[ Contents | Search | Next | Previous | Up ]
From: Klaas Bakker, MSU, Springfield MO.
Date: 15 Apr 2010
Time: 22:33:39 -0500
Remote Name: 173.22.193.98
Ken, thanks for bringing our discussion to the forum, and also thanks Robin for referring the statement I made in the JFD Special Issue on cluttering. I still believe all of this today without change. Let me make a few comments relative to the model Robin suggested, as well as an alternate model I thought of and Ken referred to in his post. I believe both are very viable approaches and represent the "homing in on the truth" of a syndrome that we are referring with the label cluttering. Yes, I believe a very productive approach would be to "let the data speak" and tell us what are possible configurations of symptoms, clinical signs and characteristics. Robin's data are clinical observations, and quality recordings that can be shared through the Internet (but with proper confidential restrictions in place). This data base, like a wiki (e.g., Wikipedia) can be refined by the international community of specialists. The "dream" approach I have suggested at several occasions is statistical in nature and starts with an interest in a large group of individuals with resesemblences to common descriptions of cluttering, but at the same time is varied in nature as well, and ideally aselectively created by a group of researchers with an open mind. Of this group we would need to determine a fairly large number of measures of interest regarding cluttering (speech rate related, disfluency related, articulation/phonology related, language related, cognition and attention related, pragmatics related and more). The next step is to apply factor analysis to determine if configurations of variables show up that could be diagnostic categories (yes, there could be more than one!). Earlier I called this my "dream" solution because there is a catch. We need a number of subjects a number of multiples of the number of variables considered (and each variable needs to be at least "interval" in nature). With cluttering being a fairly low incidence problem this may not be an easy task to accomplish and most likely suggests a multiple institution approach.