There is a kind of a split in psychology between those who do basic laboratory research and those who do more applied, field research. The people in the lab are most concerned with understanding why people do that things they do, while the people in the field are more interested in doing something about it. Ultimately, one hand washes the other, and most psychologists realize this. However, let's look into the origins of the split.

A guy named Wunt is generally recognized as the father of psychology. He was interested in fundamental lawful principles of human behavior. He studied perception mostly. How do peoples' perceptions color, brightness, etc., match objectively known stimulus values? It turns out that most people see light and hear sound in the same way (unless there is something defective about their eyes or ears). I'm sitting on a dissertation committee now where the study will have an N of 2. That is, only two subjects will be run in the experiment. The study is about how the eye adapts to different illuminations over time for different colors. The basic idea in this particular study is to describe the functioning of one aspect of the visual system. What Wunt found, and the reason that only 2 subjects will be run in the study, is that people are interchangeable in these tasks. If you run any two people with 20/20 acuity and full color vision, you get the same results, as long as they are motivated to cooperate.

The stimuli and responses in the laboratory are usually simple and well controlled. A study in learning, for example, might show people two letters flashed on a screen.

A C B C A D F G A B

C D

In this study people will be shocked every time either a C or a D appears as one of the two letters, but not when both C and D appear together. The dependent variable is GSR (galvanic skin response). People who learn that C and D are shockers sweat more when these letters appear on the screen.

On the other hand, there are stimulus situations to which people differ markedly. The first historical note on individual differences comes from astronomy just before 1800. A fellow named Kinnebrook was hired to watch stars pass the field of a telescope in England. He would meticulously record times. His boss, the Royal Astronomer, consistently recorded the passing of stars .8 seconds faster than Kinnebrook. Kinnebrook got fired. Shortly after this, however, astronomers began to notice that there were consistent differences in recording time among all the astronomers -- some faster, some slower -- and that these differences were approximately normally distributed. The upshot was that astronomers could not consider one measurement right and the others wrong.

I used to kid the learning people about the value of learning experiments. I don't see anything in the study I outlined above that will help me teach a class. The testing approach that has proved most useful in psychology came from medicine. Clinicians dealing with mentally retarded and pathological people needed diagnostic tests. Psychiatrists at the turn of the century gave patients tasks, and observed the patients' reasoning, as well a persistence on the tasks. They used this information for diagnostic purposes.

Binet's definition of intelligence. Binet was appointed to a commission to recommend how children would be classified as retarded. He set to work to devise a method for doing this. He looked at numerous measures of individual differences, including recall of digits, suggestibility, moral judgment, size of cranium, and palmistry. He found that the most useful tasks were those that forced the child to reason about objects, pictures, and words. What these tasks were measuring is somewhat obscure. They are not well defined as they are in the lab. But they were useful. Those children most superior in judgment (e.g., which of these pictures comes first?) were also superior in vocabulary, attention, etc.

Binet thought that intelligence is composed of several things:

Intelligence, in this view, is like blood pressure: it is an indication of multiple processes of the person. Just as blood pressure depends on diet, exercise, constriction of blood vessels, so does intelligence depend on motivation (goal setting), flexibility, and self correction.

"The power of good responses from the pint of view of truth or fact"

E. Thordike

"The general tendency to consistently display effective responses to problems of an abstract or symbolic nature"

(R. Thorndike)

"The capacity of an individual to understand the world about him and his resourcefulness to cope with its challenges."

(Wechsler)

In 1904, Binet tried to separate the components of intelligence, and to devise tests for each. He looked at knowledge, memory, and other components. He became pessimistic that he could measure each individually. He had some clinical friends, however, that persuaded him to use the clinical approach. They said, we don't want dissection, we want diagnosis. Binet embraced their viewpoint, and went on to develop the first intelligence test.

Most of the useful (practical) things that psychologists have developed have not come from the laboratory and theory, but rather from psychologists attempting to solve practical problems (exceptions: operant conditioning & social learning). There is a place for theory, however. One of the reasons that intelligence, or academic ability, has been so controversial, is that there is no real theory of what intelligence is. Nobody can explain fully why one person does better than another on these tests. Because the explanation is lacking, people reject the tests because they aren't obviously related to performance on everyday tasks. Civil rights litigation to validity generalization

Lots of people think ability or intelligence tests measure some innate capacity. I'll talk about the "innate" part later; now I'll talk about capacity.

draw pictures

Q: What concepts or facts are necessary to solve these problems?

Q: Is it likely that some people in non-Western cultures do not posses such facts or concepts?

There are no tests of pure learning potential.

You have to know things to do well on tests. For typical academic ability tests, some items require more knowledge and some more reasoning. Vocabulary, for example, is a test of knowledge of words. The matrices are more reasoning. Younger and older people who get the same score on an academic ability test do so for different reasons. Older people get more knowledge questions correct; younger get more reasoning questions correct. Therefore it doesn't make sense to talk about them having the same mental age.

Suppose an agricultural experiment station tests varieties of corn. For hybrid G, specimens yield, on average, 47 pounds per plant.

Q: Does it make any sense to say that hybrid G seeds have a "potential" of 47 pounds?

Q: Might seeds of hybrid F, measured at an average of 42 pounds, have a greater "potential?"

Q: Suppose that learning potential was really a function of dendritic plasticity. [This isn't really true. Dendrites are neural branches which receive information from other neurons.]

Would measuring dendritic plasticity be a better measure of academic aptitude than current tests for college admissions?

Q: Suppose we had this great test of dendritic plasticity. We tested a 5 year old boy, and found his plasticity to be greater than that of many of the great thinkers of the day. (Say, about as plastic as Carl Sagan and Henry Kissenger rolled into one.) Would the 5 year old be a good candidate for college, given that he had not yet been enrolled in first grade?

The point: there are no pure measures of potential. Test scores depend upon prior learning, motivation, and some reasoning process.

Tests of ability and tests of achievement differ in their interpretations. They may or may not differ in content. The interpretation of ability tests concerns who will learn the most or, more properly, the most quickly. The interpretation of achievement tests is about who has learned the most.

Both kinds of tests are used to make decisions about people. Ability tests are usually used to decide whether to admit someone to a course of study. That is, they are classified for treatment. For example, the SAT is sometimes used to decide who to admit to college. Achievement tests are sometimes used in this way as well. Achievement tests are also used for certification and for program evaluation. Examples of certification tests are the bar exam in law, and liscensure in nursing. An example of program evaluation is evaluating school districts with a standardized test (the Lake Wobegon effect).

If achievement tests are thought of as a measure of potential, that is, how much a person can learn, then troubles arise when we look at subsequent performance. As long as the aptitude tests predict subsequent performance perfectly, there is no problem. Of course, aptitude tests never do this. Then what happens is that we have underachievers and overachievers. Is the underachiever who gets a "B" somehow worse than a normal person who gets a "C"? The real problem is with the overachievers. These are people who learned more than their potential. Are these people somehow pathological bookworms, destined to a life of plastic pocket protectors and glasses made from the bottoms of bottled water dispenser found in office buildings? No, the problem is that logically, one's achievement cannot be greater than one's potential.

I interpret aptitude test scores as predictors of (a) speed of learning, and (b) analytical or logical problem solving ability (convergent; puzzle solving). I do not interpret such scores as measure of potential or mental age. Such interpretations are too simplistic and can be rather misleading.