Understanding Test Scores
Standard Score- Raw scores that have been converted to have a designated mean (average) and standard deviation (SD). Raw scores are converted so the position of test-takers performance is relative to other test-takers. The Mean is typically=100; SD typically=15. (A standard deviation is the square root of the variance, or spread in scores, representing the average of the squared deviations from the mean.)
A standard score of 100 is exactly average; Typically, standard scores from 90 to 109 are within the average range.
Percentile- A ranking that permits us to determine an individual’s position relative to a standardization sample (e.g. If a child scores in the 51st percentile, this means that out of 100 students his age who took the same test, the student performed better than 51 of them.
Percentiles of 25 to 73 are within the average range
Confidence Interval- a band or range of scores around the obtained score that likely includes the examinee’s true score. (Traditionally, we select points that represent the 95% level of confidence).
Some comments on grade/age equivalents:
Grade equivalents exaggerate small differences in performance; a score slightly below the median (or the middle number in a set of scores arranged by magnitude) may result in a grade level equivalent of one or two years lower.
Grade equivalents encourage comparisons with inappropriate groups. For example, we should not say that a second grader who obtains a grade equivalent of 4.1 in arithmetic is functioning like a fourth grader at the beginning of the school year; the fourth grade is the wrong comparison group. The second-grade student shares with the average fourth grader the number of items right on the test, no other attributes associated with fourth grade mathematical skills.
Age and grade equivalent scores require careful interpretation because they are deceptively straightforward in a way that is misleading. It’s too easy for a person to assume that because their child has gotten the same score (e.g. grade equivalent of 9.3) as the average person in early ninth grade, that they have therefore mastered the curriculum through eighth grade. But most tests with age or grade equivalent scores are not criterion-referenced tests (mastery of a particular skill) and cannot generate curriculum mastery information.
Reliability- consistency of measurement; Validity- accuracy of measurement