have to make some adjustments as you are nearing the end depending upon the response rate).  I then just used skip instructions based on the quota they were put in to only be asked the corresponding scenario questions.  I've found that to be the easiest way.  Each scenario should just be programmed in the same survey as unique questions.

Now, you could use one set of questions ( instead of 8 different sections) and write some Unverified Perl to change the question text or which image to be shown based on which quota they were assigned to.  This is the way to go if you really want the data combined together into the same questions.

Comments

simple and perfect approach

---

Thank you for your promt assistance!

I've programmed each scenario in the same survey as unique questions and was trying to set a quota for the 4 different scenarios. Unfortunateley, I have to admit that I do not have a clue of how to do this. I suppose this is an absolute beginner's question, but could you maybe give me some information on how to set a quota?

Thank you very much in advance! I really appreciate your help.

---

Just add a new question to your survey at the beginning and select the type to be 'Quota'.  On the cells tab that it brings you to add and name your 4 quotas and select the 'always qualify' option for each and add a cell limit to each for how many of each you want.  Then go to the advanced tab and a the bottom under 'check for cell membership' choose 'randomize all'.

That's pretty much it.  After that you can use that question for skip logic.  So, if you named the quota question 'Scenario', just write your skip logic based on Scenario=1 or Scenario=2, etc.

---

Hello! I have a similar design, i.e., eight different scenarios to be integrated into one study with each respondent being assigned to one scenario. The scenarios are different from one another regarding the starting page showing a different picture for each scenario, questions are identical across all scenarios. I've tried to program the questions for each scenario into the survey but due to my limited license I cannot add as many questions as I would need. Thus the quota solution is not feasible in my case. Do you have any alternative ideas or could you explain how to handle this Unverified Pearl approach? The more detailed, the more helpful, as this is my first Sawtooth experience. Thanks a lot!!!

---

I would still use a quota question as I described.  You can create a separate text question, each showing the image you want for the correct scenarios and as mentioned just write skips over that question if it's not in the right scenario quota.

If you want them to get the same follow up questions but with possibly different question text based on the scenario quota they fell into you can use Unverified Perl as I mentioned.  Here is an example of that:

[%Begin Unverified Perl
 if(VALUE("SCENARIO") == 1)
 {return "Scenario 1 Text Here";}
 elsif(VALUE("SCENARIO") == 2)
 {return "Scenario 2 Text Here";}
 elsif(VALUE("SCENARIO") == 3)
 {return "Scenario 3 Text Here";}
 elsif(VALUE("SCENARIO") == 4)
 {return "Scenario 4 Text Here";}
End Unverified%]