Throughout your child's education, she may take many tests. On entering school, she may take developmental tests, to evaluate her development for placement into the school's programs. While in school, she will take standardized achievement tests given by the school to see how this school and its students compare to other schools and their students, around town and across the country. She may also take intelligence tests, to measure her potential to learn, often given during the process of identifying her as a gifted student.
The results obtained from these tests can be confusing, including standard scores, age or grade equivalent scores, percentiles, stanines, and lots of other numbers. Her scores on one test may not seem to make sense compared to her scores on another test. What do all these numbers mean?
The first test your child may take is a test designed to determine her developmental readiness, often for kindergarten or first grade. These tests consist of a variety of activities, mostly verbal and hands-on, such as naming objects, throwing a ball, drawing and copying figures, and more. While these tests may be valuable to identify developmental delays, they are not designed to identify children who are developmentally advanced, and therefore are not good indicators of gifted children. In fact, some developmental tests have a standard score ceiling as low as 135; hardly a indicator of a gifted child at all.
Examples of developmental tests include the DIAL-R, Gesell Developmental and the Vineland social adaptive scale.
Achievement tests measure a student's current levels of achievement - what does this student know? Achievement tests usually cover a variety of content areas. This kind of test includes sections (sub-tests) to test a student's math and reading levels, and may include sub-tests on science, social studies, humanities, and more specific areas such as spelling, grammar, problem solving and math reasoning. Different achievement tests may be given either individually or in a group.
Examples of group achievement tests include the Common Test of Basic Skills (CTBS) and the Stanford Achievement Test (SAT), as well as a variety of U.S. state tests such as Michigan's MEAP and Texas' TAAS. Examples of individual achievement tests include the Weschler Individual Achievement Test (WIAT) and the Woodcock-Johnson (WJ).
Achievement test results are usually reported in percentiles or grade equivalences, i.e. Jon (a mythical 3rd grader) received grade equivalent scores of 5.1 for math and 4.3 for reading. While these numbers look good, the parents are often left to wonder "What does this really mean?" Does this mean that Jon is ready for 4th grade reading or 5th grade math? Not necessarily. A grade equivalent score can be misleading, and its value may vary depending on whether the test was an individual or group achievement test, as well as the norm group, and breadth and number of test items.
On group achievement tests, a score of 5.1 means that Jon did as well on this, the 3rd grade test, as an average student at the beginning of 5th grade would have done, had the 5th grader taken the same 3rd grade test. But 5th graders don't take the 3rd grade test, so this score does not mean that Jon would score as well as the average 5th grader on the 5th grade test. He would have to take the 5th grade test before you could draw that conclusion. This is known as "out of level" testing.
On individual achievement tests, the tester may have a wider range of materials to test the child on, so a grade equivalent score of 5.1 may indicate that Jon completed up through the 5th grade portion of the test, and is in fact ready for 5th grade math. An individual acheivement test may, however, have fewer items per concept. And again, the accuracy of the results depends on the content of the test and the norm group used for comparison.
Perhaps Jon's scores were reported as percentiles. Jon scored in the 99th percentile for math and the 85th percentile for reading. What do these numbers represent? If these are national scores, it means that Jon scored better in math than 99 percent of the students in his grade in the test norming group across the country. If these are local percentile scores, the comparison group might be his school district, or his local school. The percentile does not mean that Jon got 99% of the test correct. If the test was relatively difficult, he might have answered fewer than 99% correct, but still did better than 99% of the students. On the other hand, if the test was too easy, Jon might have gotten 100% correct (in reading, for example) and still scored in the 85th percentile locally, because only 85 percent of students from his grade and his school scored worse than Jon - the other 15 percent got the test all right, just as Jon did!
While it is unlikely that Jon would get all the questions correct and yet score in the 85th percentile nationally, due to the large sample size and national standardization of scores, it is more plausible that his perfect score might be reported as a lower percentile in the local sample.
The only way to know how Jon did on the test in terms of number of questions right and wrong, is to look at the raw scores. Often this is a single number, 37, or perhaps two numbers 37 / 40. The single number tells you how many questions Jon got right, but this doesn't mean much without knowing how many questions there were on the test. The two number raw score tells you that Jon got 37 questions right out of 40. Raw test scores should be reviewed if Jon's scores seem inconsistent with national and local norms, with other tests he has taken, or with his teachers or parents knowledge of his work and level. Raw scores may also be useful for a tester to compare Jon's score to scores of students in other grade levels.
It can also be useful to see how the percentile or grade equivalent scores change with one or two raw score points. The fewer questions per concept there are on an achievement test may cause significant variation in the precentile or grade equivalent scores as a result of single point variation in the student's raw test score.
Achievement test scores may also include standard scores. These scores look a lot like IQ scores, but should not be confused with actual IQ scores, which measure different constructs. Jon may have received a standard score of 145 in math, and 128 in reading. Does that mean he is gifted? He isn't gifted? More than likely, it means that Jon may be gifted, and some other measure should be used to determine his IQ.
Achievement tests cover a fixed body of knowledge. If Jon's school did not teach the specific facts included on the test, Jon might score poorly even though he is very bright. On the other hand, Jon's standard score of 145 in math indicates a highly achieving child, perhaps due to the fact he was exposed to the material, and demonstrated an exceptionally high level of mastery.
Intelligence tests are designed to measure a student's potential - how well does this student think and learn? Intelligence tests are often used to determine a student's placement in a school's gifted program. This kind of test often includes sub-tests to test a student's thinking and learning in areas that are visual, verbal, and hands-on. Intelligence tests may be given in a group or individual situation, however the individual types are generally more accurate. Group tests are often given as a screening tool, with more information gained through individual testing by a psychologist of those students who qualify based on the group test.
Many intelligence tests, despite their intention to measure potential, do so by testing achievement. A good example of this is the vocabulary section in the Weschler series of IQ tests. A child from a background offering limited exposure to these words (disadvantaged home or different culture) may not do as well on this section, no matter how bright.
Examples of group intelligence tests include the Otis Lennon Mental Abilities Test and the Ravens Progressive Matrices. Examples of individual intelligence tests include the Weschler tests: WPPSI for pre-school children, WISC for school aged children, and WAIS for adults, and the Stanford-Binet (available in two forms, IV and LM, discussed later).
Results of intelligence tests are usually reported in a single standard score, an Intelligence Quotient or IQ. IQ scores are standardized to have an average score of 100, standard deviation of 15, so that 95% of the tested population score with two standard deviations of the norm, i.e. from IQ 70 to 130. Gifted children are generally identified as having IQ scores of 130 and above, although the cutoff for gifted may vary by state and school district.
The commonly used Weschler family of tests (WPPSI, WISC, WAIS) also have 10 sub-test scores, that are combined into two separate standard scores: performance and verbal. These sub-test scores are scaled from 1 - 19, with 10 being an average score. It is important to review the sub-test scores; they can give valuable insight into the relative strengths and weaknesses of the child. One or two unusually low sub-test scores may indicate a potential learning difference, which could lower a child's overall score and make a gifted child look 'average.' Also look for ceiling effects: it is said that two or more sub-test scores in the ceiling range (17 - 19) indicate that the child reached the ceiling of the test, and should be re-tested using a test designed to differentiate highly gifted children. At this time, the only such test available is the Stanford-Binet L-M. ["Don't Throw Away the Old Binet", Silverman and Kearney, 1992]
Now you understand intelligence tests and their scoring, right? But there's a catch. Each test has a ceiling, a top score that its possible to achieve on that test. You might think since "gifted" doesn't start until IQ 130, the tests given to determine IQ would have ceilings much higher than that. Unfortunately, you would be wrong. Most IQ tests have ceilings of 155-160 or even less. Some tests given to our young children, supposedly resulting in an IQ score, have ceilings as low as 135. What does this mean? It means if Barb and Bill both took a test that has a ceiling of 135, and both scored between 130 and 135, you might be fooled into thinking that both Barb and Bill are moderately gifted. If both later take a test with a much higher ceiling, Barb might again score around 135, proving that her original score was fairly accurate for her. Bill might score much higher, perhaps achieving a score of 165 or higher (only possible on the Stanford-Binet form L-M, all other tests have lower ceiling scores). Although his initial score made Bill look only moderately gifted, his score was depressed by the ceiling of the test, and in fact Bill is exceptionally gifted.
What's the difference? Gifted children tend to think differently and learn more quickly than their peers. Compare a gifted child (IQ = 130) to an average child (IQ=100) you will see the difference: the gifted child learns quicker, thinks deeper, and draws conclusions more easily. Compare that gifted child (IQ=130) to the highly gifted child (IQ=160). Again, you will see the difference, in many of the same ways. Now compare the highly gifted child to the normal child, and you face a chasm that by the end of elementary school may place these two children as much as 5 years apart in mental age.
There's another way to look at it. The difference between the exceptionally gifted and the average child is the same as the difference between the average child and the mentally handicapped child of IQ 40. That's a big difference!
Ceiling effects may be particularly misleading in cases where the child has widely disparate abilities. A relatively low score on a subtest in a child's weak area may not be properly offset by clipped ceiling scores in his strong areas. This could make a gifted child look 'normal', and decrease his chances of being appropriately served by his education.
|180 and above||Profoundly Gifted|
For more information on tests and testing, visit Understanding Tests and Measurements for the Parent and Advocate. Although written from an LD (Learning Differences) perspective, this page offers information and insight into the issues surrounding testing for any reason.
This page last updated on 11/15/2004 by Carolyn
K. Please submit all questions and answers directly to Carolyn. Thank you.
Copyright ©1998 by Carolyn K.
All Rights Reserved