Clinical significance

[@joshboston2323]

thanks for this video. However, wouldn't a placebo also give some effect? It is called the null-distribution, yet I feel that a placebo will always generate some effect. Would it not be better to test it against "no treatment"? Thank you in advance!

[@ndevent496]

I know you asked this a while ago and im not the author of this video, but perhaps someone comes a long and is still interested in this. So, the thing with a lot of treatments is that you do not measure a ' pure' treatment effect. The treatment effect can be obscured by other effects that you are also measuring. For my students I always give the example of a social anxiety training. So these seem to work , however if you compare them to for example a wait list condition you are saying ' my treatment works compared to doing nothing' . Which is a valid research question. In some cases however we want to ' take away' some of the effects that are also measured in your SAD training such as simply going out and talking to a therapist. IN that case the treatment itself may not cause the effec tbut the ' talking to people' does. So in order to really pin point the effect of the treatment you can have an ' active control group' or a placebo group in which all other parameters are roughly equal (they have to come to the therapist in this case for example). This way the difference that you measure between groups can more precisely be attributed to the actual therapy/intervention instead of other effects. So yes you do expect the placebo to have some effect, but if your intervention is good, it should still be significantly better over the placebo group.

[@Daniel-rk2qz]

So you are saying that although something is statistically significant it doesn't really define the "impact" it has or the "magnitude" it has on the patient? Does this also mean that studies that determine number needed to treat in terms of lowering cardiovascular event risk, rather, determines this magnitude?

[@fsandesh]

thank you