An article in the August 20-27, 2007 Newsweek by science writer Sharon Begley, entitled “The Puzzle of Hidden Ability” describes the use of a common and a less familiar test to determine the intelligence of autistic children. I’ve been interested in autism for years because of its relationship to my own work so the study article immediately caught my attention. It also got me thinking about how easy it also is to judge and label companion animals based on tests about which we know little or nothing.
Getting back to the Newsweek article, anyone who has been educated in a typical public school system has probably taken an intelligence test at one time or another. The most widely used in this country is the Wechsler, in which the tester asks the student to perform a variety of mental tasks. The student’s responses are then quantified and reduced to a number referred to as the student’s intelligence quotient or IQ. This number then sets into motion expectations among parents, teachers, and others that will affect how people relate to that child from then on.
Because it’s difficult for autistic children to relate one-on-one to people, naturally a test that involves one-on-one communication with the tester is not going to produce valid results. Their awareness of this led researcher Laurent Mottron and his team to give students another test, the Raven’s Progressive Matrices test, as well as the Wechsler. Whereas the Wechsler measures what students have already learned, the Raven’s measures their ability to learn. This is a subtle, but crucial, difference.
In the most general terms, another way to look at this is that the Wechsler measures and judges how well a person was able to learn what someone believes that person should have learned using the usual teaching methods employed by that society. The Raven’s, on the other hand, measures the individual’s ability to learn in a way that is most natural/comfortable for him or her. When children with full-blown autism were given the two tests, their average Wechsler score was around the 30th percentile, earning them the label of “low average IQ.” However, on the Raven’s their average was in the 56% percentile, i.e. normal range. Even more disconcerting to the point of shocking relative to its effect on the child, not one of the autistic students scored in the Wechsler’s “high intelligence” range, whereas a third of those who took the Raven’s did. In one heart-breaking case, a child labeled mentally retarded in Wechsler terms ranked in the 94th percentile on the Raven’s. According to the article healthy (whatever that means) children’s scores on both tests were more uniform.
And therein lies the key to understanding both teaching and testing methods in children and also in companion animals. If the test or method is designed, not to assess the ability to learn but rather the test-taker’s ability to conform to the tester’s definition of what that individual ought to know and also the “right” way to gain and express that knowledge, then the results don’t really test intelligence at all. But even more troubling is that the labeling which constitutes the key reason behind the testing generates the aforementioned response from others that can cause those so labeled to behave in a manner that confirms the test results.
Unfortunately, those who support the use of the test then use this phenomenon as conformation of the test’s validity. Pretty soon both the tests and the labels they assign become so much a part of the culture that they take on a life of their own. As authority figures recommend a test and apply labels based on its results and the public accepts these (based on the notion that said authority figures surely must know what they’re talking about), the test ascends from being a useful tool in certain specifically defined situations to unchallenged standard for all. That the test’s origins might be suspect or based on conclusions that are questionable as more knowledge is gained is rarely questioned.
There are two phenomena that further make questioning this approach a worthwhile mental and moral exercise. The first takes us to a classic study called “Pygmalion in the Classroom.” In it, Rosenthal and Jacobson demonstrated how teachers exposed to information regarding student ability may become involved in a self-fulfilling prophecy instead of engaging in a meaningful teaching experience with that student. (Read more about the study here.) Although for ethical reasons the study could only look at the effect positive student labels had on teacher perception, enough other examples of how this self-fulfilling prophecy works make it reasonable to assume that negative labels would have correspondingly negative effects. A good example of this was the public response to the recent, basically normal stock market correction in response to some stupid investment schemes in the US housing market. As soon as a few people the public viewed as authoritative—including newscasters—got into their “The sky is falling” mode, the result was a downturn in stock markets worldwide. In the companion animal world, my files bulge with examples of how owners’ expectations/fears of their animal’s negative behavior greatly increased the probability of that occurring.
Hopefully, most people take any companion animal intelligence tests with at least a ton of salt. However, it is possible to find trainers, veterinarians, and other animal-care professionals who will label animals based on some other behavioral test. A pup who wriggles when put on his back gets labeled “dominant” and one who runs under a chair when someone drops a frying pan gets tagged with “timid.” Once the test is given and the label affixed, it changes the way both owners and professionals view and relate to that animal. And because we human beings do like to be right, once we’ve assigned and/or accepted a label, it can require a fair amount of energy to get us to change it.
This brings me to my second point regarding testing. Over the past three decades, there’s been a trend toward giving tests and their results more and more power. This could result because, as government funding for education declines, test-dependent commercial concerns have stepped up to fill the void. However, it could simply be that testing is easier and requires less skill than interacting one-on-one with people and their animals. When I mention the credo that governed my clinical training—”Never conduct a test to give you the answer. Always conduct it to confirm what you already know from a thorough physical examination and history.”—some of my colleagues look at me like I have two heads. I sadly must presume that this occurs because they have more faith in the tests and their results than they do in their ability to do that physical and get a good history.
None of this is to say that the tests and labels that proliferate in the realm of companion animal care are valueless. But it is to say that they’re developed by mere mortals and their value for a particular individual in a particular situation may be limited. Because of this, and going back to the mantra of our responsibility to act as advocates for our animals, before you accept any test or label for your pet, consider the following:
- Why is this test being recommended/conducted? What did the professional determine from his/her examination and the history that makes the test a valid option?
- What does the test involve?
- How will those results benefit you and your pet?
- How will any labels associated with the test results affect your and others’ perceptions of your pet? If these are negative, what can you do to eliminate this effect? (Tests that produce negative labels without simultaneously offering information about how to remedy the problem are useless to worse than no test at all.)
What if you don’t agree with the results or any labels associated with them? Get a second opinion. Sometimes the test is all right, but the person giving it lacks the necessary experience and skill to properly administer it or interpret the results. In the case of medical tests, sometimes the equipment is flawed or the person processing any sample makes a processing error. Other times, breed or individual differences may exist such that a problematic result for one breed or individual might be in the normal range of another or vice versa. For example, some lines of rottweilers use a low rumble to “talk” while other breeds might whine or moan; those unaware of this might erroneously label those rottweilers aggressive or assume the moaners are experiencing physical pain. Thyroid levels that are normal for one breed might signal problems for another. If the professional is unaware of these discrepancies, your pet could wind up with an inappropriate label that undermines his or her well-being.
Tests of all sorts are merely tests. They’re not infallible tools created by some omniscient Higher Power. The only way they can achieve that status is if we give it to them. Granted, this is the easiest thing to do and might even turn out to be the right thing most of the time. But if something about the test, the results, or the label it produces bothers you, it probably does so for a legitimate reason. You owe it to yourself and your pet to find out what this reason is.
An article in the August 20-27, 2007 Newsweek by science writer Sharon Begley, entitled “The Puzzle of Hidden Ability” describes the use of a common and a less familiar test to determine the intelligence of autistic children. I’ve been interested in autism for years because of its relationship to my own work so the study article immediately caught my attention. It also got me thinking about how easy it also is to judge and label companion animals based on tests about which we know little or nothing.
Getting back to the Newsweek article, anyone who has been educated in a typical public school system has probably taken an intelligence test at one time or another. The most widely used in this country is the Wechsler, in which the tester asks the student to perform a variety of mental tasks. The student’s responses are then quantified and reduced to a number referred to as the student’s intelligence quotient or IQ. This number then sets into motion expectations among parents, teachers, and others that will affect how people relate to that child from then on.
Because it’s difficult for autistic children to relate one-on-one to people, naturally a test that involves one-on-one communication with the tester is not going to produce valid results. Their awareness of this led researcher Laurent Mottron and his team to give students another test, the Raven’s Progressive Matrices test, as well as the Wechsler. Whereas the Wechsler measures what students have already learned, the Raven’s measures their ability to learn. This is a subtle, but crucial, difference.
In the most general terms, another way to look at this is that the Wechsler measures and judges how well a person was able to learn what someone believes that person should have learned using the usual teaching methods employed by that society. The Raven’s, on the other hand, measures the individual’s ability to learn in a way that is most natural/comfortable for him or her. When children with full-blown autism were given the two tests, their average Wechsler score was around the 30th percentile, earning them the label of “low average IQ.” However, on the Raven’s their average was in the 56% percentile, i.e. normal range. Even more disconcerting to the point of shocking relative to its effect on the child, not one of the autistic students scored in the Wechsler’s “high intelligence” range, whereas a third of those who took the Raven’s did. In one heart-breaking case, a child labeled mentally retarded in Wechsler terms ranked in the 94th percentile on the Raven’s. According to the article healthy (whatever that means) children’s scores on both tests were more uniform.
And therein lies the key to understanding both teaching and testing methods in children and also in companion animals. If the test or method is designed, not to assess the ability to learn but rather the test-taker’s ability to conform to the tester’s definition of what that individual ought to know and also the “right” way to gain and express that knowledge, then the results don’t really test intelligence at all. But even more troubling is that the labeling which constitutes the key reason behind the testing generates the aforementioned response from others that can cause those so labeled to behave in a manner that confirms the test results.
Unfortunately, those who support the use of the test then use this phenomenon as conformation of the test’s validity. Pretty soon both the tests and the labels they assign become so much a part of the culture that they take on a life of their own. As authority figures recommend a test and apply labels based on its results and the public accepts these (based on the notion that said authority figures surely must know what they’re talking about), the test ascends from being a useful tool in certain specifically defined situations to unchallenged standard for all. That the test’s origins might be suspect or based on conclusions that are questionable as more knowledge is gained is rarely questioned.
There are two phenomena that further make questioning this approach a worthwhile mental and moral exercise. The first takes us to a classic study called “Pygmalion in the Classroom.” In it, Rosenthal and Jacobson demonstrated how teachers exposed to information regarding student ability may become involved in a self-fulfilling prophecy instead of engaging in a meaningful teaching experience with that student. (Read more about the study here.) Although for ethical reasons the study could only look at the effect positive student labels had on teacher perception, enough other examples of how this self-fulfilling prophecy works make it reasonable to assume that negative labels would have correspondingly negative effects. A good example of this was the public response to the recent, basically normal stock market correction in response to some stupid investment schemes in the US housing market. As soon as a few people the public viewed as authoritative—including newscasters—got into their “The sky is falling” mode, the result was a downturn in stock markets worldwide. In the companion animal world, my files bulge with examples of how owners’ expectations/fears of their animal’s negative behavior greatly increased the probability of that occurring.
Hopefully, most people take any companion animal intelligence tests with at least a ton of salt. However, it is possible to find trainers, veterinarians, and other animal-care professionals who will label animals based on some other behavioral test. A pup who wriggles when put on his back gets labeled “dominant” and one who runs under a chair when someone drops a frying pan gets tagged with “timid.” Once the test is given and the label affixed, it changes the way both owners and professionals view and relate to that animal. And because we human beings do like to be right, once we’ve assigned and/or accepted a label, it can require a fair amount of energy to get us to change it.
This brings me to my second point regarding testing. Over the past three decades, there’s been a trend toward giving tests and their results more and more power. This could result because, as government funding for education declines, test-dependent commercial concerns have stepped up to fill the void. However, it could simply be that testing is easier and requires less skill than interacting one-on-one with people and their animals. When I mention the credo that governed my clinical training—”Never conduct a test to give you the answer. Always conduct it to confirm what you already know from a thorough physical examination and history.”—some of my colleagues look at me like I have two heads. I sadly must presume that this occurs because they have more faith in the tests and their results than they do in their ability to do that physical and get a good history.
None of this is to say that the tests and labels that proliferate in the realm of companion animal care are valueless. But it is to say that they’re developed by mere mortals and their value for a particular individual in a particular situation may be limited. Because of this, and going back to the mantra of our responsibility to act as advocates for our animals, before you accept any test or label for your pet, consider the following:
What if you don’t agree with the results or any labels associated with them? Get a second opinion. Sometimes the test is all right, but the person giving it lacks the necessary experience and skill to properly administer it or interpret the results. In the case of medical tests, sometimes the equipment is flawed or the person processing any sample makes a processing error. Other times, breed or individual differences may exist such that a problematic result for one breed or individual might be in the normal range of another or vice versa. For example, some lines of rottweilers use a low rumble to “talk” while other breeds might whine or moan; those unaware of this might erroneously label those rottweilers aggressive or assume the moaners are experiencing physical pain. Thyroid levels that are normal for one breed might signal problems for another. If the professional is unaware of these discrepancies, your pet could wind up with an inappropriate label that undermines his or her well-being.
Tests of all sorts are merely tests. They’re not infallible tools created by some omniscient Higher Power. The only way they can achieve that status is if we give it to them. Granted, this is the easiest thing to do and might even turn out to be the right thing most of the time. But if something about the test, the results, or the label it produces bothers you, it probably does so for a legitimate reason. You owe it to yourself and your pet to find out what this reason is.