ABSTRACT
In experimental research, a confound is a source of confusion regarding the explanation for a given difference. Suppose, for instance, a researcher gave an experimental group instruction in algebra using a new software program and, at the same time, a student teacher taught the same algebra skills to the participants using manipulative devices. Furthermore, suppose the control group received no algebra instruction. If the experimental group outperformed the control group, the results would be said to be confounded because there would be no way to sort out the relative influence of the software versus the manipulatives. Unfortunately, not all sources of confounding are as obvious as the one in this example.
