What Do Tests Really Tell?
Overheard at a parent-teacher night:
"My Suzi, a second grader, scored a grade equivalent of 4.7 in reading; she could be advanced to reading material at a fourth grade, seventh month level!"
"Jack's composite percentile rank dropped from 67 to 61 points from one year to the next, and I am really concerned."
"A fifth grader who scores an 8.2 GE in reading and a 7.3 in math is definitely better in reading than in math."
"Well, I think the value of a curriculum or the effectiveness of a teacher can be judged solely by standardized test results."
"Achievement tests measure almost all of the important skills and objectives that you are trying to teach."
The quotations represent misunderstandings that people often have about what standardized tests can tell them about their children, the curriculum, and even teaching methods.
The first statement reveals one of the most common misinterpretations: a grade equivalent of 4.7 in reading means the second-grade student performed as well as a fourth grader in reading. But the GE really means that Suzi can read her second-grade material as well as a student in fourth grade can read the same second-grade material.
The second statement shows how percentiles are often considered definitive and specific, when in fact they are approximate. For most students, a five- to seven-point variance from year to year is probably immaterial. This point spread (or band) will vary from subtest to subtest.
A related misconception results from the confusion between percentiles reported on achievement test results and percentages reported for teacher-made tests. A student ranking in the 65th percentile did considerably above average on an achievement test, but a 65 percent on a classroom test is usually a D.
The third statement addresses a superficial understanding: the assumption that GE scores in subtests can be compared across the board. In comparison, the two test scores may represent equally superior performances because the range of GEs for fifth graders in this case is generally greater in reading than in mathematics. And too, patterns of growth may vary from subject to subject and from year to year.
The fourth statement rests on one of the most dangerous misconceptions about standardized tests. Many factors must be considered when determining a student's achievements, a teacher's impact, a curriculum's effectiveness, or a method's worth. To base educational decisions solely on test scores is the educational equivalent of deciding to buy a house sight unseen because the address indicates a good neighborhood.
The fifth statement is an extension of the fourth. It shows how some people expect a test to tell them everything they want to find outeither about a student's progress and potential or about the teacher's work and the curriculum. Parents and teachers need to use their own observation and discernment to evaluate a student's progress, taking into account factors such as test-taking skills, maturity, and so on, and recognizing achievements in untested fields such as art, music, philosophical values, and sophisticated thinking skills.
One test, or even a series of tests, should not be used alone to determine a student's grade placement or course grade or a curriculum's value.
Standardized achievement tests can show the following:
- students' knowledge of facts, skills, and concepts common to the grade tested
- year-to-year academic development (more accurate over extended periods)
- student academic strengths and weaknesses (in individuals to a limited degree)
- students' higher-order thinking skills, although in a limited way
- where investigation into such specifics as methods and effectiveness of curriculum might be needed
A good test is a good tool for helping assess a student's progress. But it is only a tool, an indicationnot the final word. Most people would not want to buy a house based on an address, and informed parents will not want to make choices about education based only on test scores.
Explaining and Applying Test Information
When standardized test scores are returned, they create a certain amount of excitement. It is good to get some consensus in understanding the purpose and interpretation of tests before the results come back.
But even if that preemptive help cannot be arranged, explanations and questions from teachers and administrators along the way can clarify the results and make them more useful. For example, if a student scores in the 38th percentile in math problem-solving skills at the sixth-grade level, that score by no means indicates that "the student is just bad in math and will always have to work extra hard in that area."
Rather, you, the teacher, can explain that it means that the student on that particular day scored in approximately the 38th percentile in math problem-solving skills among all sixth graders who had taken the test. Then you should pose a series of questions, such as the following, that will help put the score into perspective for both you and the student and his parents.
- What should we expect from this student based on his quantitative thinking skills?
- Should we emphasize visualization and understanding more, rather than focusing on drill and memorization?
- Does the student get enough practice in problem solving?
Despite what the news media and some politicians would have us think, test scores cannot tell us whether all is well (or not well) with a student, a school, materials, or teaching methods. Test scores do not indicate whether children are learning to think from God's point of view or whether they enjoy what they are doing and are starting on the path of lifelong learning. And tests cannot tell the overall story of how a child's experience in school is preparing him for life. Such resultsthe results that matter mostmust be evaluated by parents and teachers along the way, using a multitude of tools of which a standardized test score is only one.
For a thorough review of tests and their uses, James Deuink's The Proper Use of Standardized Tests is a valuable resource (available through BJU Press).