Why Should I Have My Child Tested?
by Carolyn K., director, Hoagies' Gifted Education Page
There are two types of testing recommended in the assessment of gifted children: IQ or ability testing, and achievement testing. Before you can consider this question for your child, you should be familiar what these tests are, and are not.
Ability or IQ tests purport to show how smart a person is, how well they think, or what their capability is for learning. There are group and individual ability tests. Group tests may be administered in group or individual situations, but are normed (scored) for a very specific group of people, for example, 2nd graders in the spring of the school year. While these group ability tests have questions for other levels, the vast majority of the questions are exclusively for children who are commonly in the group being tested. This means, for gifted children, there are not a great number of questions to differentiate the gifted child form the average child, nor many, if any, questions to differentiate the various levels of gifted children.
In some cases, while the correlation between group tests and individual IQ tests is quite high for average scores, that correlation almost disappears for gifted scores. This means that an average child will score very similarly on a group IQ test and an individual IQ test, but a gifted child may not score similarly at all. There are small studies showing that group tests may even result in a negative correlation for some gifted children. This means that the more gifted the child, the lower the group ability test score! Read "Investigations of the Otis-Lennon School Ability Test to Predict WISC-R Full Scale IQ for Referred Children" by Anna H. Avant and Marcia R. O'Neal, University of Alabama, Nov. 1986, ED286883, for more details on this phenomenon.
Group ability tests include such tests as the Cognitive Abilities Test (CogAT), the Otis Lennon School-age Ability Tests (OLSAT).
Individual IQ tests tend to be more accurate for most of the population. These tests are often given to selected children, based on recommendation from a teacher, parent, or the child's scores on a group ability or achievement test. These are the tests commonly used to "identify" gifted children, for participation in school programs for the gifted.
But these tests are, in general, not designed to differentiate gifted children. In some cases, a specific test may not have been designed for use with the gifted; in other cases, getting a large enough "sample" of the various levels, at various ages, of gifted children is too prohibitive. Every test has a ceiling, a highest possible score. These scores vary by test, and each test is usually divided into subtests, each of which also have a ceiling. If a child "hits the ceiling" on several subtests, his overall score may be lowered by his lower subtest scores, resulting in an artificially depressed overall score. This is particularly true of twice exceptional, or gifted / learning disabled children.
Ceilings vary from test to test, and within a test, from subtest to subtest. They also vary based on the child's age. The WPPSI, for example, has lower subtest ceilings close to its top age of 7 years 0 months. It is suggested by the publisher that gifted children over 6 years 0 months be tested instead on its sister test, the WISC. The WISC, too, has ceiling problems near its age boundary of 17 years 0 months. Gifted children over age 12 are often recommended to wait and be tested on the WAIS (adult form) when they are old enough, at 16 years 0 months.
Test Scoring Terms
Before further discussion of testing, it is important to understand what kind of results a test provides. There are many different numbers reported from a single test. A Standard Score (SS) represents a comparison of the child to the test population, where the average score is 100 and there is a standard deviation, often 15 points. This means that the average child scores 100, and that 95% of the population scores between 70 and 130, or within two standard deviations of the norm. Scores above 130 would be considered gifted (if they are overall intelligence scores) or gifted in that subject, for individual subjects.
Tests also offer age-equivalent and grade-equivalent scores. These scores compare the child to the age and grade of average children who received similar scores. These scores do not mean that your child should jump into that grade level class, but scores more than a grade above the actual grade often do mean that the child needs something far more than is offered is his current grade level classroom.
If these age-equivalent and grade-equivalent scores are the results of a grade level achievement test, then they must be taken with a grain of salt. Say your child received a grade-equivalent score of grade 6.7. That means that your grade 2 child scored as well on this test as a child of grade 6.7 would have scored, had he taken the same grade 2 achievement test! It does not tell how your child would score on a grade 6 achievement test. Group achievement test scores are rarely valuable if they are more than 1-2 grade levels above (or below) the grade level the test was designed for.
Percentile scores compare your child to the rest of the norming population of a test. A 95th percentile score means that your child scored better than 95 percent of the norming population (or your local students, if it is a local percentile) taking the test. But... say 8 percent of the students got a perfect score on this section, or this test. Their percentile would be... 92nd percentile. That is, they scored better than 92 percent of the students taking the test. Rather misleading...
Some tests have a larger Standard Error of Measure (SEM) than others. This means that the confidence interval associated with a specific score may be quite large. In order to be confident of identifying those children with IQs at 130+, for example, one must consider all results above (130 - SEM). While on an individual IQ tests, this SEM value is commonly 4 to 7, on some group ability tests the SEM is as much as 15 or more. Given SEM = 15, one would have to select all students who score 115+ to be confident of getting all those who will be later identified as gifted kids 130+.
Achievement tests are tests designed to see what a student has already learned. Again, there are group and individual achievement tests. Group tests are usually given to a large population at once, but the same group test may be administered to a single child individually. Even administered individually, it is still a group test, subject to the advantages and disadvantages of group achievement tests.
Group achievement tests are written for a single grade level, or part of a grade level (5th grade, or spring of 2nd grade). They are given to the group, and the results are compared to a group of children. These tests have little "headroom," few questions that are more than a single grade higher or lower than the grade level of the test. Scores of 95th percentile or better do not give much information, except that the child is above his grade level in the subject (or overall). To get a more accurate idea of where a gifted child truly falls compared to his grade level, he should take an individual achievement test, or an out of level achievement test.
And group tests, whether group intelligence or group achievement tests, have another problem for some of our gifted kids: ambiguity. Since the questions on a group test are all in written format for easy group administration, our kids often see more than one answer to a question. For example, the question "A girl had 49 of something and got 7 more, what would you use to solve the question of how many does she have now? - addition - subtraction - division - multiplication. One child pointed out that it could be multiplication because 49 is 7*7, and if you did 7*8 you would get the answer. True enough! Luckily, she also decided that the test probably didn't want that answer and answered addition. But both answers could be justified, and on a group test, there is no tester interaction to prompt "why did you pick that?" or "is there another answer?" Her first answer would be marked just plain wrong. (Thanks to Amy and her daughter for this example.)
Individual achievement tests, such as the Wechsler Individual Achievement Test (WIAT), Peabody (PIAT), Kaufman (K-TEA), and Woodcock Johnson (WJ-III achievement) have fewer questions at each grade level, but cover many grade levels. Most of these tests reach up to end of high school norms - grade level 12.9. This is equivalent in math, for example, to about Algebra I level. Many of our gifted kids reach that level in elementary school! The WJ-III achievement test measures through grade level 18.9. For this reason, the WJ-III is recommended as an individual achievement test for upper elementary and older gifted students, who might reach the ceiling on a test that ends at high school level.
These tests are administered individually, usually by a psychologist or counselor, depending on the test and your state law. The results include age and grade equivalent scores, and percentiles, comparing your child to the average child at the specified grade levels. Often confusing, it is best to remember that the "expected" achievement for a high school senior, in math for example, is about the end of Algebra I. These grade levels are not grade levels for college bound seniors, and should be considered appropriately.
Out of Level Achievement Testing
Another way to use group achievement tests is for out of level (Talent Search) achievement testing. This is most commonly the SAT-I or ACT given at a younger-than-normal age (usually 7th grade). On the east / west coasts, the SAT-I is most commonly used; in the midsection of the country, the ACT is more common. Younger students are joining this trend, using tests such as the PLUS, SCAT, and Explore. Most of this out of level testing is conducted through the major and state Talent Search programs.
The advantage to out of level testing is 1) qualification for various talent search programs, including academic distance education, summer programs, and more, 2) indication to the school that the child needs something different in terms of late middle and high school program, through 3) relative comparison of child to school's graduating seniors, 4) better parental understanding of the child's difficulties in school (comparing the child's achievement levels to other students achievement levels, etc.).
Out of level testing can differentiate those scoring in the top percentiles on grade level achievement tests. Two students, both scoring in the 95-99th percentile on their grade level tests, may then score very differently on an Explore test 3 grade levels up. In fact, research has shown that taking the top 5 percent of students, and giving them out of level testing, results in another normal distribution of scores among those students, with the top students still scoring at the top end of the higher test, and students distributed across the entire score spectrum of the test. (See Discovering Highly Gifted Students by Jan B. Hansen for a table of scores.)
Curriculum Based Assessment (CBA)
No discussion of testing would be complete without mention of CBA. Curriculum based assessment (CBA) is a different kind of achievement assessment. This is arguably the most accurate means of comparing a child's achievement to the curriculum he is to be exposed to. In curriculum based assessment, the actual curriculum materials are used to assess the child's progress. This can be done by giving the child the text or teacher's comprehensive final test, or by giving the child the end-of-chapter test for each chapter in the text.
While this type of assessment sounds like the perfect way to prove whether a gifted child knows the material to be covered, there can be difficulties with curriculum based assessment. Terminology can be problematic. Consider the child who attended 3 schools in 3 years: in the first school, the teacher used the term "borrowing" when she answered the kindergarten student's questions about complicated subtraction; in the second school, the text introduced the term "regrouping" to indicate the same process in first grade. In the third school, the child was completing a chapter by chapter CBA, and was asked to "complete the following problems using renaming. When the child questioned the teacher, she was told that the teacher couldn't answer any questions. The child left those questions blank; she had no idea what renaming meant. The assessment was halted since the child did not pass that chapter, and the child returned to the grade level classroom. She learned nothing new in math all year.. except that the term renaming meant the very same thing as borrowing and regrouping.
Other difficulties with CBA include deciding what score is required to prove competence. Students are passed from grade to grade with scores of 60% or better, but gifted students trying to skip levels are required 85%, 90%, 95% or even better to pass over material they already know. While it is advisable to make sure the student is comfortable with the material, expectations higher than 85% are generally ill-advised, and push students into unhealthy perfectionism. In grades and subjects where chapter or text tests are not available, determining what to include in CBA is difficult at best.
When a child is suspected of being more than a little above the current grade's curriculum, CBA must be adjusted to begin at an appropriate level. Sitting the child in a room, day after day, week after week, completing chapter test after chapter test, beginning with the current grade level and slowly moving up through grades can be torture for the gifted child. Choosing an appropriate starting level for the CBA is imperative.
All achievement and intelligence tests have ceilings, highest possible scores. Most tests have subtests, and each subtest has a ceiling, sometimes the same as the other subtests, sometimes different, and in gifted children these subtest ceilings contribute to lowering the overall test score if a child is not evenly or "globally" gifted. Some tests, though they have fairly high ceilings, were not designed to test all the way up to those ceilings - they are only there for subtest head room for the lower scores. For example, the Wechsler intelligence tests were not designed to differentiate scores above 130 (see GT-World's FAQ on The Wechsler Intelligence Scale for Children (WISC) for a quote from Dr. Wechsler).
So how do you know if your child has reached a ceiling on a test? On the Wechsler tests (WPPSI, WISC or WAIS), a ceiling is a score of 18-19 (99th percentile) on a subtest, and the whole test may be an underestimate if 2 or more subtest ceilings are reached.
But often, parents are told that the child didn't hit a ceiling, because she didn't answer all the questions correctly, or he didn't get to the hardest questions the test had to offer. Are these not ceilings? There are two ways to identify a ceiling. First, if the child answered any more questions correctly, could she score any higher on the subtest? If the answer is no, she could not score any higher, then it is a ceiling. Second, was the termination criteria for the subtest reached? Tests have specific requirements for stopping a subtest. Commonly, the student must get less than x questions correct of the last y questions asked. If the child did not reach the termination criteria, he hit the ceiling. Either one of these events indicate a test ceiling was reached.
The next logical question is, even if my child hit the ceiling on two or three (or more) subtests, but that can't mean she hit the ceiling on the whole test, can it? Well... what it does mean is that you don't know. Ceilings are just that - you don't know how tall the building is, how many more floors it might have - once you hit the ceiling. You're just staring at the ceiling. Is it the inside of the roof? Or is there one more floor, or ten more floors? You don't know.
It's the same for tests. Once the child reaches the ceiling, there's no telling how much higher she might have gone, if there was room. Or not. The overall test score is calculated from all the subtests, so if she hit a ceiling on several subtests, and those ceilings really were her stopping point, perhaps the overall score is correct. But what if those ceilings should really have been higher... or much higher? Then the low ceilings will average in with her other subtest scores, and lower - or drastically lower - the overall score (called full scale score, GIA, etc. - varies by test). This is particularly likely in a child with uneven giftedness, a Highly Gifted child, or a Twice Exceptional child. Once she hits 2 or more subtest ceilings, all you know for certain is that her overall score is a "floor" - a least estimate of her ability.
Read What is Highly Gifted? Exceptionally Gifted? Profoundly Gifted? And What Does It Mean? for more on the specific ceilings on some of the most commonly administered intelligence tests.
Supplemental Intelligence Testing
Today there are additional scores available on the WISC, including the GAI and Extended scores, that make the use of the old SB L-M obsolete. Few if any testers in the U.S. still use this 40+ year-old measure.
Supplemental Scores on the WISC-IV
Since the WISC-IV was published, two supplemental scores have been introduced. Read What Do the Tests Tell Us? for additional details on the General Ability Index (GIA) and Extended Norms for the WISC-IV. Psychologists should know about these scores, and provide them if appropriate. But in reality, many psychologists are not aware that the test publishers have created these new scores, and will need more information to provide these scores in your child's test results. Publisher's bulletins Harcourt Assessment WISC-IV Technical Report #4 General Ability Index and Harcourt Assessment WISC-IV Technical Report #7 WISC–IV Extended Norms in conjunction with the WISC-IV scoring manual provide the psychologist all they need to know.
When to Test?
Let's start with the question of when to test. When is the gifted child too young / too old for testing? The best time to test for giftedness is when a question needs to be answered. School selection, educational placement decisions, early kindergarten, these are the kinds of questions that need an answer grounded in a comprehensive assessment, including testing. The recommended ages to begin to answer these questions, and therefore the recommended ages to test for giftedness are from ages 5 to 8. Note that there is a ceiling effect for gifted 5-year-olds on the WPPSI (Wechsler Preschool and Primary Scale of Intelligence); if possible, wait to test the gifted child on the WISC (Wechsler Intelligence Scale for Children) after the child's sixth birthday.
By age 8 most gifted children already need to have accommodations in place for their appropriate education. Negative effects, such as underachievement and withdrawal, can start in the early elementary years in gifted children. Twice exceptional gifted children's learning disabilities can seriously affect test results by age 8. Early identification is key to proper social and academic placement for the gifted child.
Schools, on the other hand, suggest testing later, often not until 3rd grade or age 8 or 9. Research shows that for the average child, IQ test scores are reliable around age 8. Observations of gifted children (real research is needed) indicate that reliability in IQ scores is obtained much younger in the gifted population. Not coincidently, most schools recommend gifted testing for the year their gifted program begins.
Meanwhile, there is truth to the oft-heard statement that "kids level out by 3rd grade." No, gifted kids don't level out, they continue to learn faster, and gain quicker, getting further ahead of their age-peers. But... Those kids who are "hot-housed," attend the most academic pre-school, are taught at home, flash carded (no, not those gifted parents who's kids *demand* flashcards, the other kind), and generally reach school already reading some sight words, perhaps even reading, doing some math... those kids often do fall back to "average" by 3rd grade, when the other kids have also learned to read.
The problem is, some gifted kids also appear to have fallen back, thanks to their development of social-self esteem - they realize that they are different from the rest of the kids, and they conclude that perhaps different is bad, or they find different to be less socially acceptable in school, so they go into hiding. For some (more often girls) this is a permanent condition; others can't take the hiding any more at some point in later schooling, and explode in frustration.
Recommended gifted testing includes an individual IQ and achievement test. What's the harm? Well, most gifted kids actually *enjoy* testing, thriving on the challenge - the hardest part for them is to convince them to give serious answers to the silly, easy, questions asked in the initial parts of the tests! Of course, if someone pressures the child, you *have* to do well on this, etc., then there is always the possibility of harm, and mis-measurement, but that's a different story - the harm isn't in the tests, it is in the folks applying the pressure.
But which is the best test choice, to learn the most information? Kathi Kearney offers insight into IQ testing:
She continues with this cogent explanation:
See An Inventory of Tests for information on each test that might be suggested.
The information gained from such testing can be valuable. It is useful to parents to know where your child stands before (or while) the teachers are saying "he's socially immature" or "she's not all that smart." Behavior problems in preschool or early elementary school are often a trigger that sends parents of gifted kids to testing. This is a difficult period for the gifted child, and the more gifted, it seems the more difficult (reported by Leta Hollingworth 80 years ago, and parents on this list, this week...).
The achievement test can combine with the IQ test to offer more understanding to the family, and a more complete picture of the child, thus explaining his frustration, or lack of fit, in his current situation. Sometimes parents do test before problems occur. Parents who've spoken to other parents with older kids may test because they know that ceilings will come into play, and they want an accurate picture. Other parents may test because they just want to know where the child fits, and how to best select their educational path.
The combination of IQ and achievement test can also show potential LDs, though there needs to be a significant difference (and the child needs to be school age, for most achievement tests) to show most hidden LDs. Young children are often tested as part of early intervention programs when difficulties are suspected. Why is such early intervention often frowned on for gifted children?
The answer to the question "Why Test?" is the same for the gifted child as for any other child: you should test to answer a question. Tests can provide detailed information about the child's learning needs to parents and teachers, including gifted identification for educational planning and gifted program participation. Tests can also offer information for early intervention of learning differences, and to facilitate an appropriate education. And you should test when you need the answer to any of these questions.
Can you test too much?
Is there such a thing as too much testing? Of course. While it is recommended to give an individual IQ test after initial testing done with a screening measure or group ability test, there is rarely a need for multiple individual IQ tests for a gifted child. The exception might be if the first individual IQ test test did not reflect the child accurately. Sometimes retesting may be done around age 8 or 9 if the child was first tested at a very young age. Taking the same test, IQ or achievement, within 12 to 24 months is not recommended, and in many cases, professionally unethical for a psychologist to administer.
Achievement tests might be used more often (though no more than annually), to determine academic placement in educational planning. At younger ages, individual achievement tests may be used. As the child approaches upper elementary school, ceilings on most of these tests may come into effect, and Curriculum Based Assessments (CBA) may be better for determining the child's placement within the school's curriculum. However some schools do not have CBAs available, nor resources to create them. In such cases, continued reliance on individual achievement tests may be needed.
Testing the gifted child, like any other psycho-educational decision, is a complex decision. Consider why you are testing, what tests will be given, and what answers you are looking for from the testing. Know what you want to learn from the testing. And make an informed decision.