The test score information sent home by one Utah school district explained that the standardized test scores were percentiles and that percentiles were the percent of the test items that were answered correctly. This is minus one for that district.
The correct answer was mentioned later in the same brochure, which explained that percentile scores indicate what percent of the test takers were above and below the individual score. A student who gets a score at the 60th percentile scored better than 60 percent of a particular group of people, and 40 percent of the same group scored better than this student. At this point, parents and educators would do well to ask questions about the group that the individual student or school is being compared to.If this information is confusing to educators who sometimes give incorrect information, it must also be confusing to parents, who are now trying to evaluate the results of the statewide testing that has just been completed in Utah.
An educator in another district told teachers to average the percentile scores of the students in each class to determine where the class stood nationally. That's minus one for another district.
Percentile scores cannot be averaged because the difference between two percentile scores that are either very high or very low is greater than the difference between two middle scores. It is confusing, but the difference in items correct between the 95th and the 96th percentile is greater than the difference between the 50th and the 51st percentile.
If you aren't confused yet then consider the statement in the Test Service Bulletin of the Psychological Corp. "A person who is at the 95th percentile is farther away from a person at the 85th percentile in units of test score than a person at the 55th percentile is from one at the 45th percentile." This may not really explain why, but percentile scores cannot be averaged.
It is true that tests are important in evaluating educational progress for individual students and for schools and education systems. It is also true that Utah school districts and the media are making an intense effort to see that the scores are correctly reported to and interpreted by the public. There are, however, other factors as important as test scores in evaluating both students and schools. After all, a test score represents only a small sample of what a student knows. It may have only taken a couple of hours for the test to extract a score that can unfairly label a student or a school for years.
This is not an essay against tests. Tests are careful samples of student behavior. Tests try to standardize the testing conditions so that each student is treated the same. Treating all the same, however, does not necessarily mean that all are treated fairly. These test samples must be put into the context of other student behaviors that are not measured on tests.
Tests don't directly measure tenacity. The tenacious student who works on the same math problem for hours at home and then checks with the teacher before the assignment is handed in so that it can be revised again if necessary may not do well on a timed math test. This student may be very much in demand by a future employer who values tenacity. Tenacity may be as important as the number correct on a timed test.
The student who had learned to use the technology of writing won't have access to the word processor/spell checker/grammar checker/thesaurus when asked to identify errors in a written passage and given a strict time limit on a test. This student may write beautiful prose but stumble on a timed test.
Writing teachers are often faced with the student who writes technically correct prose that is boring. The technician may do well on the standardized test but poorly in real-life writing.
The student who looks for creative solutions to problems may look past the obvious on a test. The tests given in this last statewide assessment don't seem to measure creativity directly.
Perhaps the tests may also have a difficult time getting at aesthetic perceptions and ethical reasoning of students. Yet we would hope that students in our system understand art. We also hope that students would develop an effective strategy for making ethical or moral decisions.
Tests give us some of the answers, but worshiping scores to the point that we teach tests may sacrifice other goals of our educational enterprise. Hopefully we will have students that not only score well on tests but are also tenacious, creative, ethical and have developed an aesthetic sense that understands the humanities and arts.