OK, this could be the design you want!

The beauty of factorial design is that you can check out a few different factors or treatments at once and see if there is a main effect or if there is an interaction effect. It could be a more realistic model and also save time by testing two are more treatments at the same time. I'm thinking of using this design so let's go through my case to give you a better idea.

Ok, in my case, I'll be working with farmers in Central America. They will be recieving courses and different types of evaluation. I want to see if the length of the course and different approaches to participatory evaluation have an effect on the satisfaction the farmers get from receiving the courses. There are the 2 factors. These factors have 2 levels each.

Courses:

one day courses to some
one week courses to others

Evaluation:

participatory evaluation conducted with the farmers and facilitators during the course.
participatory evaluation conducted with facilitators outside of course.

There will also be a control group that does not receive any treatment.

How would this look graphically?

This would be considered an incomplete factorial design because it is not fully crossed. It is a 2x2 design with a seperate control group.

To try to plan your design there are some worksheets that my help. Worksheets

And don't worry! If your plan doesn't fit exactly like this it may not be a problem. There are many variations on the theme. Factorial Design Variations

There are only 2 sites that I've found with useful beginner information on factorial design that you can read up on, although there are zillions of businesses that you can hire for a consultation!
Expanation and examples
An outline from a book on research design

Back to the definitions of main effect and interaction effect. In my example, a main effect would be if participatory evaluation always worked better if the farmers were involved too, despite the length of the course. An interaction effect would be if partipatory evaluation with the facilitators only had an effect with the week long courses.

The analysis of the data after the research is finished is where my tutorial stops. Here are some links so you can start to look into it. Good Luck!

back to the beginning

back to tutorials

to Trochim's Home Page