Scoring & Interpreting

Scoring and Interpreting Standardized Tests

1. Raw score-This refers to the actual number of items on a test a student answered correctly.

   1. These scores are usually not reported because they are not meaningful information unless tests are on the same scale.

   2. In order for raw scores to be meaningful they have to be converted to some type of derived score such as percentile rank, grade equivalent score, or standard score so that accurate comparisons can be made

2. Correct/Incorrect possible score-The number of both correct and incorrect responses out of the total number of test items are reported, typically expressed as a percentage.

   1. Generally associated with informal, criterion-based tests

   2. Commonly used to inform instruction

3. Standard score-This is the term for a variety of scores that represents performance by comparing the deviation of individual scores with the mean score in a norm group of the same chronological age or grade level.

   1. These allow for the comparison of performance across tests

4. Stanine-A combination of the words standard and nine.

   1. Scores range from one to nine, reflecting the bell curve

   2. Scores 4-6 represent the average range, stanines above that are considered above average, and achievements below this are considered below average.

   3. It may be helpful to think of each stanine in this format:

      • 9 = Very superior

      • 8 = Superior

      • 7 = Very good

      • 6 = Good

      • 5 = Average

      • 4 = Low average

      • 3 = Considerably below average

      • 2 = Poor

      • 1 = Very poor

5. Percentiles-Describes how a student compares with others in the same age group or grade. For example, a percentile of 80 means that 80% of the sample scored at or below the examinees score

   1. The most frequently used norm score

   2. Range from a lower of 1 to a high of 99 and reflects the bell curve (50th percentile is the average

   3. A percentile rank is the percentage who had scores the same as or lower than the student being tested

      • Useful for describing the student’s relative standing in the population

6. Standard error of measurement (SEM)-An estimate of how often errors are expected

   1. A low SEM is an indication of high reliability and a high SEM is an indication of low reliability.

7. Confidence interval-The range of scores in which a true score will fall within a given probability range.

8. Grade and age equivalent score-Scores in terms of grade/age levels, expressed in years and tenths of years

   1. Grade and age scores are most meaningful when the test students have taken is at the right level and is not more than a year above or below average.

   2. The APA has advocated that these scores not be reported because of their inadequate statistical properties.

   3. For example, a student who achieves the reading comprehension grade-equivalent score of 6.6 is reading like a sixth grader in the sixth month of school