Vous êtes sur la page 1sur 69

Testing and Individual Differences (Intelligence)

AP Psych. Prep 11

Introduction
Tests of mental abilities, personality, abnormal psychology, etc. are very common in psychology, but tests are also very common in our everyday lives.

This chapter will help us understand tests and testing, and then we will learn a bit about intelligence, which has been tested and explored a lot in psychology.

Standardized Tests
Psychometricians - people who make tests designed to measure psych. concepts like intelligence, stress, abnormality, etc. Standardized Tests - tests whose items have been tested with a sample (standardization sample) from the population you want to measure. This way, standards of performance (Norms) have been developed.

Standardized Tests
This allows people who take the test to be compared to the other sample (and we hope to the population that the standardization sample represents)

Standardized Tests
Many standardized tests exist. e.g. SAT - includes experimental section the test makers are testing questions to create future tests; by trying them, they know how a general sample of the population should do on those questions.

Standardized Tests
They can throw out questions that are too hard or two easy and make future tests that have been normed with variable difficulty questions (questions that can help differentiate test takers from each other).

Reliability and Validity


We studied these two terms in Chapter 2
Reliability - test can be repeated and show similar results; consistently evaluates people.

Reliability
There are different ways to measure reliability:

Split-half Reliability - compare two halves (made randomly) and see if peoples scores on those two halves are similar. If they are similar, it suggests the test is reliable.

Reliability
Equivalent-form Reliability - if theres more than one form (version) for a test, you can compare peoples performance on different forms to see if they show a similar measure.

E.g. Korean Test: different testings, but all supposed to be equivalent. If a person gets a similar score on different tests, its reliable.

Reliability
Test-retest Reliability - measure of reliability where you give someone the test more than once, and see if the scores correlate with each other.

E.g. If I write a test of career aptitude today, and one week from now, will it show a similar result?

Validity
Even if a test is shown to be reliable, we dont know for sure if its measuring what its supposed to measure. Maybe its reliably measuring wrong.

Thats why we should also worry about Validity - if the test is measuring what its trying to measure.

Validity

Validity
E.g. if a career test is properly measuring which career you would do well in / enjoy, then it is a valid career aptitude test.

If that career test is actually giving us a measure of a different thing, like chemistry knowledge, degree of depressive tendencies, etc., then it is not valid.

Validity
There are also multiple types of validity: Content Validity - how well does a test measure the range of content it should measure. E.g. a Korean test with no speaking part is not covering all of a persons Korean skill, so its content validity is low.

Validity
Face Validity - is one example of content validity. On the surface, does it look like its covering all the important parts of the thing you want to measure. (very basic evaluation) Criterion-related Validity - does test measure some criteria that we are interested in? We can check against other ways to measure those criteria.

Validity
There are two kinds of criterion-related validity:

Concurrent Validity - now / current match with criteria.


E.g. Depression measure does it measure how depressed the person is now?

Validity
Predictive Validity - the other kind of criterion validity; does the measure do a good job of predicting future. E.g. a measure of depressive risk factors does it really show us which people will have depression problems later? Predictive Validity is also important for things like Career Aptitude tests.

Validity
Construct Validity - a strong way to check for validity; does the measure correlate strongly with some other (already shown to be valid) measure?

If we have a good measure of stress now, you can see if your new stress measure gives similar scores for the same people.

Validity
Very strong, but problems with construct validity:

How do we know the other measure is valid? Are we sure its valid?
If we know the other is valid, why not just use that measure instead of making a new one?

Types of Tests
Aptitude Tests - try to measure natural ability or potential.

E.g. LSAT - Law School Admission Test; tries to measure if someone would become a good lawyer, not if they are a good lawyer now.
*Intelligence tests are usually aptitude tests.

Types of Tests
Achievement Tests - measure accomplishment or what people have learned.

Speed Test - try to measure how quickly people can answer questions; give too many questions and see how many people can get right in the time given.

Types of Tests
Power Test - try to measure the difficulty of questions a person can answer. Give enough time, but questions range from easy to very difficult, so you can find the most difficult question the person can answer.

Types of Tests
Group Tests - Where lots of people take the test at once. Easy for the test administrators, doesnt require a lot of interaction between test takers and the people giving the test.

Most tests we take in school are group tests, SAT type tests, etc.

Types of Tests
Individual Tests - test taker and test administrator interact much more. Test taker is alone, and the test administrator concentrates on the test taker.

Many intelligence tests are individual tests, as are projective tests like the TAT or inkblot tests.

Types of Tests
Many intelligence tests are individual tests, as are projective tests like the Thematic Apperception Test or Inkblot Tests.

Theories of Intelligence
Psychologists dont agree on one definition of intelligence. We commonly think of it as smartness, ability to use your mind (often for academic uses), or ability to gather and synthesize information in useful ways.

We will look at a few different Psych. ideas about what intelligence is.

Theories of Intelligence
Psychologists sometimes differentiate between two kinds of intelligence:

Fluid Intelligence - our more natural intellectual skills; not based on stored knowledge or experience. For doing things like remembering, learning new things, new skills, problem solving, etc.

Theories of Intelligence
Crystallized Intelligence - where we use knowledge, information, skills, etc., that we have collected over our lives. Involves use of our past experience.

Theories of Intelligence
Research suggests that fluid intelligence might decline over our life span...

But as we get older and collect more knowledge, skills, it seem that we can have an advantage in crystallized intelligence.

Crystallized vs Fluid Intelligence

1. General Intelligence
The first theory of intelligence we will look at comes from Charles Spearman, who thinks that intelligence is one ability, and all our intellectual abilities come from our single, general intelligence. He called this g for general.

2. Multiple Intelligences
A different idea is that we have more than one kind of intelligence, and different people could have different amounts of each type.

Howard Gardner is a strong supporter of this idea.

He listed 8 of our intelligences:

2. Multiple Intelligences
Linguistic - language ability

Musical - ability to produce music, write music, etc. Spatial - ability to think and solve problems based on space info., maps, 3D info, etc.
Logical-mathematical - math and logical problem solving skills

2. Multiple Intelligences
Bodily-kinaesthetic - knowing how to and being able to move body, gross and fine motor control, etc. (e.g. Athletes) Naturalist - knowledge and ability to work with plants, animals, the environment, etc. Someone with high Naturalist intelligence might be able to survive in the wilderness well.

2. Multiple Intelligences
Intrapersonal - intra = within; ability to understand within yourself. Understand and control own self, emotions, thinking, etc.

2. Multiple Intelligences
Interpersonal - inter = between; ability to understand relationships, get along with others, etc. High interpersonal intelligence people can probably help friends solve relationship problems, help act as diplomat between people, etc. I.e. Good social skills people.

Multiple Intelligences

3. Emotional Intelligence (EQ)


Psychologists like Daniel Goleman believe we have a separate emotional intelligence. (similar to intrapersonal and interpersonal intelligences of Multiple Int.)

Some think not just IQ, but also EQ is required for us to succeed in our lives.

4. Triarchic Theory
Robert Sternberg is a psych. who thinks we have three types of intelligence:

Componential / Analytic Intelligence - usual concept of intelligence: ability to analyze problems, explain concepts, compare and contrast things, etc.

4. Triarchic Theory
Experiential / Creative Intelligence - like creativity; using intelligence in new and creative ways, finding unique solutions to problems, etc.

Practical / Contextual Intelligence - ability to apply knowledge in real life, real situations. (apply in context)

4. Triarchic Theory

4. Triarchic Theory
Contextual intelligence suggests that intelligence might partially depend on the situation / context.

This means its not just something inside us, but intelligence might depend on our environment.

4. Triarchic Theory
If this is true, it makes it very difficult to make intelligence tests because its hard to know how to test people in many contexts, or know which contexts we should test them in.

*Most intelligence tests assume that intelligence is just inside us, not contextual.

Intelligence Tests
Stanford-Binet IQ Test:

Alfred Binet (of France) tried to measure school childrens intelligence to help teachers better teach the students.

Based on the idea of mental age - higher mental age = more intelligent

Int. Tests - Stanford-Binet


Used a standardization sample to create norms, so he could test to see what mental age people were thinking at (by comparing scores to the sample)

E.g. if a 12 year old scored like a 10 year old from the standardization sample, that child might have a problem...

Int. Tests - Stanford-Binet


Louis Terman (a Stanford professor), used Binets work to develop the Stanford-Binet test, and he made the measure we call IQ

IQ = Intelligence Quotient - because we use quotients to find the IQ:

Int. Tests - Stanford-Binet


Mental Age

IQ =

100 Chronological Age

High mental age compared to chronological age will mean a high IQ Adults arbitrarily assigned a Chron. Age of 20 to avoid strange results.

Int. Tests - Wechsler


More modern IQ tests: Wechsler Intelligence Scales.

Still called IQ, even though we dont use quotients anymore....

Based on normed results arranged as a Normal Distribution.

Int. Tests - Wechsler


Remember the special properties of normal distributions:

Int. Tests - Wechsler


Standardized Wechsler scores fall on a normal distribution with the mean/average at 100 (same as Stanford-Binet score if mental age = chronological age).

The standard deviation for Wechsler distributions is 15 points.

Int. Tests - Wechsler


There are different Wechsler tests:

Wechsler Adult Intelligence Test (WAIS)

Wechsler Intelligence Scale for Children (WISC; 6-16 years old) Wechsler Preschool and Primary Scale of Intelligence (WPPSI; 4-6 years old)

Int. Tests - Wechsler


Wechsler tests produce scores on multiple sub-scales.

E.g. WAIS - 6 for verbal intelligence, 5 for performance intelligence = 11 sub-scales.

Wechsler Non-verbal Example

Int. Tests - Wechsler


If someone has a very different score on verbal and performance scores, we might suspect that the person could have a learning disability (a particular difficulty in some aspect of intelligence / learning)

E.g. Dyslexia - difficulties with reading

Bias in IQ Testing
Many psychologists worry that IQ tests could be biased towards certain types of people.

Some cultural, racial, economic factors might put some people at a disadvantage on IQ tests...

Bias in IQ Testing
For example, in the U.S., African Americans tend to score on average 10 points lower than whites. Its thought that this is because of bias in the tests, and other environmental factors (like lower average socio-economic status - poorer on average)

Bias in IQ Testing
IQ test scores do predict college grades well, but many psych. worry that BOTH IQ tests and college grades are biased to make it easier for white, middle or upper class males to do well. Tests might assume certain vocabulary, experience, familiarity with test question topics, etc. that not all people are equally familiar with (therefore bias)

Bias in IQ Testing
This means IQ tests might not really be measuring academic potential well; It could be that if the education system generally was less biased, some people who score lower on IQ tests because of bias might do better in academics than the tests predict. I.e. their potential is actually higher than the test says....

Intelligence: Nature vs. Nurture


Psychologists are curious about how much of our intelligence comes from nature (genes, biology, etc) and how much comes from nurture (parenting, experience, environment, etc).

Intelligence: Nature vs. Nurture


Heritability - a measure from 0 to 1 of how much of our intelligence comes from genetics. (0 = none, 1 = all)

Can only be used for populations, not individuals. E.g. It doesnt make sense to compare different peoples heritability...

Intelligence: Nature vs. Nurture


We cant do an experiment to test this. It would be unethical to assign some people to good environments, and some to poor environments; it would be difficult and dangerous to change peoples genetics, etc...)

But we can find some evidence from other areas:

Intelligence: Nurture
Flynn Effect - that IQ test scores have been increasing over time. Because of this, we have to re-norm our IQ tests, so that the average score remains at 100. Human genes have remained the same over this time, so Flynn Effect suggests that changes in environment are having an effect on IQ scores (e.g. nutrition, eduction, etc)

Intelligence: Nurture

Intelligence: Nurture
Interventions like help in eduction (especially early eduction) seem to be able to help raise IQ scores, at least temporarily.

This is more evidence that nurture can have an effect on our intelligence...

Intelligence: Nature
Twin studies show that identical twins (monozygotic) IQ scores tend to correlate more closely than fraternal twins (dizygotic) scores. Suggests a strong genetic component to intelligence. But remember: we cant get all the confounding variables out of these correlational studies, so we have to be very careful about the conclusions we draw....

Intelligence: Cautions
Some things to be careful / cautious about:

Important to remember that average differences between groups of people (e.g. men and women) will be much smaller than the differences between individuals within any group.

Intelligence: Cautions
Its easy to find a person from the low average group that scores higher than a person from the high average group. Average doesnt tell us about individuals...

Intelligence: Cautions
E.g. some researchers have found differences between men and women in spacial reasoning skills. However, its easy to find many women who are better than many men in this skill.

(Also, difference found might not be real: problem with test, chance in that sample, etc

Intelligence: Cautions
Also, though we use tests a lot in our lives and in psychology, we need to understand that they arent perfect. As discussed, psychologists dont agree on what intelligence is, its hard to measure, tests might be biased, etc.

So we shouldnt put too much weight on test scores like IQ tests, or other psych. tests.

Vous aimerez peut-être aussi