Question 201:

Asked on May 20, 2008

Tags: Counterbalance

1

Answer:

No answer provided yet.

If you are only interested in counterbalancing the effects of the product order, not the test order, then there are 3 combinations of the 3 products.

1. ABC
2. BAC
3. CBA

To be evenly balance, your number of participants should be a multiple of 3, which is not always feasible. Repeat the above sequence for as many participants that you have. For example, if you are testing 12 people, the above sequence would be repeated 4 times. Then randomly assign a product to A, B or C and finally, randomly assign each participant to each order. Your final table might look something like the following, where the number represents the participant.

1. ABC : 5
2. BAC : 7
3. CBA : 3 
4. ABC : 4
5. BAC : 2
6. CBA : 1
7. ABC :11
8. BAC: 12
9. CBA : 8
10. ABC : 7 
11. BAC : 9
12. CBA : 6