When the statistical power of a research design is low, the study may fail to detect effects that are really there. Therefore, if a study with low statistical power finds a significant result, we can be confident that the effect is real, but if a study with low power fails to find a significant effect it can be hard to tell if this means that there is no real effect there or if there is a real effect but the study wasn't sensitive enough to detect it. If the power of a study is low this raises questions about negative results but does not raise questions about positive results. In looking at the literature on the importance of homework in CT, the arguement that many of the studies have relatively low power would lead to the conclusion that we can have some confidence in the studies that show that homework has important effects on outcome (such as Burns and Spangler) but may need to be skeptical about studies which fail to find an effect (unless those studies have adequate power). I don't have empirical data about the way in which Cognitive therapists use homework in day-to-day practice. However, I do know that Beck and all the leading Cognitive therapists consider homework to be very important. My clinical experience is consistent with the studies which show that clients who do homework on a regular basis improve more and improve faster. I advocate developing a homework assignment of some sort (not necessarily written) at the end of each therapy session.
Replies:
|
| Behavior OnLine Home Page | Disclaimer |
Copyright © 1996-2004 Behavior OnLine, Inc. All rights reserved.