COMPARISON OF SPEECH AND PRACTICED NONSPEECH INTRAORAL PRESSURE WAVEFORM CHARACTERISTICS
Intraoral pressure waveforms of a learned volitional nonspeech task were compared to that of a parallel speech task in order to drawl inferences regarding a possible shared sensorimotor control mechanism. Similarities between the dependent variables at question (the percents of the increase and decrease interval involved in the total duration and the slopes of the increase and decrease interval) may provide preliminary evidence of a shared generalized motor program. The nonspeech task (which was devised as part of a larger study by Shaiman et al., 2004; 2006) reflected the goal and complexity of speech, by the incorporation of intraoral pressure targets and practiced, co-articulated gestures. Six subjects participated in the study. Subjects practiced the nonspeech task over two sessions, totaling to over 600 repetitions of the task, with KR regarding accuracy of reaching the pressure target provided for 65% of trials. Nonspeech retention data was gathered at the end of both practicing sessions. Parallel speech task data were then taken. The measures of the dependent variables were calculated by the division of the pressure waveform into three distinct intervals: the increase, plateau, and decrease interval. These intervals were automatically detected using a pressure waveform analysis program, which used the first derivative of the pressure signal to mark parts of the waveform. The means for the nonspeech retention data and the speech data were taken for each dependent variable. Univariate analysis revealed no significant difference between the speech and nonspeech condition for any of the four dependent variables (p<0.05). The finding of no significant difference for any of the four dependent variables may provide preliminary evidence for a shared generalized motor program for speech and nonspeech gestures. However, future research with data from additional subjects would assess this finding. Also, descriptive observations of waveform shape during the plateau interval indicate the need for further analysis of additional waveform measurements not analyzed in the current study, and also the need to control rate and precision of production in the future.
Advisor:John Durrant, PhD; Susan Shaiman, PhD, CCC-SLP; Malcolm McNeil, PhD, CCC-SLP
School:University of Pittsburgh
School Location:USA - Pennsylvania
Source Type:Master's Thesis
Keywords:communication science and disorders
Date of Publication:09/08/2008