choice-metrics.com

Dear all,

I'm about to start my main survey with a sample size of 600-700 participants that I hope to reach. My question is, would it provide enough variation for this sample size if I create 24 choice sets in total as three blocks consist of 8 choice sets? is there any criteria for this issue? Some papers in literature apply more for ex. 80 choice sets in 10 blocks.

That depends on the number of coefficients you are using as well as the number of alternatives in your choice model.

The number of choice tasks S needs to satisfy:

S*(J-1) >= K

where J is the number of alternatives and K is the number of coefficients you are estimating (including for dummy coded variables, interaction effects, etc).

The smallest S that satisfies the above is the MINIMUM number of choice tasks you need in your design (before blocking). But generally you will want to use a larger design than that to have enough variation, so you could choose to use something like 2*S, 3*S or 4*S to make sure that you have enough variation.

You may want to refer to Sandor and Wedel (2005), https://journals.sagepub.com/doi/abs/10.1509/jmkr.42.2.210.62285.

Michiel

Thank you for your answer Mr. Bliemer.