Current survey length is quite long and am trying my best to lower the drop rate of my respondents.
For measuring reliability and validity, would these options seem okay?
Repeat 1 CVA profile for reliability test
Adding 3 holdout sets for validity test as out of sample
I am a bit concerned about these numbers.
I have 5 attributes and a total of 13 levels.
I am providing 15 scenarios (1.5 times more than the parameters) and 3 choice holdouts and 1 repeat scenario.
Would I need to increase the number of holdouts and repeat scenarios?
I haven't collected the data yet.
Thank you for your help.