Understanding Psychological Assessments and Academic Testing
Educational psychological assessment is a formal procedure undertaken individually between a psychologist and a child (or any person). After building up rapport and making the child comfortable, the psychologist will administer a test to the child. There are many tests that can be administered so I will just introduce and explain the most commonly used tests. First though, it is necessary to discuss some general principles of testing.
Tests do not tell anyone anything! They do however yield data and information that then must be intelligently consumed by the qualified assessor. The information gathered from a test must fit into the picture of the whole child’s life, background, family dynamics, learning and schooling history, motivation, health history and a thousand other variables. Anyone who takes the simplistic view that a test provides an answer that can be used to definitively unlock the riddle of a child’s learning problems is seriously mistaken.
What is Intelligence?
Arguments about the nature of human intelligence and what comprise it are centuries old. We have looked at one particular psychological test that is based on a model of intelligence that conceives it as a combination of verbal and non-verbal skills. This model feeds into the common-held understanding that being intelligent means you will ‘be good at reading and maths, you will perform well in comprehensive examinations and you will necessarily perform well in school and get into the university programme of your choice’. This is obviously a narrow model and a dangerous assumption – it’s one that is being challenged vigorously from many fronts today.
IQ is probably the most commonly understood, and at the same time misunderstood, concept about human intelligence. The problem with IQ scores is that they are far too easy to misinterpret and lead quickly to assigning people into the general categories of ’smart’ or ‘limited’. An IQ is nothing more than a mathematically derived formula to quantify various test scores. There has been a lot of research into IQ and it has been discovered that it is a good predictor of one thing, and one thing only – success in school! This is particularly true of the Wechsler model of IQ.
As stated earlier, there are other models, and a popular one is that developed by Howard Gardner*, which is known as Multiple Intelligence theory. Gardner was dissatisfied with the major model of intelligence and the type of educational structures that have resulted from this model. Gardner has stated, “I want my children to understand the world, but not just because the world is fascinating and the human mind is curious. I want them to understand it so that they will be positioned to make it a better place. Knowledge is not the same as morality, but we need to understand [morality] if we are to avoid past mistakes and move in productive directions. An important part of that understanding is knowing who we are and what we can do…”.
Gardner believes that every human being possesses a number of different capacities, all as a result of brain system functioning, which can be called ‘intelligences’. For Gardner, there are at least seven different types of intelligence and our inability and difficulty in recognising them is a result of the way we educate people, relying mostly on words and numbers.
Gardner lists his seven intelligences as:
Linguistic Intelligence – facility with words, reading, speaking.
Logical-Mathematical Intelligence – facility with numbers, logical-
sequential thinking.
Interpersonal Intelligence – being sensitive to and understanding
other people.
Intrapersonal Intelligence – knowing oneself, one’s beliefs, attitudes
and values.
Musical Intelligence – the capacity to respond to music or perform
music.
Bodily-Kinaesthetic Intelligence – ability to move, athletics, dance.
Artistic Intelligence – responding to or creating visual or plastic art.
Since creating his first list of seven intelligences Gardner has continued his research and now believes there are sufficient grounds for adding one other:
Naturalist Intelligence, which he defines as enabling “…human beings to recognize, categorise and draw upon certain features of the environment.”
Gardner continues to investigate the nature of human intelligence and is considering the possibility that he can add to his list the following:
Moral Intelligence – a concern with those rules, behaviours and attitudes that govern the sanctity of life – in particular, the sanctity of human life and, in many cases, the sanctity of any other living creatures and the world they inhabit.
Existential Intelligence – a concern with ‘ultimate issues’.
Spiritual Intelligence – exploring the nature of existence in its multifarious guises.
According to Gardner, every human being has capacities wired into their brains that are manifested in these intelligences to one degree or another. In other words, we are all intelligent – it’s just that we show our intelligence in different ways. I have oversimplified this theory to a great degree but the point I wish to draw to your attention is how much we can underestimate certain children when we conceive of intelligence merely as an IQ figure obtained from a particular test.
This ‘obsession’ with IQ testing can lead to low expectations on the part of educators when the test results are low; low expectations quickly translate into poorer teaching methods, less reinforcement in the classroom for the child and therefore lower performance on the part of the child.
The stakes are high in the IQ race and the winners are more often than not created by those who teach them than by any so-called natural intellectual endowment. Think of it this way; if a person has an IQ of 185 but possesses no ability to understand himself or others, what sort of a life will he lead? Generally speaking the answer will be a life of frustration, wrong choices, unhappiness in love and relationships, and constant disappointment.
Gardner’s theory has its critics and is not universally accepted as an alternative model of human intelligence. But whatever concerns arise about it and from it there is no disputing the fact that Gardner is responsible for bringing to the fore the issue of intelligence and expanding our understanding of what it is.
Intelligent tests require intelligent testers
Put simply, intelligence testing requires intelligent testers. Additionally, tests in themselves are not useful instruments to classify children into special education categories. I have seen far too many children with autistic spectrum disorders given tests of intelligence with the results indicating that their intelligence is significantly impaired – yet anyone working with the child, or any family member, can relate poignant stories of the child’s keen mind and different ways of thinking that clearly demonstrates their intelligence. I am therefore wary of intelligence tests when they are relied upon to provide the sole answer, solution, and source of information used to provide special education services.
The information gathered from a test falls into some general categories. If the purpose of the test is to assess intellectual skills (often referred to as cognitive skills), the information gained should shed light on most of the following:
Verbal skills
Non-verbal skills
Attention and concentration
Visual memory
Auditory memory
Short-term, long-term and immediate recall of visual and auditory information
Social judgement
Social comprehension
Hand-eye coordination skills
Perceptual organisation skills (orientation in space and time)
Abstract reasoning, both verbal and visual
If the purpose of the assessment is to investigate alleged behaviour or emotional problems, then in addition to the above, the information gathered should shed light on:
Frustration tolerance
Impulse control
Anger management
Coping skills
Interpersonal judgement
Stress tolerance
Anxiety issues
Fears and phobias
Unusual thoughts or ideas/beliefs
Knowledge of right from wrong
Social problem-solving skills
Motivation for schooling
Preoccupations and obsessions
Mood (emotions of short duration)
In general, the more information one is able to gather from assessment, the greater the possibility of putting together an intelligent formulation that helps everyone involved in the life of the child to understand the child more comprehensively. The written assessment should clearly provide insight into the referral question and all related information necessary to understand the child. It should be easily read by a teacher or parent, should not contain jargon or scores that can not be understood by all who read it and should outline specific recommendations for educational and support services and strategies necessary to enable the child to benefit from their educational programme.
Wechsler Intelligence Scale for Children
The most common assessment instrument used by psychologists is the Wechsler Intelligence Scale for Children and will therefore be the one we look at. In using this test as an example, you will hopefully get a reasonable picture of how an assessment is carried out. The Wechsler test is, essentially, a test of intelligence. It has been in use for over fifty years and has been revised numerous times to keep it up to date.
The test is divided into two sections with each section containing a number of subtests. The two broad sections of the test are the
Verbal Scale
Performance Scale
Successful completion of any item on any of the Verbal subtests requires a verbal response. On the Performance subtests, the person must do something in response to a question or task. When the entire test has been administered, the assessor calculates what is called a Composite Score, a score that takes into account both sections. Because it is a test of intelligence, the test scores obtained are called IQ scores and you will see the results stated in this format:
Verbal Scale IQ
Performance Scale IQ
Full Scale IQ (the composite score)
The Full Scale score, according to the standard interpretation, indicates the level of a person’s intelligence. A Full Scale score in the range of 90 to 110 is considered average; the person can be said to have average intelligence.
In addition to looking at the Full Scale score, the three scores, (verbal, performance and full scale) can be compared against one another. What is expected in most people is that the three scores will cluster close enough together to indicate that the individual’s verbal and performance skills are evenly developed. When there is a large difference between the two subtest scores (verbal and performance), it may indicate learning problems.
This is as good a time as any to introduce the reader to some of the common terms used in assessment:
Percentile scores
Reading age scores
Standard scores
Children are frequently referred for assessment after reading or maths tests have been administered to the entire class. The most common whole-class tests in use are the Drumcondra tests (reading) and the Micra T test (mathematics). A child’s results in these tests are reported in what are called percentile scores. A percentile score indicates where a child stands in comparison to a sample of children in his or her own age, on a given task. A score at the 50th percentile on the Drumcondra test means that the child is well within the middle range (49 children out of 100 score higher, 49 score lower).
Some tests yield what are called reading age scores. Reading age scores do not yield significant information, their use has been criticised and has been discouraged in the learning support teachers’ written guidelines. A two-year difference in reading age in 5th class may not be terribly significant.
Standard scores are also frequently reported following assessment. The average standard score is 100, which is at the 50th percentile, meaning the child’s score isn’t significantly different in that test than other, same- age children. Standard scores must differ from one another by about fifteen points in order for the difference to be of any real significance. About two-thirds of all children have standard scores on a test that are between 85 and 115, that is, the 16th percentile and the 84th percentile (see table below). Scores in this range are not particularly noteworthy (there are exceptions to this, which will be presented when we explain tests of children’s intelligence).
The following chart will be helpful in translating standard scores, scale scores, standard deviations, and percentile scores into understandable and meaningful information. Standard deviations tell us how much confidence we can place in a given score. Any time a test is administered there will be a certain range of scores obtained that don’t have any significance in the actual test results. In psychological assessment the usual standard deviation of significance is three points or more. So if a child measures 12 points on a test and 11 on another there is no real significance to this difference. Although an oversimplification it is helpful to consider the standard deviation in scores to determine whether or not a strength or weakness is actually present upon assessment.
The most useful scores to interpret for common sense purposes are therefore percentile scores. I recommend you ask for percentile scores when test results are being reported. Most importantly, do not expect reading- or mathematics-age scores to be useful for educational planning or for reviewing the effectiveness of educational interventions. We will refer to percentiles again throughout this section.
Standard Scores
The Verbal Scale, Performance Scale, and Full Scale scores are all Standard Scores. Previously I stated that standard scores all have 100 as their average, with the range of average being from 90 to 110. About two-thirds of all children will score between 85 and 115 on these three scales and scores within this range are not highly significant.
At the risk of getting bogged down in too much information, it’s worth having a more detailed look. For example, let’s take a look at the Verbal Scale. The subtests that are administered are in bold and I have included what they are trying to assess:
Information: factual knowledge, long-term memory, recall.
Similarities: abstract reasoning, verbal categories and concepts.
Arithmetic: attention and concentration, numerical reasoning.
Vocabulary: language development, word knowledge, verbal fluency.
Comprehension: social and practical judgment, common sense.
Digit Span: short-term auditory memory, concentration.
On the Performance Scale, the following subtests are administered (bold) and what they are trying to assess is indicated:
Picture Completion: alertness to detail, visual discrimination.
Coding: visual-motor coordination, speed, and concentration.
Picture Arrangement: planning, logical thinking, social knowledge.
Block Design: spatial analysis, abstract visual problem solving.
Object Assembly: visual analysis and construction of objects.
Symbol Search: visual-motor quickness, concentration, persistence.
Mazes: fine motor coordination, planning, following directions.
An example will help illustrate the fine points of interpreting this test. Suppose Patricia is referred for an educational psychological assessment, having progressed through Stages One and Two.
The Wechsler test is administered and she obtains the following results (this is a crude example for illustrative purposes and the numbers are not meant to be accurate representations of what a real test profile would look like). Individual subtest scores range from a low of one to a high of nineteen. Remember that differences of three points or less between them are not particularly significant. When the difference exceeds three points it may indicate a difficulty with the underlying brain processing tasks that were described above.
Verbal Scale Performance Scale
Information 8 Picture Completion 9
Similarities 3 Coding 10
Arithmetic 9 Picture Arrangement 11
Vocabulary 9 Block Design 2
Comprehension 18 Object Assembly 9
Digit Span 9 Symbol Search 8
Mazes 14
Using the conversion tables available in the Wechsler test manual, the results of these subtests yield the following scale scores:
Verbal Scale IQ 109
Performance Scale IQ 113
Full Scale IQ 110
Patricia is in the average range, right? Looking at the three Scale scores, you would think so. But if we take a closer look at the individual subtest scores, something interesting comes into view. On two subtests that assess abstract thinking (Similarities and Block Design), Patricia’s subtest scores are quite low. Subtest scores have an average of ten and there is little significance in a variation of three. However, Patricia’s score of 2 on Block Design and 3 on Similarities indicates a real weakness in abstract thinking, verbally and non-verbally, despite her average intelligence. This weakness may well indicate learning problems.
I described percentile scores earlier. These scores help us to compare a child’s test results with those of other, same-age children. Let’s see how Patricia compares with other girls her age by looking at the percentile scores that correspond to each of her scores above, as follows:
Verbal Scale Performance Scale
Scale Score percentile Scale percentile
Information 8 25 Picture Compilation 9 37
Similarities 3 1 Coding 10 50
Arithmetic 9 27 Picture Arrangement 11 63
Vocabulary 9 27 Block Design 2 1
Comprehension 18 99 Object Assembly 9 37
Digit Span 9 37 Symbol Search 8 25
Mazes 14 91
The results of all these subtests yield the following scale scores:
Scale Score percentile
Verbal Scale IQ 109 73
Performance Scale IQ 113 81
Full Scale IQ 110 75
Taking a look at the percentile scores tells us more about how Patricia compares to children her own age.
Now, let’s suppose that Patricia was initially referred because she was having considerable difficulty learning to read. I was at pains to point out in the earlier section that the assessor must take into account all the factors that might result in Patricia’s difficulty, before drawing conclusions She may have had health problems which caused her to miss one-third of the school year over each of the past several years; what if her parents were members of the Travelling Community and moved her from school to school five times each year? What if, for the past two years she has had three different teachers, as a result of staff illness, and two of them had no teaching qualification? There may be personal issues (family bereavement etc) that may have relevance. Any of these factors, and more, could be the real cause of Patricia’s reading problems. The assessor will have to take everything into account and put it together in a way that makes sense to all.
What I am saying here is that there are a great many factors which can account for the scores obtained and that it is the responsibility of the examiner to be sure the results are an accurate picture of the child’s intellectual skills and not an artefact of other influences which mask the true skill levels.
It is only possible to make full sense of test scores if they are stated in full in the assessment written report. It is often the case that the psychologist will only report a range of scores, for example, “Verbal IQ: Average Range”, “Performance IQ: Borderline Range” This sort of report writing can raise more questions than answers because sometimes the numbers are at the fringes of a range. For example a score of 90 and a score of 109 are both within the ‘Average’ range but are both at the extreme range, with one Low Average and one High Average. Without stating the exact numbers, it is impossible to get an accurate picture of the child’s level of abilities. I suggest that parents request the complete test data, (the actual numbers themselves) – it will be a useful means to compare results if an assessment is re-administered sometime in the future.
Scale Scores
This brings us to the range of scale scores and what they represent. The Wechsler test is supposed to be a test of intelligence and for these purposes, the three scale scores that are calculated correspond to a range of intelligence ‘category’, from Gifted to Learning Disabled:
Scale Score IQ Intelligence Range/Special Ed Category
130 and above Exceptionally Able/Gifted
90-110 Average (not a special education category)
70-79 Borderline General Learning Disability
50-68 Mild General Learning Disability
35-49 Moderate General Learning Disability
Below 35 Severe/Profound General Learning Disability
You might wonder what happens to those children whose scale scores fall between 80-89. The short answer is that they are not generally eligible for special education services; if their reading of mathematic ability is below the 10th percentile they will be looked after by the learning support teacher. If not, they are deemed to be doing as well as other children and will not receive any specialist assistance.
As has been stated earlier, observations are a helpful source of information but it must be remembered that all observations are subjective (liable to be distorted by hidden bias and differing levels of tolerance for learning differences and differences in behavioural skills). In any assessment the sole reliance upon observation and teacher-made tests is inappropriate. Assessment instruments that generate quantifiable data are a necessary part of the assessment process.
David J. Carey, Psy.D.
297 Beechwood Court
Stillorgan
Dublin, Ireland
http://www.davidjcarey.com