Professor Leanne Togher
Discipline of Speech Pathology, The University of Sydney
The data for Jonathan suggest that intervention led to the change in behaviour.
However, there may be other explanations.
Alternative explanations can arise from lack of experimental control.
In research, we describe these other factors (e.g., a change in the client’s health, the client maturing, the client receiving another intervention in parallel) as threats to internal validity.
To account for threats to internal validity, we need to move beyond simply describing behaviours (case studies) to experimental designs.
Think about your own client or the case example. Write down 10 things that could potentially influence his or her progress in intervention. Then, for each factor (threat to internal validity), think of and write down a strategy for how you could control for this factor. While some can be controlled for simply through actions (e.g., ensuring a client does not receive another intervention at the same time) other factors (e.g., maturation) can only be controlled for using an experimental design.