What is the difference between norm-and criterion-referenced?
- A norm-referenced test compares a child's score to a group of typically developing children of the same age across the country.
- A criterion-referenced test measures mastery of a specific set of skills based on the child's age or grade level.
What is meant by reliable and valid?
- A reliable assessment is one that provides the same results with repeated administrations. Typically, you should be able to administer a test on two different days and obtain consistent scores.
- A valid test is one that measures what it is supposed to measure and also gives the same results as a similar instrument. For example, if you give the CAAP® and the SPAT-D II and you should see the same error patterns.
- A standard score is a score based upon the standard deviation of the test population. Typically, a standard score will have an average or mean of 100 and a standard deviation of 15. Thus, a child receiving an average score on a test would fall one standard deviation above or below the mean, and his/her score would range between 85 and 115. In most school districts, a child will qualify for services if his/her scores are 1 1/2 to 2 standard deviations below the mean.
- A percentile rank is a converted standard score. It refers to the percentage of children who fall at or below a specific score. For example, a percentile rank of 90 means that the child scored better than 89% of the test population.
- An age-equivalent score indicates that the child's performance on the test may be typical of that particular age group. For example, a child may be 4 years old, but receives an age equivalent of 3 1/2 based on his/her responses to the test.
What is a confidence interval?
- A confidence interval is the range of a child's test score. For example, a child may receive a standard score of 85 with a confidence interval of +8 or -8. Upon repeated administration, the child's score may vary
from 77-93.
