Page 33 - teachers.PDF

Test developers use the model of the normal curve in developing and norming tests. In this guide, we use it to show similarities between different types of normative test scores -- test scores that describe individual student performance in comparison to the actual performance of a large group of students.
Two statistics are helpful in discussing test score distributions:
• the mean and
• the standard deviation.
The mean is frequently called the average score. You compute the mean by adding all the
scores then dividing the sum by the total number of scores.
A deviation score is how far away the score is from the mean. For example, on a test with a mean of 15, a score of 20 deviates 5 points from the mean. The deviation score alone does not tell you whether this is a big difference or not. Rather, the standard deviation gives you a framework for interpreting this test score variability. You compute the standard deviation by taking the square root of the average of all the squared deviations. You can interpret standard deviation as an average distance that the scores deviate from the mean.
What are the advantages of raw scores?
• They are easy to compute.
• One of the most accurate ways to analyze a student's gains in achievement is to
compare the raw scores from two administrations of the same test.
What is the limitation of raw scores?
Raw scores do not contain a frame of reference for indicating how well a student is performing.
Total percent correct scores
Total percent correct scores tell you the percentage of items that a student answers correctly out of the total number of items on a test. Like raw scores, total percent correct scores do not reflect varying degrees of item and test difficulty. They are of limited value in making comparisons.
Note that total percent correct scores are NOT the same as percentile scores. (We discuss percentile scores later in this section.)
What are the advantages of total percent correct scores?
• They are easy to compute.
• They adjust for differing numbers of items.
What are the limitations of total percent correct scores?
Rudner, L. and W. Schafer (2002) What Teachers Need to Know About Assessment. Washington, DC: National Education Association.
From the free on-line version. To order print copies call 800 229-4200
28