The Belief Systems of Psychologists and Clients about Intelligence and Intelligence Tests

 This is a transcription of a lecture I gave a few years ago at a seminar in a small town near Tel Aviv. I have divided the lecture into four posts.

The lecture discusses theories of intelligence and presents PASS theory. After a fairly long and somewhat philosophical introduction, the theory and its applications are presented. The lecture also has a slightly unusual ending, in which the boundaries of intelligence are expanded.

Here is the first part:

Why Do We Need Theories of Intelligence?

Two Creation Stories

In the Book of Genesis, we encounter two creation stories. Genesis 1 presents a sequential, serial account of creation, in which something new is created on each day. On the first day, God creates light and separates light from darkness. On the second day, He separates the heavens from the waters. On the third day, dry land and vegetation are created. On the fourth day, the greater light—the sun—and the lesser light—the moon—are created. On the fifth day, the sea creatures and birds are created, and on the sixth day, the land animals and human beings are created.

Genesis 2 presents a more holistic, simultaneous creation story. Nevertheless, this account also contains a certain broad sequence: first, God separates the heavens from the waters; then the human being is created; afterward, the Garden of Eden is created, with its vegetation and the rivers flowing from it; the animals are then created; and finally, Eve is created from the rib.

These two stories present two theories that attempt to answer the question of how the world was created. Each theory has different implications, for example, for the place of humanity and the place of women in the world. According to the theory presented in Chapter 1, human beings are created last, and woman is created together with man: “male and female He created them.” According to the theory presented in Chapter 2, man is created first and has a central role as a participant in the process of creation, in the sense that he gives names to all the animals. Eve, by contrast, is created last, from the rib, in a manner that may imply that she is not equal to man.

During the summer, I read a fascinating book with an impressive title: The Big Picture: On the Origins of Life, Meaning, and the Universe Itself. The book was written by the physicist Sean Carroll, who, in my view, is no less a philosopher than he is a physicist.

Sean Carroll presents an approach that he calls poetic naturalism. According to this approach, there is only one way in which the world is constructed, but there are many ways of talking about the world. The different theories of physics are different stories about the world, operating at different levels of generality. Each theory has its own vocabulary and its own rules, which together create a coherent system of beliefs about the subject addressed by the theory.

Just as the planets in the solar system hold one another in place through gravity, the beliefs within a theoretical system support one another. Each belief depends on and reinforces the others, creating a coherent theoretical system.

When I read this, I wondered: What is our system of beliefs, as psychologists, regarding diagnostic assessment? What is our clients’ system of beliefs regarding assessment?

Because we are discussing theories of intelligence, we will narrow the question and speak specifically about our beliefs, as psychologists, and our clients’ beliefs about intelligence—and, more specifically, about intelligence tests.

What did we think in the past—say, ten years ago—and what do we think today about intelligence tests? I would like to emphasize that I am not claiming that what we thought then was entirely wrong, or that what we think now represents the absolute truth. These are spiral processes. Some of the beliefs that were common “then” and subsequently “went out of fashion” may eventually return in a new form.

What Did We Think “Then” About Intelligence Tests?

First, we regarded intelligence as atheoretical. Most of us used an intelligence test that was not based on a theory of intelligence: the Wechsler test.

We regarded intelligence as a single entity—general intellectual ability—or as two entities: verbal intelligence and performance intelligence. Consequently, we tended to administer the entire Wechsler test to every child who came for an assessment, in order to derive these one or two entities.

We believed that discrepancies between subtests on an intelligence test had clinical significance. When a child received a score of 8 on Similarities and a score of 10 on Vocabulary, we believed that this discrepancy was clinically meaningful and interpreted it accordingly.

We believed that an uneven subtest profile was an indicator of a learning disability or some other problem. We assumed that a normally functioning child should have a flat, uniform profile.

We also believed that the child’s general intelligence level was essential for determining the presence of a learning disability. We worked according to the discrepancy formula, which defined a learning disability as a discrepancy between ability and achievement that could not be explained by exclusionary factors. Ability was measured by the general IQ score.

What Is Our Current System of Beliefs About Intelligence Tests?

We now see intelligence as both one and many. There is a general level of intelligence, but there are also multiple cognitive abilities. We regard these cognitive abilities as more important and meaningful than the general intelligence score. Familiarity with a child’s cognitive abilities may help us plan appropriate interventions more effectively than knowledge of the child’s general IQ.

We regard cognitive abilities as causally related to achievement in reading, writing, and mathematics. When a child has a specific difficulty in reading, for example, we may argue that the difficulty is caused by a weakness in a particular cognitive ability, provided that exclusionary factors, such as inadequate instruction, do not offer a better explanation.

To improve the child’s reading, we may therefore recommend an intervention designed to improve the child’s functioning in the weakened cognitive ability.

We now regard an uneven subtest profile as normative and believe that it cannot, on its own, serve as evidence of a problem. For example, we know that in the WAIS-R standardization sample, the average discrepancy between the highest and lowest subtest scores was 6.7 points, or 2.1 standard deviations, since the subtests have a mean of 10 and a standard deviation of 3.

Approximately 69% of the participants in the standardization sample had a discrepancy of six points or more—two standard deviations or more—between their highest and lowest subtest scores.

According to the WAIS-III examiner’s manual, 90.3% of the population has a discrepancy of at least three points—one standard deviation—among the scores on the six verbal subtests. In other words, substantial variability among subtest scores is the rule rather than the exception (Binder et al., 2009).

We now believe that interpretation at the level of broad cognitive abilities is preferable to interpretation at the level of individual subtests. A broad ability is measured by several tests and is therefore more reliable.

Today, when we use an intelligence test, we work in light of a specific diagnostic question. The diagnostic question, together with our view that broad abilities are more important than the general intelligence score, leads us to administer intelligence tests flexibly, with particular emphasis placed on the abilities relevant to the diagnostic question.

We now believe that determining the presence of a learning disability generally does not require measurement of the child’s general intelligence level. The discrepancy formula as a definition of learning disability has “gone out of fashion” and has been abandoned in the United States and, subsequently, in Israel. We now prefer to work according to Flanagan’s approach but also use the DSM-5 approach.

According to the Flanagan approach, a learning disability involves a “consistent correspondence between low achievement and low ability, while most other cognitive abilities are intact,” provided that exclusionary factors do not offer a better explanation for the low achievement.

In other words, a learning disability is a situation in which a child has low achievement in reading, writing, or mathematics; a weakness in one or two cognitive abilities can explain the child’s specific weakness in achievement; and the child’s other cognitive abilities are intact.

The child’s general intelligence level plays no role in this definition. It therefore does not need to be measured in order to determine the presence of a learning disability, unless a differential diagnosis involving intellectual disability or an intellectual developmental disorder is required.

Many of us psychologists hold additional beliefs about intelligence tests that are not reflected in the intelligence tests currently familiar to us in Israel. These include the belief that intelligence tests should assess both attention and emotional intelligence.

We recently learned that emotional intelligence is defined as a broad cognitive ability in the latest version of CHC theory. In my estimation, however, many years will pass before we are able to assess this ability as part of an intelligence test.

With regard to attention, PASS theory addresses the need to assess this area as part of intelligence.

One final point: we now have multiple theories of intelligence. We have moved from an atheoretical position to a situation in which several different theories are available.

What Is Our Clients’ System of Beliefs About Intelligence Tests?

By “clients,” I mean the educational staff, the parents, and the child.

I suspect that our clients believe that an intelligence test shows how smart a person is in the sense of wisdom. We are very familiar with the distinction between IQ and wisdom. Perhaps this is the same distinction—or the same gap—that tests of emotional intelligence may help to bridge, at least in part. Some of our clients, however, may not recognize these distinctions.

I think our clients believe that it is possible to pass or fail an intelligence test and that failing an intelligence test may result in placement in special education.

Some of our clients probably believe that intelligence tests discriminate against minority groups or groups from different cultural and linguistic backgrounds, such as Ethiopian Israelis. They may therefore hesitate to consent to intelligence testing.

We know that when an intelligence test is administered with an awareness of the different ways in which children from diverse cultural and linguistic backgrounds may approach the tasks included in it, children can be assessed fairly.

Our clients’ willingness to implement interventions based on the findings of our intelligence assessments depends on whether they regard intelligence and cognitive abilities as fixed or as capable of change and development—or, in Carol Dweck’s terms, whether they have a fixed mindset or a growth mindset.

Clients who regard intelligence and cognitive abilities as fixed will invest less effort in improving functioning in cognitive abilities. Clients who regard intelligence as changeable will be more willing to engage in interventions intended to improve cognitive functioning.

The view that cognitive abilities can be changed underlies the work of institutes that prepare people for the SAT or ACT, or examinations used to identify gifted children.

Our system of beliefs—and that of our clients—about intelligence tests and intelligence itself is important because it affects the way we work with our clients.

Our beliefs as psychologists are influenced by the models of intelligence with which we are familiar. Models of intelligence affect the way we perceive and conceptualize the child’s difficulties and strengths, as well as the recommendations we provide to our clients.

What are the leading theoretical models of intelligence?

And what does PASS theory contribute?

These questions will be addressed in Part Two.

 

Binder, L. M., Iverson, G. L., & Brooks, B. L. (2009). To err is human: “Abnormal” neuropsychological scores and variability are common in healthy adults. Archives of Clinical Neuropsychology, 24(1), 31–46.