I am asking this question because 2 factorial design normally deals with 2 factors and 2 treatments. This means that every combination of the treatment is possible. The issue that arises is where Treatment B is created based on Treatment A but using a different design technique. Because A and B contain the same information, using within-subject sampling, where each subject is given treatment A and B could bring about a learning effect. In this case, the subjects become too knowledgeable with the two treatments. Will an in-between subject suffice in resolving this learning effect? Where each subject is given one treatment each, but randomly. Expert views will be appreciated.
I may have misunderstood, so please correct me if I'm wrong.
Would it be possible for 50% of your sample do complete treatment A followed by B, and the rest complete B followed by A; this would reduce order/learning effects.
But not sure, if you can do this with your data.
If not, I think the random allocation of participants to different treatments would work.
Hope this helps.
Thank you Lee! Your contribution is in order. However, I am more concerned with the learning effect than the order effect. I strongly feel that having subjects use one treatment to perform the task will influence how they perform on the same task but using another treatment which contains the same type of information but different design technique. This will surely constitute a threat to validity since subjects may end up performing tasks based on prior knowledge gained from the previous task. I think sticking with the in-between 2 factorial design is fine, and again, as you suggested, the random assignment will be necessary. Thank you once again.
No worries, and I think you're right, the between subjects design may be necessary here.
Good luck :).
- Badges
- Science method
- Similar topics
- Teaching