Glossary » single-case experimental design
single-case experimental design
a method for determining the effect of an experimental manipulation with only one test subject (one user, one organization, etc.), basically by applying and removing the intervention over time:
- measure the dependent variable initially
- apply the intervention, and measure again
- remove the intervention, and measure again
If the intervention has an effect, then the dependent variable should change after the intervention, and return to its initial value when the intervention is removed (removing the intervention again helps to rule out the possibility that something else changed at the same time as the intervention which caused the actual effect).
This technique may be useful, for instance, in testing changes to a groupware system being used by a large organization. Since it is difficult to find other organizations using the same system in the same way, it may be difficult to create a control condition.