In order to
determine the intellectual ability of a
person, psychiatrist use intelligent test. This is done in medicine to
ascertain the cognitive functioning of a
patient who might have suffered brain damage. The test also provide information
and insights about a person's intellect. Intelligence test is also widely used
in other fields other than medicine. For instance in business, employers use
intelligence test coupled with aptitude test select their employees. In the military, intelligent test is one of
most reliable methods used in selecting
their personnel. According to Encarta, these test helps employers to predict
which employee has the ability to
acquire new information and this is very helpful in selecting people for
complex and intellectually tasking job.
Since World War I, the United States military has had one of the most comprehensive testing programs for selection and job assignment. Anyone entering the military takes a comprehensive battery of tests, including an intelligence test. For specialized and highly skilled jobs in the military, such as jet pilot, the testing is even more rigorous. Intelligence tests are helpful in the selection of individuals for complex jobs requiring advanced skills. The major reason intelligence tests work in job selection is that they predict who will learn new information required for the job. To a lesser extent, they predict who will make “smart” decisions on the job.
Although intelligent
tests have played major roles in offering admission to students into colleges
and universities, hiring employees and drafting of military personnel, many people, especially psychologist and education
experts have express their concern over
the possibility of accurately measuring a person's success both academically
and socially. In the following article, professor Robert J. Sternberg of Yale
University in New Haven, Connecticut, provides insight into the accuracy of
intelligent test and also recommends ways of improving the method.
How Intelligent Is
Intelligence Testing?
By Robert J.
Sternberg
A typical American
adolescent spends more than 5,000 hours in high school and several thousand
more hours studying in the library and at home. But for those students who wish
to go on to college, much of their fate is determined in the three or so hours it
takes to complete the Scholastic Assessment Test (SAT) or the American College
Test (ACT). Four years later they may find themselves in a similar position
when they apply to graduate, medical, law or business school.
The stakes are high.
In their 1994 book The Bell Curve,
Richard J. Herrnstein and Charles Murray pointed out a correlation between
scores on such tests and a variety of measures of success, such as occupational
attainment. They suggested that the U.S. is developing a 'cognitive
elite'—consisting of high-ability people in prestigious, lucrative jobs—and a
larger population of low-ability people in dead-end, low-wage positions. They
suggested an invisible hand of nature at work.
But to a large
extent, the hand is neither invisible nor natural. We have decided as a society
that people who score well on these high-stakes tests will be granted admission
to the best schools and, by extension, to the best access routes to success. People
have used other criteria, of course: caste at birth, membership in governmental
party, religious affiliation. A society can use whatever it wishes—even height,
so that very soon people in prestigious occupations would be tall. (Oddly
enough, to some extent Americans and many people in other societies already use
this criterion.) Why have the U.S. and other countries chosen to use ability
tests as a basis to open and close the access gates? Are they really the
measures that should be used? The answers lie in how intelligence testing
began.
A
Brief History of Testing
Sir Francis Galton,
a cousin of [British scientist] Charles Darwin, made the first scientific
attempt to measure intelligence. Between 1884 and 1890 Galton ran a service at
the South Kensington Museum in London, where, for a small fee, people could
have their intelligence checked. The only problem was that Galton's tests were
ill chosen. For example, he contrived a whistle that would tell him the highest
pitch a person could perceive. Another test used several cases of gun
cartridges filled with layers of either shot, wool or wadding. The cases were
identical in appearance and differed only in weight. The test was to pick up
the cartridges and then to discriminate the lighter from the heavier. Yet
another test was of sensitivity to the smell of roses.
James McKeen
Cattell, a psychologist at Columbia University, was so impressed with Galton's
work that in 1890 he devised similar tests to be used in the U.S. Unfortunately
for him, a student of his, Clark Wissler, decided to see whether scores on such
tests were actually meaningful. In particular, he wanted to know if the scores
were related either to one another or to college grades. The answer to both
questions proved to be no—so if the tests didn't predict school performance or
even each other, of what use were they? Understandably, interest in Galton's
and Cattell's tests waned.
A Frenchman, Alfred
Binet, got off to a better start. Commissioned to devise a means to predict
school performance, he cast around for test items. Together with his colleague
Theodore Simon, he developed a test of intelligence, published in 1905, that measured
things such as vocabulary ('What does misanthrope
mean?'), comprehension ('Why do people sometimes borrow money?') and verbal
relations ('What do an orange, an apple and a pear have in common?'). Binet's
tests of judgment were so successful at predicting school performance that a
variant of them, called the Stanford-Binet Intelligence Scale (fourth edition),
is still in use today. (Louis Terman of Stanford University popularized the
test in the U.S.—hence the name.) A competing test series, the Wechsler
Intelligence Scales, measures similar kinds of skills.
It is critical to
keep in mind that Binet's mission was linked to school performance and,
especially, to distinguishing children who were genuinely mentally retarded
from those who had behavior problems but who were able to think just fine. The
result was that the tests were designed, and continue to be designed, in ways
that at their best predict school performance.
During World War I
[1914-1918], intelligence testing really took off: psychologists were asked to
develop a method to screen soldiers. That led to the Army Alpha (a verbal test)
and Beta (a performance test with pantomimed directions instead of words), which
were administered in groups. (Psychologists can now choose between group or
individually administered tests, although the individual tests generally give
more reliable scores.) In 1926 a new test was introduced, the forerunner to
today's SAT. Devised by Carl C. Brigham of Princeton University, the test
provided verbal and mathematical scores.
Shortly thereafter,
a series of tests evolved, which today are used to measure various kinds of
achievements and abilities, including IQ (intelligence quotient), 'scholastic
aptitude,' 'academic aptitude' and related constructs. Although the names of
these tests vary, scores on all of them tend to correlate highly with one
another, so for the purposes of this article I will refer to them loosely as
conventional tests of intelligence.
What
Tests Predict
Typically,
conventional intelligence tests correlate about 0.4 to 0.6 (on a 0 to 1 scale)
with school grades, which statistically speaking is a respectable level of
correlation. A test that predicts performance with a correlation of 0.5,
however, accounts for only about 25 percent of the variation in individual
performances, leaving 75 percent of the variation unexplained. (In statistics,
the variation is the square of the correlation, so in this case, 0.52 = 0.25.)
Thus, there has to be much more to school performance than IQ.
The predictive
validity of the tests declines when they are used to forecast outcomes in later
life, such as job performance, salary or even obtaining a job in the first
place. Generally, the correlations are only a bit over 0.3, meaning that the
tests account for roughly 10 percent of variation in people's performance. That
means 90 percent of the variation is unexplained. Moreover, IQ prediction
becomes less effective once populations, situations or tasks change. For
instance, Fred Fiedler of the University of Washington found that IQ positively
predicts leadership success under conditions of low stress. But in high-stress
situations, the tests negatively predict success. Some intelligence tests,
including both the Stanford-Binet and Wechsler, can yield multiple scores. But
can prediction be improved?
Curiously, whereas
many kinds of technologies, such as computers and communications, have moved
forward in leaps and bounds in the U.S. and around the world, intelligence
testing remains almost a lone exception. The content of intelligence tests
differs little from that used at the turn of the century. Edwin E. Ghiselli, an
American industrial psychologist, wrote an article in 1966 bemoaning how little
the predictive value of intelligence tests had improved in 40 years. More than
30 years later the situation remains unchanged.
Improving
Prediction
We can do better. In research with Michael
Ferrari of the University of Pittsburgh, Pamela R. Clinkenbeard of the
University of Wisconsin-Whitewater and Elena L. Grigorenko of Yale University,
I showed that a test that measured not only the conventional memory and
analytical abilities but also creative and practical thinking abilities could
improve prediction of course grades for high school students in an introductory
psychology course. (A direct comparison of correlations between this test and
conventional tests is not possible because of the restricted sample, which
consisted of high-ability students selected by their schools.)
In these broader
tests, individuals had to solve mathematical problems with newly defined
operators (for example, X glick Y = X + Y if X < Y, and X - Y if X ≥ Y),
which require a more flexible kind of thinking. And they were asked to plan
routes on maps and to solve problems related to personal predicaments, which
require a more everyday, practical kind of thinking. Here is one example:
The following
question gives you information about the situation involving a high school
student. Read the question carefully. Choose the answer that provides the best
solution, given the specific situation and desired outcomes.
John's family moved
to Iowa from Arizona during his junior year in high school. He enrolled as a
new student in the local high school two months ago but still has not made
friends and feels bored and lonely. One of his favorite activities is writing
stories. What is likely to be the most effective solution to this problem?
A. Volunteer to work
on the school newspaper staff B. Spend more time at home writing columns for
the school newsletter C. Try to convince his parents to move back to Arizona D.
Invite a friend from Arizona to visit during Christmas break Best answer: A
Creativity can
similarly be measured. For example, in another study, Todd Lubart, now at René
Descartes University-Paris V, and I asked individuals to perform several
creative tasks. They had to write short stories based on bizarre titles such as
The Octopus's Sneakers or 3853, draw pictures of topics such as the
earth seen from an insect's point of view or the end of time, come up with
exciting advertisements for bow ties, doorknobs or other mundane products, and
solve quasiscientific problems, such as how someone might find among us
extraterrestrial aliens seeking to escape detection. The research found that
creative intelligence was relatively domain-specific—that is, people who are
creative in one area are not necessarily creative in another—and that creative
performance is only weakly to moderately correlated with the scores of
conventional measures of IQ.
The implications for
such testing extend to teaching. The achievement of students taught in a way
that allowed them to make the most of their distinctive pattern of abilities
was significantly higher than that of students who were taught in the conventional
way, emphasizing memory. Indeed, further research done by Bruce Torff of
Hofstra University, Grigorenko and me has shown that the achievements of all
students improve, on average, when they are taught to think analytically,
creatively and practically about the material they learn, even if they are
tested only for memory performance.
Interestingly,
whereas individuals higher in conventional (memory and analytical) abilities
tended to be primarily white, middle- to upper-middle-class and in 'better'
schools, students higher in creative and practical abilities tended to be
racially, socioeconomically and educationally more diverse, and group
differences were not significant. Group differences in conventional test
scores—which are common and tend to favor white students—therefore may be in
part a function of the narrow range of abilities that standard tests favor.
Tests can also be
designed to improve prediction of job performance. Richard K. Wagner of Florida
State University and I have shown that tests of practical intelligence in the
workplace can predict job performance as well as or better than IQ tests do, even
though these tests do not correlate with IQ. In such a test, managers might be
told that they have a number of tasks to get done in the next three weeks but
do not have time to do them all and so must set priorities. We have devised
similar tests for salespeople, students and, most recently, military leaders
(in a collaborative effort with psychologists at the U.S. Military Academy at
West Point). Such tests do not replace conventional intelligence tests, which
also predict job performance, but rather supplement them.
A
Question of Culture
Cultural
prerogatives also affect scores on conventional tests. Grigorenko and I, in
collaboration with Kate Nokes and Ruth Prince of the University of Oxford,
Wenzel Geissler of the Danish Bilharziasis Laboratory in Copenhagen, Frederick
Okatcha of Kenyatta University in Nairobi and Don Bundy of the University of
Cambridge, designed a test of indigenous intelligence for Kenyan children in a
rural village. The test required them to perform a task that is adaptive for
them: recognizing how to use natural herbal medicines to fight illnesses.
Children in the village knew the names of many such medicines and in fact
treated themselves once a week on average. (Western children, of course, would
know none of them.) The children also took conventional IQ tests.
Scores on the
indigenous intelligence test correlated significantly but negatively with
vocabulary scores on the Western tests. In other words, children who did better
on the indigenous tests actually did worse on the Western tests, and vice
versa. The reason may be that parents tend to value indigenous education or
Westernized education but not both, and they convey those particular values to
their children.
People from
different cultures may also interpret the test items differently. In 1971
Michael Cole, now at the University of California at San Diego, and his
colleagues studied the Kpelle, who live in western Africa. Cole's team found
that what the Kpelle considered to be a smart answer to a sorting problem,
Westerners considered to be stupid, and vice versa. For instance, given the
names of categories such as fruits and vegetables, the Kpelle would sort
functionally (for instance, 'apple' with 'eat'), whereas Westerners would sort
categorically ('apple' with 'orange,' nested under the word 'fruit').
Westerners do it the
way they learn in school, but the Kpelle do it the way they (and Westerners)
are more likely to do it in everyday life. People are more likely to think
about eating an apple than about sorting an apple into abstract taxonomic
categories.
Right now
conventional Western tests appear in translated form throughout the world. But
the research results necessarily raise the question of whether simply
translating Western tests for other cultures makes much sense.
Toward
a Better Test
If we can do better
in testing than we currently do, then, getting back to the original question
posed at the beginning of the article, how have we gotten to where we are?
Several factors have conspired to lead us as a society to weigh conventional
test scores heavily:
1. The appearance of precision. Test scores look
so precise that institutions and the people in them often accord them more
weight then they probably deserve.
2. The similarity factor. A fundamental principle
of interpersonal attraction is that people tend to be attracted to those who
are similar to them. This principle applies not only in intimate relationships
but in work relationships as well. People in positions of power look for others
like themselves; because they needed high test scores to get where they are,
they tend to seek others who have high test scores.
3. The publication factor. Ratings of
institutions, such as those published annually in [the news magazine] U.S. News and World Report, create intense
competition among colleges and universities to rank near the top. The
institutions cannot control all the factors that go into the ranking. But test
scores are relatively easier to control than, say, scholarly publications of
faculty, so institutions start to weigh test scores more heavily to prop up
their ratings. Publication of mastery-test scores by states also increases the
pressure on the public schools to teach to the tests.
4. Confirmation bias. Once people believe in the
validity of the tests, they tend to set up situations that confirm their
beliefs. If admissions officials believe, for example, that students with test
scores below a certain point cannot successfully do the work in their institution,
they may not admit students with scores below that point. The result is that
the institutions never get a chance to see if others could successfully do the
work.
Given the
shortcomings of conventional tests, there are those who would like to get rid
of standardized testing altogether. I believe this course of action would be a
mistake. Without test scores, we are likely to fall into the trap of
over-weighting factors that should matter less or not at all, whether it is
political pull or socioeconomic status or just plain good looks. Societies
started using tests to increase, not to decrease, equity for all.
Others would like to
use only performance-based measures,
such as having children do actual science experiments. The problem with such
measures is that, despite their intuitive appeal, they are no less culturally
biased than conventional tests and have serious problems of statistical
reliability and validity that have yet to be worked out.
A sensible plan
would be to continue to use conventional tests but to supplement them with more
innovative tests, some of which are already available and others of which have
to be invented. Unlike most kinds of companies involved in technology, testing firms
spend little or nothing on basic research, and their applied work is often
self-serving. Given the monopoly a few companies have in the testing industry
and the importance of tests, we might think as a society of strongly
encouraging or even requiring the testing companies to modify their approach.
Or the public could fund research on its own. The innovations should be not
just in the vehicles for testing (such as computerized testing) but in the very
content of the tests. The time has come to move testing beyond the horse and
buggy. We have the means; we just need the will.
About
the Author
ROBERT J. STERNBERG
… is professor of psychology and education at Yale University.
Source: Microsoft
® Encarta ® 2009. © 1993-2008 Microsoft Corporation. All rights reserved.
0 comments:
Post a Comment