ABSTRACT
Within-subject designs are one of a class of related designs—the scores in different conditions are related in that they belong to the same subject. Within-subject designs are frequently used when the independent variable has many levels. Many levels of the independent variable means many experimental conditions. A special case of a within-subject design is the pre-test/post-test paradigm. In this paradigm subjects are measured on the dependent variable once, then they are exposed a single treatment, and then they are measured on the dependent variable again. Within-subject designs require fewer subjects than between- subjects designs. The major limitation of within-subject designs, order effects, can be addressed through experimental procedures. An alternative procedure for minimising order effects when student researcher have a lot of experimental conditions is randomisation. An alternative method of achieving incomplete counterbalancing is to use a procedure called randomisation and rotation.
