High-Stakes Standardized Testing [WLO: 1] [CLOs: 2, 6]

accelerated and enriched so that he might eventually be placed with the gifted and talented. Age-Equivalent Scores Age equivalents are used when norming tables convert raw scores to ages rather than grades. For example, an age-equivalent score might indicate that 6-year-old Sofia is reading at an age level of 10.2. Again, this does not mean that Sofia has the same vocabulary and reading skills as the average 10.2-year-old. It’s likely that no 10-year-olds actually took the test, so we do not know how they might have fared on what would be, for most of them, a very easy test. © 2019 Zovio, Inc. All rights reserved. Not for resale or redistribution. Interpreting Standardized Tests Section 10.2 And if Sofia were to take a standardized achievement reading test designed specifically for 10-year-olds, she might well fall off the bottom of the norming scale, no matter how clever she now seems. One important limitation of age- and grade-equivalent scores is that the differences between scores are not equal across all age and grade levels. For example, reading at a second-grade level when you are in first grade is probably far more significant than reading at an 11th-grade level when you are in 10th grade. That is because many more new reading skills and competencies develop in the earlier years than in the later grades. In much the same way, mathematics competencies tend to develop unevenly, so that differences between some grades are more meaningful than between others. For this reason, some standardized achievement test publishers no longer provide age and grade equivalents in their norming tables. Percentile Ranks Percentile ranks are another way to describe a student’s performance on a standardized test. As we saw in Chapter 9, percentiles indicate the percentage of scores that fall at or below a given point. Thus, the 75th percentile is the point at or below which 75% of all observations fall. If a student scores at the 50th percentile on a standardized test, that student’s score is exactly in the middle. Note that a score corresponding to the 40th or the 35th percentile is not a failing score; it is simply the score at or below which 40% or 35% of observations fall. Conversions Based on the Normal Curve Raw scores are often converted to one of a variety of scores that have clear meaning because they correspond to the normal distribution. As we saw in Chapter 9, the normal distribution (or normal curve) has two important characteristics that make it ideal for norming tests designed for large-scale administration—as are most commercially prepared standardized tests. 1. It is a remarkably accurate description of many naturally occurring events, so that if you plot a sufficient number of numerical values associated with random, chance happenings, they will eventually reveal a normal distribution. Even nonrandom variables such as test results often conform to a normal distribution if you collect enough of them. 2. Normal distributions have certain mathematical characteristics that allow us to predict where different proportions of scores will lie. All we need to know is where the mathematical center of the distribution is and how scores are spread around its center. Given the mean and the standard deviation of a distribution, we can determine much about the meaning of any student’s score relative to the norming population. With the same information, we can also compare the performance of entire classes—or even schools or school systems—to the performances of national or statewide groups. © 2019 Zovio, Inc. All rights reserved. Not for resale or redistribution. Interpreting Standardized Tests Section 10.2 Test makers sometimes provide their own conversions based on the normal curve by using means and standard deviations of their own choosing. Often, however, they use one or more of three common types of converted scores, referred to as standard scores (see Figure 10.1). Figure 10.1: A normal distribution with commonly used converted scores Conversion tables that accompany most standardized achievement tests might include one or more of these norms (percentiles, Z-scores, T-scores, and stanines) and perhaps age and grade equivalents as well. Z-scores, T-scores, and stanines are all standard scores with a predetermined mean and standard deviation. They are used to simplify interpretation of test results. Because raw scores on different tests vary a great deal, as do means and standard deviations, simply knowing that a person has a score of 112 or 23 or 1,115 is meaningless unless we know what the mean and standard deviation are for a comparable group on that test. Yet if these raw scores are transformed into one of the standard scores, they become highly meaningful. Z-scores are standard scores with a mean of 0 and a standard deviation of 1; T-scores have a mean of 50 and a standard deviation of 10; and a stanine (short for standard nine) score is a nine-point scale with a mean of 5 and a standard deviation of 2. The meanings of these standard scores, relative to each other, are shown in Figure 10.1. As you can see, a T-score of 80 would be an extremely high score (three standard deviations above the mean is at the 99th percentile); the equivalent Z-score would be 3. There is no exactly equivalent stanine score, because the highest score possible on this nine-point scale is 9, which is two standard deviations above the mean. Converting raw test scores to one of these standard-score scales is usually very simple because virtually all tests that use them provide conversion tables. These tables typically allow you to read the standard-score equivalent directly once you know the student’s raw score and age © 2019 Zovio, Inc. All rights reserved. Not for resale or redistribution. Views of Intelligence Section 10.3 or grade. As long as you know what the mean and standard deviation are for these standard scores, they will be meaningful. Otherwise, they are just numbers. The fact that standard scores are just numbers unless you know what they mean is one of the limitations of their use. Most parents, and many teachers, don’t know what a stanine of 8, a T-score of 65, or a Z-score of 1.5 mean—much less that they are roughly equivalent. Most people think they understand percentiles, and even more are convinced they know exactly what an age equivalent and a grade equivalent mean. Unfortunately, as we saw, their knowledge of these last two values may be misleading. 10.3 Views of Intelligence Focus Question: What are some useful, current views of intelligence? Standardized tests are available for almost any purpose imaginable. There are national police officer tests; standardized tests for admission into the ranks of accountants, lawyers, and even teachers; statewide exams for getting a driver’s license; federal and state tests for boating licenses; and standardized tests for citizenship. There are mandatory standardized tests that assess a baby’s condition at birth, and there are even standardized procedures for determining if a person is brain dead. Of particular interest in educational assessment, hundreds of different standardized tests have been designed to assess aptitude, achievement, and personality. The term aptitude, as we saw earlier, refers to a specific ability or talent. For example, mathematical aptitude refers to an inherent ability to learn mathematics—a talent for mathematics. Very closely related to the term aptitude is the term intelligence, which refers not to a specific ability or talent but to far more general abilities. One way of defining intelligence is to say that it is a talent for learning and performing that encompasses a wide variety of areas. Early approaches to understanding intelligence identified a general factor, labeled the g factor (the g stands for general), which appears to underlie intelligence. In a sense, the g factor accounts for the generally high correlation among different cognitive tasks. What this means is that individuals who do well on one cognitive task are likely to do well on others. This high correlation among the different abilities measured by most intelligence tests is what accounts for the g factor (Eid, Geiser, Koch, & Heene, 2017). IQ scores are sometimes interpreted as being a measure of an individual’s g factor. But that is only one of many approaches to intelligence. We look at several more here. Boring’s Definition of Intelligence Intelligence, said Boring (1923) “is what the tests test” (p. 35). In other words, we do not really know how to define intelligence, but we have intelligence tests that correlate relatively highly with success on tasks we think require intelligence—such as doing well in school. Therefore, © 2019 Zovio, Inc. All rights reserved. Not for resale or redistribution. Views of Intelligence Section 10.3 even if we cannot agree on what intelligence is, we might agree that we can actually measure it. So, intelligence is what the tests test. Still, this is not an especially productive or useful definition. Definitions and explanations that look at the actual processes, capabilities, and actions involved in intelligent behavior are more practical. Cattell: Fluid and Crystallized Abilities Intelligence, explains Cattell (1971), involves two different kinds of capabilities. On the one hand are abilities that seem to underlie much of our intelligent behavior, termed fluid abilities. Fluid abilities are involved in problem solving, representing symbolically, manipulating, organizing, and learning. These abilities are essentially nonverbal and relatively unaffected by culture or experience. They are evident in speed of processing, general reasoning, memory, and attention span. As neurological functioning deteriorates with old age, so too do fluid abilities (McDonough et al., 2016). In Cattell’s second group of abilities are the crystallized abilities. These are mostly verbal, and they are heavily influenced by culture, experience, and education. Crystallized abilities are evident in vocabulary tests, tests of general information, and arithmetic skills. Because these abilities depend on the acquisition, storage, and recall of information, they tend to increase with increasing experience. As a result, unlike fluid abilities, crystallized abilities may increase with age, sometimes even into very old age (Yuan, Voelkle, & Raz, 2018). See Table 10.3. Table 10.3: Cattell’s theory of fluid and crystallized intelligence Crystallized abilities Fluid abilities Description Mainly verbal Mainly nonverbal Examples Extensive vocabulary; mathematical skills; large store of information Quick reaction and decision time; logical reasoning skills; abstract thinking skills Effect of experience Heavily influenced by experience, culture, education Relatively unaffected by experience Age-related changes May increase into old age Decline with deteriorating neurological functioning Tests that tap relevant abilities Vocabulary tests; tests of arithmetical skills; tests of general information Speed of recognition tests; reaction time measurer; attention and memory tests © 2019 Zovio, Inc. All rights reserved. Not for resale or redistribution. Views of Intelligence Section 10.3 Sternberg: Successful Intelligence The most useful approach to understanding intelligence, says Sternberg (1996, 2015b), is to focus on people who are successful in adapting to the world. If we do that, we find that these intelligent people not only learn well, they are also exceptionally skilled at selecting and shaping their environments. These people have successful intelligence. Note how this view emphasizes adaptation. In effect, explains Sternberg, the most intelligent individuals are those who exercise control over their environments by selecting or changing them as well as by modifying their own behavior. Thus, a successfully intelligent student not only selects classes and programs that fit her interests and abilities, but also strives to improve relevant skills and knowledge to maximize her success (her adaptation). Successful intelligence is evident in how well the individual is able to evaluate options, devise strategies for attaining goals, and assess the extent to which goals are being met. In other words, successful intelligence requires analytical abilities. These allow the individual to evaluate options, devise strategies, and monitor progress in adapting. Analytical abilities involve judging, evaluating, contrasting, comparing, and analyzing. Creative abilities are also essential to successful intelligence. These allow the individual to discover options, generate ideas, and try new ways of selecting and shaping the environment and of adapting to it. Creative abilities are apparent in activities such as discovering, imagining, inventing, and supposing. Finally, successful intelligence requires the practical abilities that are necessary for carrying out options. Without a high level of practical abilities, it is difficult to put into practice the behaviors and skills that are involved in selecting, shaping, and adapting to environments. Analytical, creative, and practical skills are the three components of successful intelligence. They make up Sternberg’s triarchic theory of successful intelligence. This theory suggests that conventional measures of intelligence might not be as useful as educators had thought. That’s because they seldom measure either creative or practical abilities. Thus, they might provide a very misleading picture of this thing called successful intelligence (see Figure 10.2). © 2019 Zovio, Inc. All rights reserved. Not for resale or redistribution. Views of Intelligence Section 10.3 Figure 10.2: Sternberg’s triarchic theory of successful intelligence Sternberg’s view of intelligence stresses adaption and suggests that our current measures tap only part of what successful intelligence really is. Sternberg’s view of successful intelligence has the advantage of making the concept less abstract; it brings it down to a more concrete and more easily understood level. Perhaps even more important, it emphasizes that intelligence is not a fixed, unchanging quality that we cannot alter. It echoes the contemporary view that intelligence is highly malleable rather than fixed—it isn’t something that we have just so much of, and that’s it. Gardner’s Multiple Intelligences Sternberg’s model describes intelligence as an information processing activity—an activity that requires analyzing, being creative, and putting intentions into practice. This informationprocessing view contrasts sharply with earlier views that saw intelligence as a mysterious something that some people had a whole lot of and others, not so much. In Sternberg’s view, intelligence is a little like a box filled with the tools we use to play the game of cognition. We might not all have exactly the same tools in our boxes; our genes have seen to that. But we can certainly learn to take better advantage of the ones we have. And we might also benefit from seeing what tools others use and how they use them. Some of us have rusty old tools we have never used. © 2019 Zovio, Inc. All rights reserved. Not for resale or redistribution. Views of Intelligence Section 10.3 Ask people how smart they are, says Gardner (Chen & Gardner, 2005), and many will hesitate. They don’t quite know how to answer. “It depends,” they say. Maybe it depends on whether you are asking about how easily they can remember facts and figures or how easily they can solve problems. Or perhaps it depends on whether you are asking them how smart they are compared to their peers or to neurosurgeons and rocket scientists or to a family member. Yet what if instead of asking, “How smart are you?” you ask, “How are you smart?” The answer might be immediate and highly revealing. Now people say, “Hey, I never forget a name or a face,” or “I’m a whiz at math, but I’m not so good at reading people,” or “I’m a great storyteller but I’m not so hot at writing stuff.” Gardner’s theory of multiple intelligences says that we have not one but at least eight largely unrelated kinds of intelligence: logical-mathematical, linguistic, musical, spatial, bodily kinesthetic, interpersonal, intrapersonal, and naturalistic. Gardner (2006b, 2016) also considered the possibility that we might have additional types of intelligence—such as moral intelligence, spiritual intelligence, and an aptitude for humor— but concluded that these traits do not meet the criteria for distinct intelligences, although they might be included in a broad category called existential intelligence (Gardner, 1999). The eight distinct kinds of intelligence are explained in Table 10.4. Table 10.4: Gardner’s multiple intelligences Gardner’s eight kinds of intelligence Main characteristics Possible occupations Logical–mathematical intelligence Ability to think rationally, to reason logically Scientist, mathematician, accountant, computer programmer Linguistic intelligence Ability to produce and understand oral and written language Orater, poet, journalist, textbook writer, newscaster Musical intelligence Appreciation of and ability to produce musical forms Composer, musician, conductor, pop star Spatial intelligence Ability to visualize and manipulate mental representations Graphic designer, architect, artist, draftsman, builder Bodily kinesthetic intelligence High physical coordination and body control Dancer, athlete, choreographer, bullfighter, fashion designer Interpersonal intelligence Ability to detect and understand emotion in self and others and respond appropriately Teacher, psychologist, coach, professor, con artist Intrapersonal intelligence Deep understanding of personal strengths, weaknesses, and emotions Guru, writer, philosopher, theorist, visionary Naturalistic intelligence Ability to detect patterns and organization in nature Biologist, evolutionary theorist, archaeologist, ecologist, naturalist We don’t assess multiple intelligences very well, note Chen and Gardner (2005). Most of our intelligence tests tend to focus on mathematical, linguistic, and logical tasks because we’ve © 2019 Zovio, Inc. All rights reserved. Not for resale or redistribution. Views of Intelligence Section 10.3 assumed that these sorts of tasks are most closely related to intelligent behavior. However, these tasks are unrelated to our kinesthetic, intrapersonal, interpersonal, naturalistic, or musical aptitudes. Basically, the Gardner view of multiple intelligences requires the development of a range of new intelligence tests to tap these neglected competencies—most of which are not easily measured by means of paper-and-pencil or computer-based tests. In addition to its implications for testing intelligence, this highly popular theory has important applications in education. Gardner (2006b) argues that schools need to pay far more attention to what and how they teach. A guiding principle for approaches to instruction based on the theory of multiple intelligences suggests that children should have each of their intelligences stimulated every day. A well-rounded curriculum would include daily opportunities for every student to participate in music, bodily kinesthetic activities (sports and dance, for example), social activities, and activities requiring spatial skills, as well as a range of more traditional activities involving logical, mathematical, and linguistic skills. Schools, Gardner (2006a) explains, are facing tremendous challenges. They need to be concerned with the kinds of minds people will need if they are to be prepared for the future. In Gardner’s own words from a recent interview: In the future, individuals who succeed will need to have a disciplined mind (knowing how to become an expert in one or more fields), a synthesizing mind (which can put together disparate information in a useful way), a creating mind (raising new questions and solving difficult challenges), a respectful mind (being able to deal well with individuals near and far), and an ethical mind (handling the challenges that arise in the workplace and in one’s role as citizen, where one needs to behave in a moral way). (Sole 24 Ore, 2015) Gardner’s view of multiple intelligences is highly compatible with the contemporary belief that the most useful assessments are those that identify important differences among learners and that provide information (feedback) that serves to enhance each learner’s progress and achievements. That, as we saw in Chapter 5, is what is meant by differentiated assessment and differentiated instruction. Differentiated approaches to learning and instruction are especially important for children with special needs—a group that includes not only children who experience significant difficulty in learning and ELL students, but also those with remarkable gifts and talents. Dweck’s Mindsets: Entity Versus Incremental Theories of Intelligence Some people, Dweck (1986, 1999) informs us, view intelligence as fixed and unchanging— something that certain people possess a great deal of and others, not so much. These people are said to have an entity theory of intelligence; they view ability as an unchanging entity. © 2019 Zovio, Inc. All rights reserved. Not for resale or redistribution. Views of Intelligence Section 10.3 Others see intelligence as something that is malleable. They are said to have an incremental theory of intelligence; they view it as a thing that can be increased with effort and determination. These two ways of viewing intelligence—Dweck (2006) calls them mindsets—have tremendously important implications for effort and achievement, both in education and in other endeavors. Our mindsets regarding intelligence shape our attitudes, goals, efforts, and ultimately, behaviors and achievements. Those who believe that intelligence does not change—who have a fixed mindset—often go out of their way to demonstrate that they have a lot of this thing they view as intelligence. As a result, they tend to avoid difficult tasks and challenges that might expose their weaknesses, much preferring undertakings where success is highly likely. In contrast, those who believe that intelligence can be increased—who, using Dweck’s terms, have growth mindsets—are convinced that with dedication and effort, they can develop their skills and abilities and increase their personal competence. Research lends considerable support to Dweck’s belief that these mindsets can determine important outcomes. Growth mindsets, explain Dweck and Yeager (2018), lead students to seek challenges that foster learning: Their ultimate goal is to increase their competence and to master learning objectives. In contrast, goals associated with fixed mindsets typically reflect the confidence individuals have in their abilities. Those who are highly confident and think they have a generous amount of this thing called intelligence are more likely to undertake difficult and challenging tasks than those who don’t believe they possess very much intelligence. In the face of failure, both those who are highly confident and those less confident are likely to feel helpless, not persist when success seems unlikely, and avoid future challenges. Dweck argues that schools and parents need to consider interventions that focus on developing growth mindsets. These, she insists, have tremendous potential for increasing academic performance and for reducing feelings of inadequacy and helplessness that are often associated with fixed mindsets. Unfortunately, many children grow up with the notion that intelligence is fixed and that little can be done to change it. However, considerable research supports the belief that growth mindsets can be fostered in various ways, including workshops, tutoring sessions, online programs, and other interventions that emphasize how brains develop, grow, and change (e.g., Yeager et al., 2016). These interventions typically underline that it is especially important not to repeatedly praise children for being intelligent—an act that implies that intelligence is fixed. It is even more important not to berate children for apparent shortcomings, which also suggests the unchanging nature of intelligence. Instead, as Haimovitz and Dweck (2017) suggest, the most effective approach is to direct praise at efforts and strategies that are successful. One of the fundamentally important tasks of educators is to provide children with challenging tasks at which, with effort and guidance, they will succeed.