Standardized Testing
Introduction
The Arizona Revised Statutes, Section 15-744, mandates that a nationally standardized norm-referenced achievement testing program in the subjects of reading, grammar, and mathematics be adopted and implemented for use in Arizona schools. The legislation exempts the testing of pupils with specified disabilities and allows the governing board of a school district to exempt pupils who are limited English proficient for up to three consecutive years.
Test scores are reported by subject, subtest and skill at the pupil and classroom levels. Aggregated scores are provided for the school, district, county, and state levels. PUSD students consistently score above county, state, and national averages in reading and mathematics.
The Scores
Definition. National percentile ranks indicate the relative standing of a student in comparison with other students in the same grade in the norm (reference) groups (in this case, the nation) who took the test at a comparable time. Percentile ranks range from a low of 1 to a high of 99, with 50 denoting average performance for the grade. The percentile rank corresponding to a given score indicates the percentage of students in the same grade in the norm group obtaining scores equal to or less than that score. For example, a student earning a percentile rank of 62 achieved a score that was equal to or better than the scores earned by 62% of the students in the national sample.
Uses Percentile ranks are useful for comparing a student's performance in a particular subtest relative to the performance of other students. Percentile ranks are also useful for comparing a student's performance across subtests in a score profile.
Percentile ranks must always be interpreted with reference to the group from which they were derived. Percentile ranks from two different test batteries should not be directly compared with each other unless they are derived from the same samples of students or unless the score scales have been previously equated.
Limitations Percentile ranks do not represent actual amounts of ability. Further, they do not represent equal units along the score scale. For example, the difference between percentile ranks of 5 and 10 does not reflect the same difference in performance as the difference between percentile ranks of 50 and 55. Since percentile ranks do not represent equal units, and since their interpretation is limited to the reference group from which they were derived, they are best used for reporting scores when position in relation to the reference group is of primary interest.
