Identification of Children With Language Impairment: Investigating the Classification Accuracy of the MacArthur–Bates Communicative Development Inventories, Level III Purpose This study tested the accuracy with which the MacArthur–Bates Communicative Development Inventories, Level III (CDI–III), a parent report measure of language ability, discriminated children with language impairment from those developing language typically. Method Parents of 58 children, 49 with typically developing language (age 30 to 42 months) ... Research Article
Research Article  |   August 01, 2009
Identification of Children With Language Impairment: Investigating the Classification Accuracy of the MacArthur–Bates Communicative Development Inventories, Level III
 
Author Affiliations & Notes
  • Elizabeth Skarakis-Doyle
    University of Western Ontario, London, Ontario, Canada
  • Wenonah Campbell
    University of Western Ontario, London, Ontario, Canada
  • Lynn Dempsey
    University of Western Ontario, London, Ontario, Canada
  • Contact author: Elizabeth Skarakis-Doyle, School of Communication Sciences and Disorders, University of Western Ontario, London, Ontario N6G 1H1. E-mail: eskaraki@uwo.ca.
  • Lynn Dempsey is now at the Department of Applied Linguistics, Brock University, Ontario, Canada.
    Lynn Dempsey is now at the Department of Applied Linguistics, Brock University, Ontario, Canada.×
Article Information
Development / Research Issues, Methods & Evidence-Based Practice / Language Disorders / Research Articles
Research Article   |   August 01, 2009
Identification of Children With Language Impairment: Investigating the Classification Accuracy of the MacArthur–Bates Communicative Development Inventories, Level III
American Journal of Speech-Language Pathology, August 2009, Vol. 18, 277-288. doi:10.1044/1058-0360(2009/08-0035)
History: Received May 25, 2008 , Accepted January 21, 2009
 
American Journal of Speech-Language Pathology, August 2009, Vol. 18, 277-288. doi:10.1044/1058-0360(2009/08-0035)
History: Received May 25, 2008; Accepted January 21, 2009
Web of Science® Times Cited: 8

Purpose This study tested the accuracy with which the MacArthur–Bates Communicative Development Inventories, Level III (CDI–III), a parent report measure of language ability, discriminated children with language impairment from those developing language typically.

Method Parents of 58 children, 49 with typically developing language (age 30 to 42 months) and 9 with language impairment (age 31 to 45 months) completed the CDI–III, a 2-page questionnaire that includes 100 vocabulary items, 12 sentence pairs, and 12 questions regarding linguistic concepts.

Results A discriminant analysis indicated that the CDI–III total score together with age classified children into language status groups with 96.6% accuracy overall. The corresponding likelihood ratios supported this strong level of accuracy, although precision may not be as high as indicated by broad confidence intervals.

Conclusions Results of this study contribute to the accumulating evidence on the types of valid inferences that may be made from the CDI–III, specifically its classification accuracy. Further research should continue to investigate classification accuracy in larger samples with broader maternal education levels and with different types of language impairments. Additional research should also investigate the classification accuracy when the CDI–III is used in combination with other tests.

Acknowledgments
This work was supported by grants from the Toronto Hospital for Sick Kids Foundation Grant XG01-081, the Ministry of Health and Long Term Care–Children With Special Needs Branch, the University of Western Ontario Academic Development Fund New Research and Scholarly Initiative Award to the first author, and Social Science and Humanities Research Council of Canada Doctoral Fellowships to the second and third authors.
We express our sincere appreciation to Dr. Philip Dale for sharing unpublished data from the CDI–III norming study. Additionally, we are grateful to Joselynne Jacques, project manager; Melanie Beaudin, Sarah Pifher, and Brooke Thornton for their assistance in data collection; and Jayna Amting, Allison Estabrooks, Sarah Gillespie, Carla Montgomery, and Megan Wells for assistance in preparation of the data. We extend gratitude to the children, parents, clinicians, and teachers who participated or otherwise assisted in this study.
This article was a result of collaborative effort among the three authors; hence authorship is shared equally. Portions of this article were presented at the Symposium on Research in Child Language Disorders in Madison, WI, in June 2006, and at the American Speech-Language-Hearing Association Annual Convention in Boston in November 2007.
Order a Subscription
Pay Per View
Entire American Journal of Speech-Language Pathology content & archive
24-hour access
This Article
24-hour access