For decades, psychologists have been concerned with IQ measurement. It’s always been a method of assessing a child’s intelligence and a criterion for their educational placement. However, as important as IQ measurements are, very little attention has been given to the types of IQ measures. Generally speaking, there are two types of IQ measures: verbal and nonverbal. Wechsler’s original IQ scales were the first to notice the difference between the two types of measures since they included verbal scores and performance (nonverbal) scores. So what’s the difference between the two? Well, put simply, the difference is in their language requirements. Verbal IQ measures rely heavily on a child’s language abilities. On the other hand, nonverbal IQ measures can effectively measure a child’s level of intelligence, regardless of their linguistic abilities.
The origin of nonverbal IQ measures is actually quite interesting. They were first developed, not by psychologists, but by the US military during World War I. You see, not all recruits spoke English profoundly. Some showed limited English proficiency while others were illiterate altogether. Nonverbal IQ measures were an effective method of assessing them all equally. Later on, these nonverbal IQ measures were used for special populations such as those with hearing loss, neurological damage, learning disabilities, and speech-language impairment.
As ambiguous as the concept of intelligence is, there are quite a few IQ measures available on the market. So how do you choose an efficient one? For our purposes, we’ll be talking about 5 criteria that have helped us with our choice. First of all, the measure has to be marketed and commonly used for measuring general cognitive functioning. Second of all, it has to provide a standardized score for nonverbal ability. This means that it relies heavily on visuospatial skills without requiring a verbal response from the examinee. Third, the measure has to be relatively recent, which means that it had to be revised or developed within the last 15 years. The only exception to this rule is the Columbia Mental Maturity Scale—Third Edition, which was published 30 years ago, but is frequently used in the field of speech-language pathology. Another important criterion is that measures have to be suitable for preschool or school-age children. Last but not least, an effective measure has to be suitable for the general population, not just specific groups such as those with visual impairment or deafness.
When comparing different nonverbal IQ measures, it’s important to specify whether or not each nonverbal component has a verbal subscore. An example of this would be a test that requires a child to identify an object or picture then name it. Another important factor to consider when comparing nonverbal IQ measures is the form of instructions given before administration, which can be verbal or nonverbal. It’s not uncommon for a nonverbal test to have verbal instructions. For example, if the nonverbal test required a child to simply point at a picture, the verbal instructions could be “I’m going to show you some pictures. Please point to the picture that is missing something”. In other words, one might say that nonverbal measures are actually “language-reduced”, not completely devoid of language. Even if the instructions were nonverbal, the language factor could still exist. For example, when one child was asked to recite the order by which some pictures had appeared, the child labeled them “man, boy, girl, and baby” to help remember their order. Another thing to keep in mind is that the measure should be suited to the age you’re dealing with. You should take into consideration that some tests are dependent on time constraints, so if a test allots 2 minutes per question, allowing more time would nullify the score.
Nonverbal IQ measures are always compared to normative samples to makes sure the results are accurate. But here’s the question, how do you know the normative samples themselves are accurate? Well, there are generally 3 criteria for a good normative sample. First, it should consist of a large number of participants. The larger the sample is, the lower the error possibility. After all, you can’t expect ten 6-year-olds to represent all the 6-year-olds in the US, can you? It’s preferred that the number of participants be at least 100 for each child group. Second of all, the normative sample should be representative. This means that it includes all variables of race/ethnicity, socioeconomic status, and disabilities. So it should be large and representative, what else? It has to be recent. This is important because IQ scores change across time, according to the Flynn effect, which shows a 3-point increase per decade. Normative samples within the last 15 years should be acceptable.
In addition to having sufficient normative samples, test developers have to prove their tests were reliable and valid. For something to be reliable, it has to be consistent across items (internal reliability), across time (test-retest reliability), and across examiners (interrater reliability). It should also have an available standard of error measurement (SEM), which provides a sense of how much error is likely to occur. For example, the Standard Battery Full-Scale score SEM is 4. This means that if a child were to score an 87, their real score could be anywhere between 83 and 91.
There are a lot of methods to determine validity. For starters, the test should be consistent with developmental trends, meaning the test’s raw scores should increase with age. That makes sense because performance increases with age, doesn’t it? Second of all, the test should have good correlation with similar tests. It should also give similar results when administered to groups who have been already diagnosed.
Now, for the most important part: how to select an appropriate nonverbal IQ measure. There’s no golden rule but there are a few recommendations. First and foremost, the measure should be psychometrically sound. It should take into consideration the special needs of the individual being evaluated. Be especially careful of linguistic demands when dealing with language difficulties. The American Educational Research Association specifies that as long as you’re not specifically measuring linguistic or reading ability, you should keep them to a minimal. The purpose of assessment is also important in your choice of nonverbal IQ measure. For high-stakes assessments like diagnoses, treatment eligibility, or educational placement, go for a multidimensional battery like UNIT or WISC-V. For low-stakes assessments, such as research purposes, you’re better off with a one-dimensional battery such as TONI-3 or WASI.
When interpreting the results of nonverbal IQ measures, keep in mind that they are less effective than verbal measures in predicting academic achievements. After all, they exclude linguistic abilities, and language plays an important role in the academic learning environment. That’s why it’s difficult for children with poor language skills to succeed in school, regardless of how high their level of cognition is.
All nonverbal IQ measures have pros and cons so no matter which one you choose; you have to be knowledgeable of its strengths and limitations. No measure is perfect but with proper judgment, you can get as close to perfection as possible. And remember, the IQ test you choose could very well change a child’s entire life, so choose wisely.