How Norm Groups Are Used in Psychological Testing
When designing a test of something—for instance, academic ability or signs of depression—it’s important for the people making the test to understand the group that they are testing. They also need to identify what is considered normal within that group. Take, for example, the SAT (originally named the Scholastic Aptitude Test and later the SAT Reasoning Test). Published by the College Board, the standardized test measures academic potential. The SAT is taken by high school juniors and seniors throughout the United States each year. Therefore, the normative group for the SAT is a randomized, cross-cultural group of American junior and senior high school students who accurately reflect the diversity (and thus, the average) of that group of test-takers. A psychology example could be a test intended to diagnose depression in American children between the ages of five and 10 years old. In this test, the normative group would be a sample of five- to 10-year-olds from various demographic groups within the United States.
How Norm-Referenced Tests Are Assessed
Norm-referenced tests are assessed differently compared to criterion-referenced tests. Criterion-referenced tests are the typical test format you’d find in school: all questions have right answers and wrong answers, and scores are graded out of a perfect score. By contrast, it’s not possible to “pass” or “fail” a norm-referenced test. Rather, it will give results based on performance compared to a normative group. One of the main types of norm-referenced tests is an Intelligence Quotient or IQ test. Intelligence test scores typically follow a normal distribution, which is a bell-shaped curve where the majority of scores lie near or around the average score. For example, the majority of scores on the Wechsler Adult Intelligence Scale-Fourth Edition (WAIS-IV) tend to lie between plus 15 or minus 15 points from the average score of 100. This means that approximately 68% of people who take the WAIS-IV test will score somewhere between 85 and 115.
Percentiles as an Expression of Performance
Norm-referenced tests can also be presented as a percentile. The percentiles are based on a bell curve with the “norm” falling in the middle. The percentile range is demarcated as deviations (either above or below) from the norm. If you’ve taken a standardized test such as the SAT, you may have noticed that you got both a score that was a number based on the total number of points you could have received, as well as a percentile that reflected how you did in relation to other test takers. The farther away from the norm you are, the further away from the 50th percentile your score will be. So, for instance, an SAT score in the 99th percentile means you scored better than 99% of the other test-takers.
