When conducting a study with multiple groups, there are two primary approaches for assigning participants to those groups.
In this approach, each participant is in one, and only one, group. In other words, half of the participants receive the treatment and half receive the control. For a true experiment, the assignment of participants to these groups should be done randomly.
In this approach, each participant is in all groups. In other words, each participants receives both the treatment and the control. For example, if the treatment is use of a new feedback tool for assignments, the participants would have some assignments where they receive feedback from the tool (treatment) and some where they receive the traditional feedback (control).
This approach has both advantages and disadvantages over the Independent Groups approach.
Counterbalancing helps researchers address the order effects of the Repeated Measures design. In a counterbalanced design, the researcher divides the participants into groups. Each group receives the treatment and the control in a different order. The picture below illustrates this design.
Depending on the nature of the treatment, there are times where it is not feasible to use this approach. In a case where the application of the treatment fundamentally changes the participant (e.g. they learn a structured way to do a formerly unstructured task), counterbalancing is not possible.