ABSTRACT
The major argument of this article is that failing to measure what is taking place in treatment and control conditions can lead to scientifically invalid conclusions. It is argued that researchers are ethically responsible for being aware that variables related to the therapist, client, and the therapeutic relationship (as well as their interaction) might play a confounding role when treatment and control conditions are compared. As a consequence, they should either measure these variables or be tentative in their interpretation of their findings.
