Heather Campbell1, Daphna Harel2, Elaine Hitchcock3, Tara McAllister Byun1. 1. a Department of Communicative Sciences and Disorders, NYU Steinhardt School of Culture , Education, & Human Development , New York , NY , USA. 2. b Center for the Promotion of Research Involving Innovative Statistical Methodology , New York , NY , USA , and. 3. c Department of Communication Sciences and Disorders , Montclair State University , Montclair , NJ , USA.
Abstract
Purpose: A current need in the field of speech-language pathology is the development of reliable and efficient techniques to evaluate accuracy of speech targets over the course of treatment. As acoustic measurement techniques improve, it should become possible to use automated scoring in lieu of ratings from a trained clinician in some contexts. This study asks which acoustic measures correspond most closely with expert ratings of children's productions of American English /ɹ/ in an effort to develop an automated scoring algorithm for use in treatment targeting rhotics. Method: A series of ordinal mixed-effects regression models were fit over a large sample of children's productions of words containing /ɹ/ that had previously been rated by three trained clinicians. Akaike/Bayesian Information Criteria were used to select the best-fitting model. Result: Controlling for age, sex, and allophonic contextual differences, the measure that accounted for the most variance in speech rating was F3-F2 distance normalised relative to a sample of age- and sex-matched speakers. Conclusion: We recommend this acoustic measure for use in future automated scoring of children's production of American English rhotics. We also suggest that computer-based treatment with automated scoring should facilitate increases in treatment dosage by improving options for home practice.
Purpose: A current need in the field of speech-language pathology is the development of reliable and efficient techniques to evaluate accuracy of speech targets over the course of treatment. As acoustic measurement techniques improve, it should become possible to use automated scoring in lieu of ratings from a trained clinician in some contexts. This study asks which acoustic measures correspond most closely with expert ratings of children's productions of American English /ɹ/ in an effort to develop an automated scoring algorithm for use in treatment targeting rhotics. Method: A series of ordinal mixed-effects regression models were fit over a large sample of children's productions of words containing /ɹ/ that had previously been rated by three trained clinicians. Akaike/Bayesian Information Criteria were used to select the best-fitting model. Result: Controlling for age, sex, and allophonic contextual differences, the measure that accounted for the most variance in speech rating was F3-F2 distance normalised relative to a sample of age- and sex-matched speakers. Conclusion: We recommend this acoustic measure for use in future automated scoring of children's production of American English rhotics. We also suggest that computer-based treatment with automated scoring should facilitate increases in treatment dosage by improving options for home practice.
Authors: Benjamin Munson; Jan Edwards; Sarah K Schellinger; Mary E Beckman; Marie K Meyer Journal: Clin Linguist Phon Date: 2010-01 Impact factor: 1.346
Authors: Sarah Hamilton Dugan; Noah Silbert; Tara McAllister; Jonathan L Preston; Carolyn Sotto; Suzanne E Boyce Journal: Clin Linguist Phon Date: 2018-05-24 Impact factor: 1.346
Authors: Nina R Benway; Elaine R Hitchcock; Tara McAllister; Graham Tomkins Feeny; Jennifer Hill; Jonathan L Preston Journal: Am J Speech Lang Pathol Date: 2021-07-07 Impact factor: 2.408