Why Psychological Assessment Needs Rethinking
In Part 1 of this series of posts Iof this series of posts for the Mensa Foundation’s Insights blog, I explained that it is important to talk about intelligence because it relates to many life outcomes. In the follow-up post, I balanced that out with a message that intelligence is not the only thing that matters in life; narrower cognitive abilities, personality, non-cognitive behaviors, ethics, and other concepts are important to consider when evaluating a person. In this final post, I describe what the psychological assessment process should look like, given what we have learned in the first two posts in this series.
The Role—and Limits—of Intelligence Tests
First, a full psychological assessment of a person should include an intelligence test. The choice of intelligence test is important, though. There is no test that is appropriate for all examinees in all settings. The right test will depend on the examinee, the reason for the assessment, and the constraints of time, money, and available expertise.
One essential requirement is that the examinee should belong to the population that the test was designed for (or a population that later research showed can take the test). Professional test creators are very clear about who their tests are designed for, and so this information is often easily available. For example, the Wechsler Intelligence Scales for Children, Fifth Edition (WISC-V) is designed for American children, ages 6 through 16. It is the psychologist’s duty to ensure that the examinee taking the WISC-V is an American child between the ages of 6 and 16. If the child does not belong to that population, then the psychologist should select a different test, perhaps the WISC-V’s Spanish-language version, one of its foreign versions, or a completely different test.
If this guideline sounds basic, that is because it is basic. But this guideline is violated distressingly often. Several years ago, a friend who was born outside of the United States and does not speak English as his first language had to take an intelligence test as part of a job application process. The test he took is designed for adults born in the United States who speak English as a first language. Even though it was obvious that my friend does not belong to this population, the psychologist administered the test to him anyway. Unsurprisingly, the test underestimated his IQ, especially on the verbal sections.
Purpose Matters: Matching Tests to Context
In addition to the test’s target population, a psychologist should also consider the purpose of the assessment. If the test is being administered for a high-stakes purpose, such as competency to stand trial or diagnosis of a disability, then a high-quality test administered in person is best. The face-to-face format allows a psychologist to make observations about the examinee’s behavior that may provide vital clues to any deficits or strengths the examinee has.
On the other hand, an intelligence test may not be central to the overall assessment purpose, which opens up other test options to the examiner. For example, the National Football League administers an intelligence test called the Wonderlic as part of the series of assessments of potential players. This series, called the “NFL Scouting Combine,” includes measures of speed, strength, and physical agility; a drug screening; and an interview. If you’re looking for a great football player, the intelligence test is not the most important part of the Scouting Combine. That is why NFL teams manage just fine with a quick-and-dirty 12-minute intelligence test.
Designing Better Tests for Broader Intelligence
Ideally, an intelligence test should consist of a wide variety of tasks so that the test can measure the full breadth of a person’s intelligence. If a test has a limited number of tasks, then it may produce a distorted score for some examinees. This was seen historically with the 1916 version of the Stanford-Binet intelligence test, which was heavily stacked with verbal tasks. When its creator, Lewis Terman, used the test to identify over 1,500 children with high IQs as “geniuses,” he unwittingly missed two future winners of the Nobel Prize in Physics. While the heavily verbal content of the test was not the only reason Terman decided that these children were not “geniuses,” their IQs were surely underestimated by a test that favored verbal abilities over the mathematical and spatial thinking needed to excel in physics.
The need for many different tasks is why the test I have created, the Reasoning and Intelligence Online Test (RIOT), consists of 15 subtests, each with a different task format. The wide variety of subtests and tasks ensures that no one type of ability is heavily favored and that people with a wide variety of strengths can have a chance to demonstrate their intelligence. There are many paths to high IQ, and a test should have many tasks on it to reflect that fact.
Balancing Ideal Testing with Real-World Constraints
This discussion is very idealized, but reality puts constraints on testing options. Some individually administered full-length intelligence tests can take over 2 hours to administer. If there is not the manpower and time to administer a 2-hour test to every person individually, then the psychologist must choose another option, such as a shorter test or a test that can be administered to a group of examinees at once. Young children, non-English speakers, and people with disabilities often have special considerations that frequently require alternative tests in order to measure their intelligence fairly.
Intelligence is Just One Piece of the Puzzle
While measuring intelligence is important, it should never be the only component of the assessment process. Other tests and measures should be part of the process of evaluating a person because it is unethical to use a test score as the sole basis for making an important decision about an examinee.
What those other assessments should be, though, depends on the purpose of the assessment. The tasks that future athletes must engage in at the NFL Scouting Combine are very different from the tests that a school psychologist gives to a child who may have dyslexia.
Psychologists and other scientists have developed a wide variety of tests that are often useful for assessment. Even though intelligence is usually a better predictor of many outcomes, personality tests still capture important information that intelligence tests do not. Using both intelligence and personality tests in conjunction is better than using either alone. Mental health tests are valuable in clinical settings, and employers often have a fixed set of tasks they give to potential employees so that every applicant can be judged by the same standard. In all of these situations, intelligence matters, but not always to the same degree, and where it fits in the assessment process can vary greatly.
Technology and the Future of Testing
The 21st century is an exciting time to work in testing. The field is taking advantage of technological developments to make testing more efficient and accurate. Educational testing is leading the way in this regard. Readers my age (or older) will associate standardized tests with #2 pencils and scantron sheets. Paper-and-pencil tests are largely obsolete in American educational testing, and most students take professionally developed tests on computers today.
In psychological assessments, more tests are given either on computers or with technological aids, such as smartphones. The Wechsler intelligence tests can now be administered via video calls, and some tests can be administered with the assistance of tablets and other electronic devices. The Reasoning and Intelligence Online Test (RIOT) is part of this technological trend: in addition to being fully computer-administered, the test is developed and updated continuously using the same philosophy and process that is used in software development. Unlike traditional tests, the RIOT will not go years between updates; this will keep the test responsive to user needs and the latest research.
Despite the exciting technological advances, paper-and-pencil tests still have their place. The traditional administration format is better for people with certain disabilities that keep them from using a computer or a mouse. Young children or people without much exposure to technology may also struggle with computerized assessment.
Professional Judgment and a Holistic View
So, there is no single right way to measure a person’s intelligence, personality, or other important characteristics. What assessment should look like will depend on the person, the reason for the assessment, and the constraints of the situation. An intelligence test should usually be a part of this process, but it is almost never the only source of data in the assessment process.
A variety of conditions will impact which intelligence test a psychologist chooses to administer. This makes professional judgment very important, and psychologists should be prepared to justify their choice of tests and ensure that they can explain how those tests align with the examinee, the purpose of the assessment, and the needs of stakeholders.
What matters most is getting a holistic view of an individual. People are more than a single test score, and the assessment process should reflect that. A multiprong assessment process that is customized to the situation is the best way to measure potential and help examinees reach the highest level of development that they can.
Dr. Warne is a leading intelligence researcher and psychometrician with 15+ years of experience in cognitive assessment. At RIOT IQ, he oversees the scientific development of intelligence tests, from subtest design to data analysis. He is the author of In the Know and Statistics for the Social Sciences (Cambridge University Press) and has published over 60 scholarly articles. Outside of research, he’s a father of four, a theater critic, and a lifelong musician.
Comments (0)