Defining Intelligence PDF

# Defining Intelligence - Is not easy. - Some attempts - The mental abilities that enable one to adapt to, shape, or select one's environment. - The ability to deal with novel situations. - The ability to judge, comprehend, and reason. - The ability to understand and deal with peop...

# Defining Intelligence - Is not easy. - Some attempts - The mental abilities that enable one to adapt to, shape, or select one's environment. - The ability to deal with novel situations. - The ability to judge, comprehend, and reason. - The ability to understand and deal with people, objects, and symbols. - The ability to act purposefully, think rationally, and deal effectively with the environment. - None are fully satisfactory. - Those terms need definitions themselves. - Use of terms such as: - Judge - Understand - Comprehend - Think rationally. It would be nice if we could say that psychologists first analyzed *learning*, *cognition* and *memory* and then built upon that knowledge to understand intelligence. - Psychological researchers began by testing the ability to do well in school and then conducted research to discover what the tests measure. - It may seem odd to measure something before defining what it is. - The same is true in other fields. ## Physicists measured: - Gravity - Magnetism - Electricity - And other forces before studying what they were. ## Spearman’s Psychometric Approach and the "g" factor - First program in psychology was Charles Spearman’s (1904) psychometric approach to intelligence, based on the measurement of individual differences in performance. - Measured how well people performed tasks such as: - Following directions - Doing arithmetic - Judging musical pitch - Matching colors - Founded that performance on any tasks correlated positively with performance on any of the other tasks. - Therefore, inferred that all tasks have something in common: - To perform well on any test of mental ability, argued that people need a “general” ability, which he called *g*. - The symbol *g* is always italicized and lowercase, like mathematical terms *e* (the base of natural logarithms) or *i* (the square root of -1). - To account for the fact that performances on various tasks do not correlate perfectly, suggested that each task also requires a “specific” ability (*s*). ## Conclusion - Thus, intelligence consists of a general ability plus an unknown number of specific abilities, such as: - Mechanical - Logical - Musical - Spatial - Arithmetical - His theory called “monarchic” theory of intelligence because it included a dominant ability (*g*) or monarch that ruled over the lesser abilities. According to Spearman (1904), all intelligent abilities have an area of overlap, which he called “g” (for “general”). Each ability also depends on an “*s*” (for “specific”) factor. ## Later researchers confirmed that: - Scores on virtually all kinds of cognitive tests correlate positively with one another within almost any population, including non-Western nations. - Noticeable trend: a student who does well in one course generally does well in others. - Only under special conditions do most of the individuals with high scores on one test get low scores on the other. For example, in one study, rural Kenyan children who did well on an academic test did poorly on a test of knowledge about traditional herbal medicines, and those who did well on the test of herbal medicines did poorly on the academic test. - Evidently, the two groups of children had been exposed to different experiences. - Intelligence consists of a general ability plus an unknown number of specific abilities. ## Explanations for *g* - Why do people who perform well on one type of test generally perform well on others also? - The simplest interpretation is that all the tasks measure a single underlying ability. - For example, most people who excel at running a 100-meter race also do well at the high jump and the long jump. - They have to, because all three events depend on the same leg muscles. - Measurements of sprinting, high jumping, and long jumping correlate with one another because they all depend on the same leg muscles. Similarly, the *g* factor that emerges in testing could reflect a single ability. - Similarly, perhaps people perform well on a variety of intellectual tests because all the tests depend on one underlying skill. - Underlying skill could be working memory capacity or processing speed, which correlate highly with each other and with cognitive performance. - One possibility is the efficiency of mitochondria that generate energy in cells. Greater efficiency of mitochondria means greater potential for brain activity, needed to take specialized skills seriously. - Some people, especially those with brain damage or genetic mutations, show impairments on one type of task while performing normally on others. A similar idea could be true for intelligence: - Perhaps we have many abilities. - Every cognitive task uses a combination of them. - That any pair of tasks uses at least a few overlapping abilities. - The results would be positive correlation between any pair of tasks, even without a single general factor. - Another possible explanation for *g* is that several types of intelligence correlate because they grow in the same ways. - By analogy, consider three body parts. The rule is: most people with a long left leg also have a long right arm and a long left index finger. The factors that control the size of such body parts are genetics, nutrition, and health. - Similarly, all forms of intelligence depend on genes, health, nutrition, and education. - Most people who have good support for developing one intellectual skill also have good support for developing others. - Do the various intellectual skills correlate with one another because they all measure a single underlying ability or because any two tasks use a few overlapping abilities, or because all abilities grow together? ## Hierarchical Models of Intelligence - Suggested the existence of *s* factors, but task fails when other psychologists try to describe them. - Intelligence as a hierarchy of components: - Language - Perceptual processing - Spatial relationships - They are language, short-term memory, and reasoning. ## Fluid intelligence - Is the power of reasoning and using information, and solving new problems. - Enables you to learn skills in a new job. - Peaks before or soon after age 20. ## Crystallized intelligence - Consists of acquired skills and knowledge and the ability to apply that knowledge in specific situations. - Peaks in middle age and remains nearly constant. - Includes the job skills you have already acquired. The difference between fluid and crystallized intelligence is sharper in theory than in practice, as almost any task taps some combination to fluid and crystalized intelligence. ## Results are sometimes misleading when: - Older people are not sufficiently motivated to perform on seemingly unimportant tests. ## Gardner's Theory of Multiple Intelligences - Howard Gardner found and argues: - Multiple Intelligences: independent forms of intelligence. - Such as: - Languages - Musical Abilities - Logical - Mathematical reasoning - Spatial reasoning - Self-control - Self-understanding - Sensitivity to other people's social signals - Someone seems intelligent in *one* way, be mediocre or worse in *another* because different skills require not only different kinds of practice but also different brain specializations. - Different skills should be valued, different abilities are independent and unrelated. ## IQ Test - Analogy doing sports and choosing the best ones. - Aptitude: is the ability to learn, or fluid intelligence. - Achievement: is what someone has already learned, or crystallized intelligence. ## Additionally - Aptitude and achievement are hard to separate because aptitude leads to achievement, and past achievement increases future ability to learn. - The first IQ test was created by Alfred Binet and Theophile Simon (1904). - French Ministry of Public Instruction wanted a fair way to identify who had such serious intellectual deficiencies that they needed to place special classes, *without* decision to a teacher's opinion. - Measured the skills that children need for success in school: - Counting - Remembering - Following instructions - Understanding language - Stanford Psychologists translated *English*; they changed the emphasis to identifying the best students who would profit from accelerated classes. ## Objective tests help identify students with good abilities, including some whom their teachers had overlooked or underestimated. ## Intelligence quotient (IQ) tests: - Use procedures that try to predict someone’s performance in school and similar settings. - The term *quotient* originated when IQ was determined by dividing mental age by chronological age and then multiplying by 100. - Mental age reflects the average age of children who perform as well as a specific child. - For example: an 8-year-old who performs like an average 10-year-old has a mental age of 10, 10/8 x 100 = 125 ## The Stanford-Binet Test | Age | Sample Test Items | |---|---| | 2 | Test administrator points at pictures of everyday objects and asks, “What is this?” “Here are some pegs of different sizes and shapes. See whether you can put each one into the correct hole.” | | 4 | “Why do people live in houses?” “Birds fly in the air; fish swim in the.” | | 6 | “Here is a picture of a horse. Do you see what part of the horse is missing?” “Here are some candies. Can you count how many there are?” | | 8 |“What should you do if you find a lost puppy?” “Stephanie can’t write today because she twisted her ankle. What is wrong with that?” | | 10 | “Why should people be quite in a library?” “Repeat after me: 483714.” | | 12 | “What does regret mean?” “Here is a picture. Can you tell me what is wrong with it?” | | 14 | “What is the similarity between high and low?” “Watch me fold this paper and cut it. Now, when I unfold it, how many holes will there be?” | | Adult | “Make up a sentence using the words celebrate, reverse, and appointment.” “What do people mean when they say, ‘People who live in glass houses should not throw stones’?” | ## Adaptive Testing - Testing method in which the range of items used is adapted to the performance. - Provides overall IQ score, Verbal and Nonverbal IQ scores. - Subs-scores reflecting visual reasoning, short term memory, and other skills. - Modified, but similar to the original Binet Simon test. - Items for children designated age. ## The Wechsler Test - Produces same average and almost same distribution of scores as Stanford-Binet. - Test overall IQ: Verbal, performance, comprehension, and processing speed. - Ex: of a processing speed item is, "Put a slash (//) through all circles and an X through all the squares, as quickly as possible." - Calls attention to *strengths* and *weaknesses*. - Ex: a child who learned English as a second language might to best on items that call for *nonverbal* answers. ## Culture-Reduced Testing - Educators can use either test to identify possible learning disabilities. - Furthermore, a translated item may be easier or harder than the original item. - Ex: some items refer to information that is more familiar in one culture than other. ## Progressive Matrices - Most widely used culture-reduced test. - It attempts to measure *abstract reasoning* (**fluid intelligence**) without use of language or factual information. - **Advantage**: The test is fairer for someone who is not a native speaker of English. - **Disadvantage**: It provides only a single score instead of identifying someone's strengths and weaknesses. ## Individual Differences in IQ Scores - Francis Galton (1869/1978) was the first to argue for the importance of heredity. - His evidence was that politicians, judges, and other eminent and distinguished people generally had distinguished relatives. - **Family Resembles**: - Score of monozygotic (identical”) twins correlate with each other about 0.85, significantly higher than dizygotic twins or non-twin siblings. - Similar brain volume and in *working memory*, *attention, reading, and mathematics*. - The greater similarity between monozygotic than dizygotic twins implies a genetic influence, although it may overstate the genetic impact, because monozygotic twins are more likely to be treated the same. - **Correlation of IQ scores for people of various degrees of genetic relationships**. - In most cases, their environments were separate but similar. - Identical twins continue to resemble each other throughout life, even beyond age 70. - Similar IQ growing older. - Older individuals have more control of their environment. - People who start with an intellectual advantage gravitate toward activities that sustain and increase that advantage. *Called multiplier effect*. ## Twins and Single Births - Twins resemble each other more closely than other siblings. - Higher correlation between the IQs of brothers born within a couple of years of each other than those born further apart. ## Adopted Children - Correlation is lower than the correlation between biological brothers or sisters. - IQs of young adopted children correlate moderately with those of their adoptive parents. - Growing older, their IQ scores gradually correlate more with those of their biological parents. - Unrelated children of the same age growing up in the same family. - For example, parents might adopt two children of the same age, or have a child of their own and adopt another of the same age, or have a child of their own and adopt another of the same age. ## Gene Identification - Some low-IQ parents who put their children up for adoption are impoverished and probably do not provide good prenatal care. - Poor prenatal care correlates with decreased IQ for the offspring throughout life. - Adopted children can resemble their biological parents for non-genetic as well as genetic reasons. - Young children correlate with biological parents. - Genetic variations reportedly correlate with measures of intelligence, but no common variant has a large effect. - Intelligence depends on many genes making small contributions and epigenetic influences. - How genes are expressed rather than differences in the genes themselves are key factors. ## Environmental Influences - Important about is especially apparent when children from a low-quality orphanage get adopted into a good home, or when a family with small children moves to a more prosperous country. - IQ scores can increase substantially. - Improved environment has the greatest effect on the youngest children. - Still, the finding is that mean IQ is consistently lowest in the countries and states where children have the highest exposure to infectious diseases such as: - Tetanus - Malaria - Tuberculosis - Hepatitis - Cholera - Measles - Education also promotes intelligence. ## Brain Size and Intelligence - Human brain volume correlates about 0.24 IQ scores, according to one study; about 0.19 according to another, larger study. - The sibling with the larger brain also shows higher intelligence, with a correlation of 0.18. - On average, men have larger brains than women, but women are equal to men in intelligence. - Women have more deeply folded cerebral cortex, so surface area is nearly equal. - Men and women have approximately the same number of neurons. - Other animals have larger or small brains, but none have more neurons. - Humans have more total neurons than any other species, having more than monkeys and more neurons than elephants or whales because their neurons are larger and less numerous. ## Standardizing IQ Test - Standardization: the process of evaluating questions, improving ambiguous items, and establishing rules for administering a test and interpreting the scores. - Norms: descriptions of how frequently various scores occur. - Standard deviation: measures variance among individuals. Most scores are close to the mean. ## The Distribution of IQ Scores - The authors determine the mean and the distribution of scores for a random or representative sample of population. - IQ tests have a mean of 100 and a standard deviation of about 15 or 16, depending on the test. - The mode (most frequent score) is higher than 100, and a bulge of lower scores exists. ## The bulge at the lower end represents people with disabilities, described as mentally challenged. - For example, people with Down syndrome have a variety of impairments as a result of having an extra copy of chromosome #21 generally including impairments in speech development, memory, motor skills, and cognition. - Some vary and some have IQ scores close to the mean. - The term *mentally challenged* or *mentally disabled* refers to people *more* than two standard deviations below average, depending on the test, but this is arbitrary. - An IQ of 130 or more is *gifted range*. ## The Flynn effect - Pattern noted by James Flynn that decade by decade, generation by generation, people’s raw scores on IQ tests gradually increased, requiring test makers to make the test harder. - The results vary, but a typical figure is a gain of about three IQ points per 10 years. - The size of the Flynn effect – that is, the rate of increase in IQ scores over time. - Hypotheses suggest Flynn effect may be caused by: - Improved education - Increase cognitive stimulation - Decrease in mental retardation - Increased tendency for people to marry outside their own neighborhood. - It is argued that the increase in IQ scores does not indicate equally increase intelligence over time. ## Evaluating Tests ### Reliability - Repeatability of its scores. - Calculated with a correlation coefficient. - Measures how accurately we can use one measurement to predict another, on a scale from 0 to 1. - Compare two tests: first and second halves of the test or the scores on the test’s odd-numbered and even-numbered items, called *test-retest reliability*; the correlation between scores on a first test and a retest. - On a test with high reliability, people get similar scores each time they take the test. On a test with low reliability, scores fluctuate randomly. ## Validity - Defined as the degree to which evidence and theory support the interpretations of test scores for the intended purposes. - Indicates how well the test measures what it claims to measure. To determine the validity of a test, researchers examine several types of evidence. - **Content**: the content of a test should match its purposes. - **Response process**: If a test claims to measure a certain skill, then the test takers should need to use that skill instead of using shortcuts. - **Usefulness for prediction**: Most importantly, if a test is valid, the scores predict some performance. - **Consequences of testing**: Test produce benefits, but also certain unintended consequences. - Measuring the validity of a test can be difficult. ## Interpreting Fluctuations in Scores - When tests are not perfectly reliable, your scores fluctuate. The lower the reliability, the greater the fluctuation. - When people lose sight of this fact, they sometimes draw unwarranted conclusions. - If you had selected infants with the lowest scores and then said “hocus pocus” and tested them again a few days later, most of their scores would have improved, simply because they had nowhere to go but up. - Similarly, for those with the lowest scores on the first test, we can predict that, on average, their scores will improve. This tendency is called **regression to the mean.** When tests are not perfectly reliable, scores fluctuate; lower reliability = greater fluctuation. Regression to the mean. IQ tests for infants have low reliability—the scores fluctuate widely. ## Measuring Test Bias - Bias is the tendency for test scores to overstate or understate the true performance of one or more groups. - If one group scores better than another, the difference by itself does not necessarily indicate bias. - Need to determine the bias, or lack of it, for any potential use of any test. ## Evaluating Possible Bias in Single Test Items - SAT test was very biased against women of question. A bias test misestimates the performance by members of some group. Ex: If an IQ test is biased against black Students’, then black students who score, say, 100, will do better in school than white Students with the same score. Remove items that tend to be easy for one group of people to answer but difficult for another group. Evaluate whether the whole test makes equally accurate predictions for all groups. Poverty, poor prenatal health, distractions, and worry are other explanations for differences in test scores. ## Stereotype Threat - Is a person’s perceived risk of performing poorly and thereby supporting an unfavorable stereotype about their group. ## Ways to Combat Stereotype Threat - Tell people about stereotype threat before the test and urge them not to let the stereotype bother them. - Writing interventions. ## Hypothesis - If black Students believe they’re taking the kind of test on which black students on average do not perform well, then they worry that their own performance may reflect poorly on their group.