Contrbutions From Our Members: Common Problems With Standardized Tests—And How to Cope With Them Since the 1980s, the quality of standardized speech and language tests for children has steadily improved (Balthazar, 2003). Nonetheless, no test is perfect, and it is incumbent upon the user to know the limitations of the test being used. All clinicians should be familiar with the basic criteria for a ... Contributions
Contributions  |   July 01, 2003
Contrbutions From Our Members: Common Problems With Standardized Tests—And How to Cope With Them
Author Affiliations & Notes
  • Catherine Balthazar
    Department of Communication Disorders, Governors State University, University Park, IL
Article Information
Contributions
Contributions   |   July 01, 2003
Contrbutions From Our Members: Common Problems With Standardized Tests—And How to Cope With Them
SIG 1 Perspectives on Language Learning and Education, July 2003, Vol. 10, 17-18. doi:10.1044/lle10.2.17
SIG 1 Perspectives on Language Learning and Education, July 2003, Vol. 10, 17-18. doi:10.1044/lle10.2.17
Since the 1980s, the quality of standardized speech and language tests for children has steadily improved (Balthazar, 2003). Nonetheless, no test is perfect, and it is incumbent upon the user to know the limitations of the test being used. All clinicians should be familiar with the basic criteria for a well-standardized test, such as minimum sample size, reliability, and validity (Hutchinson, 1996; McCauley & Swisher, 1984a; Sabers, 1996), and should evaluate the characteristics of the normative sample relative to those of the child to be tested (Balthazar, 2003).
The normative sample must be of adequate size (100 individuals per relevant and distinct group) and must be randomly drawn from a population that is representative of the individuals to be tested. Generally, a test manual ought to report the sample characteristics thoroughly, including racial, ethnic, and linguistic background, socioeconomic status, geographic region, and disability status. Problems with the size and composition of the sample can result in low reliability for all or some of the sample and can make it inappropriate to use the scores to determine whether an individual’s performance falls outside the normal range (McCauley & Swisher, 1984a).
First Page Preview
First page PDF preview
First page PDF preview ×
View Large
Become a SIG Affiliate
Pay Per View
Entire SIG 1 Perspectives on Language Learning and Education content & archive
24-hour access
This Issue
24-hour access
This Article
24-hour access
We've Changed Our Publication Model...
The 19 individual SIG Perspectives publications have been relaunched as the new, all-in-one Perspectives of the ASHA Special Interest Groups.