[ Contents | Search | Post | Reply | Next | Previous | Up ]
From: Tom
Date: 03 Oct 2009
Time: 02:14:21 -0500
Remote Name: 91.148.89.192
""I am still not sure what the point is about validity of post-therapy speech done when the client knows who the listener is, as opposed to anonymously. As researchers, we often feel that this can make results look better than they might, but surely the speaker would know the answer. Am I missing something?"" I do not understand what you do not understand! :-) I am saying that the fluency varies depending on who you are speaking to, even in real life. So when the researcher calls an ex-patient up, the ex-patient will change their behaviour because s/he is put back into a treatment setting which will affect fluency or tries to put on best behaviour because they are recorded. Ideally, one should call them up anonymously or record their speech the whole day. For example, in the Pagoclone study they had a significant reduction in dysfluency in the control group! Where does this effect come from? Just placebo effect? I doubt it. That's what I call the measurement problem. And all outcome studies suffer from this effect in my view; so one probably needs to reduce the positive outcome by 20-30%. Of course, in terms of questionnaires for subjective measures the effect is probably not present or much lower.