Any child tested multiple times would produce a range of values due to test conditions, examiner, the child’s state (e.g., fatigue, boredom, lack of motivation, illness), etc. These scores may not be reflective of the child’s “true score.” For this reason, it is important to include a confidence interval whenever a score is reported. The confidence interval is a range of values surrounding the score obtained from the administration of a standardized test.
Since the standard score may not be the child’s true score, the confidence interval increases the likelihood that the test has produced accurate information by including a range of probable scores. New York State has recognized the importance of treating a standard score as within a confidence interval. “Assessment professionals should be careful to treat each score from standardized tests as falling within a confidence interval whose size is determined by the reliability of the test. . . . This represents a more appropriate description of the student’s ability. It also makes a clear statement of our recognition of the inherent limitation in the technology of standardized tests.” (NYCDOE CSE SOPM 2008, p. 52)
A confidence interval changes depending on the level of confidence. Most tests provide confidence intervals at 90% confidence level and at 95% confidence level. This percentage indicates the likelihood that the “true score” falls within the confidence interval. The confidence interval gets bigger as the confidence level increases because a wider range of scores must be included to ensure that the “true score” falls within it. Confidence intervals, however, also demonstrate the lack of information gained from using and scoring standardized tests in order to identify language impairment or cognitive delay. Consider the scores included within confidence intervals in the following standardized tests:
WIPPSI-III: 90% confidence intervals for 3-year-olds
VIQ 10 to 11 SS range- VIQ 71 score has a confidence interval that ranges from 67 to 78 or a mild delay to a moderately severe delay. The test has told us that the child has a mild to moderately severe delay. Many states require that a child be a certain number of standard deviations (e.g., 2) from the mean to qualify for services. However, the score gained from this indicates that the child may or may not qualify for services. What have we really learned from the score?
PIQ 14 to 15 SS range- PIQ 76 score ranges from 71 to 85, or a moderate delay to within normal limits.
FSIQ 10 to 11 SS range- FSIQ 74 ranges from 70 to 81 or a moderate delay to low average.
Resources
NYCDOE Initial Guidance Document for Speech and Language Evaluators
References
New York City Department of Education (2008). Standard Operating Procedures Manual.http://schools.nyc.gov/documents/d75/related/counseling/SOPM_08.pdf
