Quanti-Scaling of Articulatory Defectiveness in Cleft Palate Speakers

Abstract

In recent research (3)' a new method of scaling articulatory defective— ness for cleft palate speakers was developed. This method, describable as quanti-scaling, allows ratings of articulatory proficiency on a one to seven scale. Although resultant ratings are similar to ratings obtainable from the equal-appearing-intervals rating procedure ('7), the method used to obtain the ratings is entirely different and novel. Designed originally as a test of attitude, the equal—appearing-intervals rating procedure of Thur-stone and Chave (7) provided descriptive terms for rating a given atti— tude on any one of several points separated by equal—appearing-intervals on a continuum scale. In past research this rating procedure has been utilized as a seven point scale for a number of studies of nasality, most of which used cleft palate speakers as subjects (I). It has provided a basis for assessment of harshness among non—cleft palate speakers (5) as well as harshness, breathiness, and hoarseness among cleft palate speakers (2). As a nine point scale, the method of equal—appearing-intervals rating has also been used to assess normal and articulatory defective speakers (4). The quanti-scaling technique for rating articulatory defectiveness of cleft palate speakers (3) uses a different approach from the descriptive guidelines of the equal-appearing-intervals rating procedure. This new rating technique was developed experimentally as an attempt to obtain ratings of articulatory proficiency on the basis of explicit, less ambiguous guidelines than are available in conventional procedures (7). It was reasoned that improved criteria for rating articulatory proficiency might necessitate use of fewer judges (assuming proper qualification and train— ing) than classic scaling procedures, which employ descriptive terms (4, 6), usually employ. A discrete numerical rating scale ranging from one to seven is employed. In the original procedure (3) each test sentence to be rated contained three consonants—/p/, /t/, and /k/ or /f/, /s/, and /tf/ ——for quanti—scaling. In each test sentence as it appeared on the rating sheet, each of the three appropriate consonants was underlined as a re— minder to judges of the delimitation of the judgments. If all three conso