Reply To: FULLY RANDOMIZED CHOICE/CONJOINT EXPERIMENT

[David Thul]

Hello Annina,

you could implement the planned project as follows:

Create two static lists (one list per package).

Static list 1 (Package A)

Static list 2 (package B)

 

Then you create the corresponding list elements within the respective list:

Static list 1

Package A: Attributes 1 / Level0 -> always in random selection

Package A: Attributes 1 / Level1 -> always in random selection

… and so on

Static list 2

Package B: Attributes 1 / Level0 -> always in random selection

Package B: Attributes 1 / Level1 -> always in random selection

… and so on

 

Now we have two static lists, which are filled with the required list elements. The list elements are randomly selected via the inclusion condition (always in random selection)

For the next step you have to create 12 user-defined variables.

 

We need 6 user-defined variables per package (6 for package A, 6 for package B)

 

We let these user-defined variables be filled via a list trigger.

For this, we create two list triggers on a page that is sure to be seen by all people (who are supposed to answer the question).

 

Type: List Trigger

Execution position: directly

Condition: execute always

 

IMPORTANT: Check ”randomize list elements” (second blue underline)

These triggers ensure that the label of the respective list elements is written into the user-defined variables when the trigger is triggered.

Finally, you can output the user-defined variables (which are filled via the list triggers) via the corresponding placeholders within a matrix question.

 

Please check extensively how the corresponding approach can be evaluated for you.

Tip: Create another 12 user-defined variables and have them filled with the element number instead of the label.

Kind regards

David