my aim is to implement multiple choice tasks in our survey. Basically this worked, but now we want to ask the participants to choose from the options of the same choice set again after they chose a specific answer.
As can be seen here, we have 4 choices. If option 4 is chosen, I want to show the same choice set again (same specificity of attributes), but only with the options 1-3.
If it’s implemented from the same experiment a different choice set will be shown. If we generate a new experiment without the 4th option and show it, the values of the attributes of the choice sets will differ.
Is there a solution which can solve this specific problem?
Without seeing your allocation scheme it’s hard to say why this is or isnt working.
Here is the suggested workflow.
Get your ‘opt out’ experiment working as you want
Duplicate this - call it forced or something
Delete the 4th alternative
(If you want to clean up the warning - download codeplan, delete the additional redundant column for option 4 and re-upload)
This next bit is important as you want both your experiment design AND your allocations to be synchronised.
Use a fixed design row to allocate the ‘opt out’ - either explicit (1,2,3) or use some calculation of block+ offset. Anyway imagine this is called allocation1 (or $allocation1 if referring to it in a program)
As in the tutorial - detect if they selected option 4 using the a derived value with the custom function. So call the derived value ‘selected1’
SE_data_by_design_row($allocation1,”pref1”, name of opt out experiment)
then test and apply as a branch or other logic to control whether it is applied ($selected1 ==4)
Then you display the forced experiment with the same design row as the opt out using $allocation1
Hello, I’ve now duplicated the experiment and deleted the 4th option via Spec and Layout. In the Design area it says that the new experiment design has 20 columns, but 15 are used.
Then I’ve implemented the original experiment and the new adjusted experiment into different pages. The picked row out of the design for the opt-out experiment is set on default (random row design). I’ve then generated a new Derived variable on the same page called allocation1 with the code $current_design_row. And also generated a variable on the same page called selected1 with the code SE_data_by_design_row($allocation1,“pref1”,“SFDR”) where SFDR is the name of the experiment.
On the following page, I’ve implemented the new experiment and replaced random row design by allocation1 which is the generated variable on the page before.
Somehow the experiment still picks other design rows than the experiment before and is shown no matter which option I pick in the first experiment altough I’ve set a branch display condition $selected1==4
Do you see where it say’s ‘random design row’ in your scenario?
That’s why. Unless you specify the design row a random one will be chosen (without replacement).
What 99% of researchers do is to explicitly allocate rows as they need fine control over:
block allocation, consistency and dominance checks.
A noddy way to do this is just allocate 1,2,3,4 etc but if you have blocks or want randomisation you should code this.
In the tutorials have a look at the ‘Blocking’ example.
