Individuals with Disabilities Education Act

bundet

model, this form of identiﬁcation has been criticized because it does not lead to interventions and the fact that it relies too heavily on IQ test scores. In its place, the local education agency may use a process that deter- mines if the child responds to scientiﬁc, research-based intervention as a part of the evaluation procedures. This process, also known as Response to Intervention, is not without its critics as well. As a result, the process for identifying learning disabilities will continue to be highly debated. In addition to modifying identiﬁcation procedures, IDEAwasmadetocloselyaligntoanotherfederaledu- cation law, No Child Left Behind (NCLB). In accord with NCLB, special education teachers must now be highly qualiﬁed. In this inst ance, ‘‘highly qualiﬁed’’ means that all special education teachers must have a state special education certiﬁcate and may not hold an emergency, temporary, or pr ovisional certiﬁcation. In addition, if these teachers are responsible for content courses, they must be licensed in the subjects taught. Eligibility for Services Under IDEA To be eligible for services under IDEA, a child must have a speciﬁc disability that has an adverse effect on the student’s performance. The following disabilities are recognized by IDEA: Autism Deaf-blindness Deafness Emotional disturbance Hearing impairment Mental retardation Orthopedic impairment Other health impairment (e.g., attention deﬁcit hyper- activity disorder) Speciﬁc learning disability Speech or language impairment Traumatic brain injury Visual impairment Evaluation for IDEA Services An evaluation for services under IDEA consists of procedures to determine whether a child has a disabil- ity. During this procedure, the child must be assessed in all areas related to the suspected disability. In general, a comprehensive assessment consists of the following: 1. Individual psychological examination (e.g., general intelligence, academic func tioning, social emotional functioning, and learning strengths and weaknesses), 2. Social history based on parent and teacher interviews, 3. Academic history based on teacher and parent interviews, 4. Classroom observation of the student, 5. Functional behavioral assessment to determine the relationship between the child’s performance and the variables that are related to its occurrence, 6. Other assessments as necessary depending on the spe- ciﬁc disability evaluation category (i.e., speech lan- guage evaluation, occupational therapy evaluation, bilingual assessment, audito ry and visual assessment). Parental Consent Before any assessment or student placement, parents must be informed and consent to the evaluation. This is done to protect the legal rights of parents and their children. Parental consent consists of the following: 1. The parent has been informed of all information relevant to the activity for which consent is sought. 2. The parent agrees in writing to the proposed assess- ment procedures. 3. The parent is notiﬁed that consent is voluntary and can be revoked at any period during the assessment procedure. Individualized Education Program The Individualized Educ ation Program (IEP) is a written document that details the educational objec- tives and expectations for students with disabilities. In accordance with IDEA, speciﬁc goals must be devel- oped based on the strengths of the child, the concerns of the parents, and the results of the most recent evalu- ation.MandatoryinformationinanIEPincludesthe following: 1. A statement of the child’s present levels of aca- demic achievement and functional performance; 2. A statement of measurable academic and functional annual goals; Individuals with Disabilities Education Act 521 3. A description of how the child’s progress toward meeting the annual goals will be measured; 4. Periodic reports on the progress the child is making toward meeting the annual goals; 5. A statement of the special education, related services, and supplementary aids with which the student will be provided; 6. An explanation of the extent, if any, to which the child will not participate with children without disabilities in the general education classroom; 7. A statement of any accommodations that are neces- sary to measure the academic achievement and functional performance of the child on state and district-wide assessments; 8. If the IEP Team determines that the child must take an alternate assessment instead of a particular regular state or district-wide assessment of student achievement, a statement of the reasons for that determination; and 9. The projected date for the beginning of the services, in addition to the anticipated frequency, location, and duration of these special education services. Individual Education Program Team The Individual Education Program (IEP) team is a multidisciplinary team that is concerned with deter- mining if a child has a disability, and if so, the services that are necessary to ensure a free and appro- priate education. This team is mandated by IDEA so that the child is guaranteed that any evaluation is comprehensive and conducted by different profes- sionals to decrease subjectivity. In accordance with federal mandates, the team must use a variety of assessment measures and gather information from multiple sources. Members of the multidisciplinary team generally include the following: parents of a child with a disability; general education teacher; special education teacher; school psychologist; school social worker; guidance counselor; school nurse; phy- sician; parent advocate; and whenever appropriate, the child with the disability. Least Restrictive Environment As a result of the passage in 1975 of P.L. 94–142, all children with disabilities were guaranteed the right to be educated in the least restrictive environment. The least restrictive environment refers to the fact that children with disabilities should be placed in the envi- ronment that is best suited for their educational needs. By law, students with disabilities should be educated with children without disabilities to the greatest extent possible. Transition Services Transition services are a coordinated set of activi- ties designed to provide students with the practical skills and knowledge that will allow them to transition to adulthood. Beginning when the child turns 16, an Individual Transitional Education Plan (ITEP) must be in place for students with disabilities. The ITEP must include the following: 1. A statement of measurable postsecondary goals related to training, education, employment, and independent living skills; 2. A statement of long-term outcomes for the student; and 3. A list of participants involved in the planning and development of the individualized transitional edu- cation program. Importance Before its inception, millions of children with disabil- ities were denied a free and appropriate education, and as a result, many students were unable to reach their full potential. From educating all children regardless of their disability to providing transition services to adulthood and changing how society views the capabilities of individuals with disabilities, this legislation has changed the educational landscape. Scott L. Graves, Jr. See also Disabilities; Individualized Education Program; No Child Left Behind; Special Education; Standardized Tests Further Readings Batshaw, M. (Ed.). (2002). Children with disabilities (5th ed.). Washington, DC: Brookes. Council for Exceptional Children. (2004). The new IDEA. Arlington, VA: Author. Education for all Handicapped Children Act of 1975, 20 U.S.C. 1400 et seq. Individuals with Disabilities Education Improvement Act of 2004, P.L. 108–446, 20 U.S.C. 1400 et seq. President’s Commission on Excellence in Special Education. (2002). A new era: Revitalizing special education for children and their families. Washington, DC: Author. 522 Individuals with Disabilities Education Act Smith, T. (2005). IDEA 2004: Another round in the reauthorization process. Remedial and Special Education , 26 , 314–319. Telzrow, C., & Tankersley, M. (2000). IDEA amendments of 1997: Practice guidelines for school-based teams. Bethesda, MD: National Association of School Psychologists. I NDUCTIVE R EASONING Inductive reasoning, or induction, refers to inferences from evidence that are more or less plausible; a good inductive inference is likely to be true. In contrast, the conclusions of deductive inferences are guaranteed to be true by the truth of the premises upon which they are based. Deductive inferences occur in mathemati- cal or logical contexts; almost all other judgments involve induction. Consider a gardener wishing to buy fencing for a square plot 18 feet on a side. Concluding that the perimeter of the plot equals 72 feet is deductive. What it means to be 18 feet on a side is to have a 72 feet perimeter. This deductive inference is an important piece of solving the problem. Ultimately, however, the gardener has to make an inductive judgment about how much fence he or she will need. Constructing a fence always involves some amount of waste and obstructions to be fenced around. In planning a fencing purchase, it is wise not to take the premises as conclu- sively true; the plot is unlikely to be exactly 18 feet, nor perfectly square. In a mathematics classroom, the problems demand deduction; outside, in the garden, the hard problems demand induction. Because almost every judgment has some element of induction, the study of induction is a rich and var- ied enterprise. Questions about the nature and justiﬁ- cation of inductive judgments have a long history in philosophy. In psychology and education, there are two major approaches: a focus on form and a focus on content. These perspectives have led to different pictures of the development of induction and different approaches to teaching good inductive practices. From a formal perspective, induction involves assessment of probability or association. A weather forecaster’s statement that ‘‘there is a 70 % chance of rain tomorrow’’ reﬂects the association between cur- rent conditions and future rain: On 70 % of the days like today, it has rained tomorrow. Statistics and probability theory provide formal means for calculat- ing association and probability (P). Bayes’s theorem deﬁnes the probability that a hypothesis is true given some data (e.g., that it will rain tomorrow given today’s conditions) in terms of the ‘‘prior probabili- ties’’ of the hypothesis and the data (how likely rain is on any day, how common are today’s conditions) and the ‘‘likelihood’’ of the hypothesis—the probabil- ity of the data when the hypothesis is true (how often we see today’s conditions when it actually does rain the next day). P(hypothesis | data) = P(hypothesis) * P(data | hypothesis)/P(data) : Probability judgments are a basic element of psy- chology. Even traditional learning theory (behaviorism) assumes that learners can calc ulate associations. Ani- mals and infants are able to learn quite complex pat- terns of probabilities. Whether people’s intuitive judgments conform to or violate axioms of probability theory is a major focus of debate. One inﬂuential theory is that people reason with heuristics that are efﬁcient and effective given the environments typically encoun- tered. The cognitive abilities underlying valid statistical inference may require particular environmental support (e.g., schooling) or developmental achievements (e.g., Piaget’s formal operations). Some psychologists have proposed that people should b e explicitly taught good principles of statistical reasoning. In education, a formal approach to inductive infer- ences has often focused on teaching the scientiﬁc method, such as control of variables in experimenta- tion. More generally, critical thinking is a set of skills for evaluating arguments based on formal principles of logic. Students who learn to construct valid experi- ments, recognize confounded evidence, uncover assumptions, and avoid contradictions will make bet- ter inductive inferences. Formal principles are independent of the particular content (weather, gardens) that is the subject of judg- ment. Inductive arguments are often characterized as empirical; they derive not from ﬁrst principles, but from knowledge of particular facts. People with better facts make better judgments. No matter how skilled at critical thinking or probability analyses, a novice will often make worse predictions than an expert. A clas- sic ﬁnding in psychology is that skilled reasoners in one domain are often terrible on analogous problems with unfamiliar content. People seem to reason with Inductive Reasoning 523 informal heuristics built up from experience, or they have causal knowledge and explanatory theories that warrant conclusions. Educators often observe that learning good inferential skills in one domain does not transfer to another. One way to improve inductive reasoning is to teach content knowledge rather than (or in addition to) general decision-making strategies. An emphasis on core knowledge holds that sound judgments result from having the requisite background knowledge to identify plausible or workable solutions. Advocates of a domain-speciﬁc approach to cognitive development suggest that adult performance results not (just) from general improvements in critical thinking or intellec- tual abilities, but from greater knowledge and better theories. Children are novices in many areas. Making good decisions involves the expertise that is often characterized as common sense. A content-focused approach to induction empha- sizes abduction, or ‘‘inferenc e to the best explanation.’’ Content knowledge is require dtoidentifygoodexpla- nations. Faced with brown, wilted tomato plants, an expert gardener may draw on knowledge about common pests and properties of tomatoes to identify the problem. Although her reasoning may involve an (implicit) application of Bayes’s theorem, the appropri- ate weighting of probabilities (the priors) would be the discriminating feature of expert judgment. Abduc- tive inference involves testing hypotheses against theory (‘‘does this make sense?’’) as much as testing against data. Content-focused accounts of inductive inference face the question of how expert knowledge is acquired. One inﬂuential suggestion is that much of our inductive success is the result of innate knowl- edge. Evolution may have equipped us with the priors we need. An alternative is that content knowledge is the result of formal inference principles. Attending to patterns of covariation and probabilities in the envi- ronment is a general way to develop speciﬁc knowl- edge. Bayesian networks are one approach to the question of how domain-general principles can yield domain-speciﬁc theories. These models have been very inﬂuential in machine learning. Whether human learning can be understood along the same lines is an open research question. Charles Kalish See also Correlation; Deductive Reasoning; Scientiﬁc Method Further Readings Gilovich, T., Grifﬁn, D., & Kahneman, D. (2002). Heuristics and biases: The psychology of intuitive judgment. New York: Cambridge University Press. Holland, J. H., Holyoak, K. J., Nisbett, R. E., & Thagard, P. R. (1986). Induction: Processes of inference, learning, and discovery. Cambridge: MIT Press. Koslowski, B. (1996). Theory and evidence: The development of scientiﬁc reasoning. Cambridge: MIT Press. I NFERENTIAL S TATISTICS Two basic types of reasoning underlie all of science— deductive reasoning and inductive or inferential reason- ing. Deduction involves reasoning from a set of starting assumptions or general principles to a speciﬁc future observation. Induction requires reasoning from a lim- ited set of observations to draw conclusions about things that have not been observed or to derive general principles. That is, induction (inference) is the process by which researchers attempt to learn what the general principles or scientiﬁc laws are. Deduction will fail if one or more of the general assumptions is false. If one has the following starting assumptions: All cows give milk. Ferdinand is a cow. The deduction that ‘‘Ferdinand gives milk’’ follows logically. If one were to observe Ferdinand and ﬁnd out that he does not give milk, one of the two starting assumptions must be false. In this case, the most likely reason for the failure of the prediction is that Ferdinand is not a cow but a bull (but note that the ﬁrst assumption might also not be true). Science uses deduction to make predictions about future observations based on past observations. Sup- pose a researcher has a theory (general assumption) that verbal praise will increase the reading level of second graders. The researcher uses deduction to make a prediction about what will happen if verbal praise is used as part of the program of reading instruction for a sample of second graders. The researcher predicts that their reading scores will rise more than the scores of a control group that does not receive verbal praise. That is, the researcher uses 524 Inferential Statistics deduction to reason from a theory or set of beliefs to a hypothesis or prediction about future observations. Now the researcher wants to test the hypothesis that students who receive verbal praise will show more growth in reading scores than students who do not receive praise. Because the researcher cannot administer the treatment to all second graders, he or she must reason from a limited set of observations to a general conclusion about the effectiveness of praise on learning to read. Here, the researcher must use inferential or inductive reasoning to go from a speciﬁc set of observations (researcher’s data) to derive or support a general principle. Unfortunately, for reasons discussed shortly, if the researcher observes higher reading scores in a group of students who have received verbal praise, he or she cannot proceed directly to the conclusion that praise is effective. Inferential statistics, as a branch of applied mathe- matics, provides a way to reason inferentially with quantitative information such as reading test scores. But before discussing inferential statistics in detail, it is necessary to distinguish between two contexts in which one might want to use quantitative data. The basic issue is whether one has data on all of the enti- ties with which he or she is concerned. If, for exam- ple, the Board of Education of the state of Texas wants to know the level of mathematics achievement of eighth-grade students in the state in the current year, it would be theoretically possible to test the mathematics achievement of every eighth-grade stu- dent and compute the mean, standard deviation, and other statistics. These statistics describe the group that was tested and, disregarding the possibility of mea- surement or computational errors, the results accu- rately reﬂect the mathematics achievement of Texas eighth-grade students. The mean and other statistics computed on the group that has been tested are called descriptive statistics. If the Board has no interest in going beyond this speciﬁc group, then no inference is involved. However, it may not be practical to test every student for ﬁnancial or other reasons. In this case, the entire group about which the Board wants to make a statement is called a population and the Board could select a subgroup from this population, called a sample , for testing and compute the mean, standard deviation, and so on of the sample. The results com- puted on the sample would be descriptive statistics for the sample, but the Board’s objective is to reason or generalize from the sample to the population. There- fore, because they want to use the results from the sample to draw conclusions about the population, the Board must use inferential statistics. Whenever one uses the results from a sample to draw conclusions about a larger population, he or she is engaged in sta- tistical inference. Notice that in statistical inference, descriptive statistics from the sample are used to reach conclusions about the population. The heart of statistical inference is probability. Because the fact that using inferential statistics implies that not all of the cases of interest are mea- sured, there is always the chance that the conclusions drawn from the sample will be incorrect. Researchers would like to know and, if possible, control the proba- bility that a mistake will be made or an incorrect con- clusion will be reached. There are several ways that researchers can ask questions of the data they have gathered or plan to gather. Exactly how they use inferential statistics will be determined by how they phrase the questions. Some ways may be more useful than others for some purposes. Testing Null Hypotheses The origin of inferential statistics is bound up with what is called testing a null hypothesis. During research, researchers derive what they expect to ﬁnd using deduction from earlier work, but because they wish to generalize to unobserved cases, they must use inference. If an individual draws three marbles from a bag containing 100 marbles and all of the marbles drawn are green, it does not necessarily follow that all of the marbles in the bag are green, but if three green marbles are drawn and then a blue one, the individual knows for certain that not all of the mar- bles in the bag are green. This is the logic behind test- ing null hypotheses. A research hypothesis states what the researcher believes is true but cannot assert its truth based on a limited sample. What researchers do instead is develop a hypothesis, called the null hypothesis , which, if they can reject it as untrue, implies that the research hypothesis is true. For exam- ple, if a researcher believes that children who receive verbal praise read at a higher level than students taught another way (the research hypothesis), then the null hypothesis would be that students taught using verbal praise do not read at a higher level. If the researcher conducts a study and obtains data that are unlikely if the null hypothesis is true, then he or she concludes that the null hypothesis is false, and this Inferential Statistics 525 gives support for the research hypothesis. That is, if, based on the data, the researcher ﬁnds that students taught with praise read at a higher level, he or she has the statement ‘‘students taught using verbal praise do not not read at a higher level.’’ The data are incom- patible with the null hypothesis, producing the second not , and the double negative afﬁrms the positive. What is done statistically when stating a null hypothesis is to specify the distribution of outcomes one would expect to observe in a very large number of repetitions of an experiment if the null hypothesis is true. These outcomes will have different probabili- ties of occurring. For example, suppose a couple is planning a family of 10 children. There are 11 differ- ent numbers of boys and girls that they could have, but these 11 outcomes have different probabilities. If one assumes (the null hypothesis) that the chance of any child being a boy is .5 and the chance of a girl is also .5, the chance (probability) that all 10 children will be boys (or all girls) is about 1 in 1,000, whereas the chance that there will be 5 boys and 5 girls is about 1 in 4. Suppose that after 20 years of marriage, the couple has produced 10 boys. One of two conclu- sions can be reached: either that the result happened by chance (unlikely with a probability of 1/1,000), or that the probability of a boy at each birth is higher than that of a girl. The null hypothesis was that the probabilities of boys and girls were equal. The out- come observed was unlikely to have occurred by chance under the conditions speciﬁed in the null hypothesis , so the conclusion that the null hypothesis is not true is reached. A researcher can reject the null hypothesis if his or her observation is either larger than would be expected or smaller than would be expected. If the null hypothesis is concerned only with differences in a particular direction (for example, a researcher might be concerned only with whether verbal praise raised reading achievement, not with whether it reduced achievement), the researcher rejects the null hypothe- sis only if the difference was in the predicted direc- tion. This is called a one-tailed null hypothesis. If the researcher does not care about the direction, he or she would reject the null hypothesis when he or she observes an unlikely result in either direction, in which case the null is called a two-tailed hypothesis. Testing null hypotheses has been widely criticized andvigorouslydebatedformorethan50years.The criticisms have ranged fro m the assertion that the null hypothesis is ‘‘quasi-always’’ false, and therefore rejecting it gives researchers no information (see Meehl, 1978) to the observation that the question asked by null hypothesis testing is not the question researchers want to ask (Cohen, 1994; Schmidt, 1996). Because null hypoth- esis testing has had such a central place in statistical inference and continues to be the most widely used infer- ential procedure, this entry considers the processes involved in some detail. Later, this entry looks at some of the alternatives that have been proposed. Sampling Distributions and Test Statistics In a research study such as the one concerning the effect of verbal praise on learning to read, if the researcher expects verbal praise to be effective, the null hypothe- sis would be that praise will have no effect or a nega- tive effect. This means that if the null hypothesis is correct, the researcher would expect, in a large num- ber of repetitions of a study of praise, that the mean reading score of praised students would be the same as (or less than) the mean of students not receiving praise. In any single study, the means might differ by a small amount. The question that statistical infer- ence procedures attempt to answer in the case of null hypothesis testing is whether the difference the researcher observes is too large to have occurred by chance if the null hypothesis is true. If the probability of getting a result such as the one obtained is less than 5 % (or 1 % , the conventional critical probabilities), the researcher concludes that the result is statistically signiﬁcant, that he or she can reject the null hypothe- sis, and that the results support the research hypothe- sis. Note that what the phrase statistically signiﬁcant really means is ‘‘unlikely to have occurred by chance under the conditions speciﬁed in the null hypothesis.’’ Statistical inference is based on the concept of the sampling distribution. Let us conduct a thought experi- ment. Suppose that the Texas Board of Education wishes to know what the level of mathematics knowl- edge is among eighth graders. The mean for all stu- dents in the state is a population parameter, a number that characterizes the entire population. It is conven- tional to use Greek symbols to represent parameters, so the symbol µ (mu, the Greek lower-case letter for M ) is usually given to the population mean. If the Board measures the mathematics knowledge of a sample of N (for example, 25) students from this population, the result is a descriptive statistic for the sample, but it can be used to draw an inference about the nature of the 526 Inferential Statistics population. Roman letters are usually used for sample statistics, so M is the symbol often used for the mean. In this thought experiment, the Board draws not one sample but a very large number of samples (all of the same size) and computes the mean (and other descriptive statistics) for each sample. They now have as data not a distribution scores for individuals but a distribution of means for samples, each mean based on the scores of N students. This distribution of means is called the sampling distribution of the mean,and it is used as the basis for sta tistical inference about the population mean. The sampling distribution of the mean contains means ( M s) from samples of size N : The other descriptive statis tics that might be computed (such as the standard deviation) also have sampling dis- tributions, and inferences about these statistics would be based on those sampling di stributions. Because most of the practice of statistical inference is focused on means, the mean will be used for the development of inferential procedures, but readers should be aware that exactly the same logical processes can be applied to standard deviations, correlation coefﬁcients, and other descriptive statistics. The only difference is in the test statistic that is used. In ever y case, the null hypothesis speciﬁes a value for the parameter of interest and deﬁnes the sampling distribution of the statistic, and sta- tistical inference allows researchers to state how likely the particular sample statistical value is if the null hypothesis is true. Two additional pieces of information are needed to complete this inferential picture. First, we need to know the mean and standard deviation of the sam- pling distribution under the null hypothesis. Next, we need a way to determine the probability of any partic- ular outcome given the sampling distribution. Statistical theory has shown that the mean of the means from a large number of independent random samples from a population is equal to the mean of the population. That is, the mean of the sampling distribu- tion, the mean of M s, is equal to µ . It can also be shown that if the population standard deviation (called σ , Greek lower-case letter sigma for S ) is known, then the standard deviation of the sampling distribution of M is σ M = σ ﬃﬃﬃﬃ N p : When σ is not known, σ M must be estimated from the data. One way to get a good estimate of σ M is with the formula ^ σ M = S ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ N − 1 p , where S is the standard deviation computed from the sample and ^ σ indicates an estimate rather than an exact value. Now the mean and standard deviation of the sam- pling distribution are known if the null hypothesis is true. All that remains is to develop a way to deter- mine the probability of any sample result from this sampling distribution. This involves the use of what is called a test statistic. For the current example, one test statistic will be needed if σ M is known, and a different one if ^ σ M must be used. The test statistic when σ M is known is called Z : The value of Z reﬂects how far a given observation is from the mean of the distribution in units of the standard deviation of that distribution. It can be computed as Z = M − µ σ M : Z will be negative if the observed mean is below µ and positive if it is above µ. Statistical theory tells researchers that Z follows the normal distribution, also oftencalledthe bell-shaped distribution or bell curve. Most statistics books contain a table that includes the probability of obtaining a value of Z of a particular size. For example, the probability of obtaining a Z of +1.96 or greater is 2.5 %. Therefore, if a researcher conducts a study using a one-tailed h ypothesis and obtains a Z for the sample mean of +1.96, the researcher can be conﬁdent that if he or she rejects the null hypothesis he or she will make an error no more than about 3 % of the time because a result this different from the population mean would occur by chance only about two to three times in 100 repetitions. (For a two-tailed hypothesis, the probability from each end of the distribution is included, so the probability would be 5 %.) For exam- ple, if the null hypothesis says that the mean reading score in the population is 50 or less (assume it is known that the standard deviation in the population is 10) and a mean of 55 in a sample of 25 students is obtained, the value of σ M is 10  ﬃﬃﬃﬃﬃ 25 p = 2, so the Z statistic for these data is Z = 55 − 50 2 =+ 2 : 5 : Consulting the probabilities for the normal distri- bution, the researcher would ﬁnd a value of .006. This Inferential Statistics 527 means that, on the basis of the results, the researcher can reject the null hypothesis that µ ≤ 50 with a prob- ability that he or she is making a mistake of less than .01. A sample mean of 55 is so unlikely to occur in a sample of 25 from a population with a mean of 50 or less that the researcher concludes that the mean is not 50 but some other (unknown) value. Note that rejecting the null hypothesis does not tell what the population parameter is, only what it is not ( µ is not less than or equal to 50). This feature is the source of one of the main criticisms of null hypothesis testing. When σ is not known, the logical process is identi- cal, but the details are a little more complicated. In this case, the test statistic is known as t,butadifferent probability distribution must be used depending on sample size. That is, for the Z test statistic, there was only one table of probability values, but for t , there is a different table for each possible sample size. The value for t is computed using the formula t df = M − µ ^ σ M : The df subscript tells the researcher he or she needs to look in a table of probability based on the sample size, df equals N − 1, and there is a different probability table for each value of df (most statistics books include a single table of critical values of the t statistic—values with a probability of .10, .05, .025, and .01 for dif- ferent values of df —because these are the most com- monlyusedvaluestomakeadecisionaboutthenull hypothesis). Using the above example, if the researcher did not know the population standard deviation was 10 but had computed that value from the data, ^ σ M would be 10  ﬃﬃﬃﬃﬃ 24 p = 2 :04 and t 24 = 55 − 50 2 : 04 = 2 : 45 : Using the appropriate table, the researcher would ﬁnd that the probability of a t this large is just over .01. That is, the researcher would once again reject the null hypothesis, but the risk that he or she is mak- ing an error is slightly larger. When the researcher does not know σ , the distribution of the test statistic is slightly wider, so a particular difference from the null value is slightly more probable. The larger the sample size, the smaller the difference between Z and t : Most research situations involve comparing the means (or other statistics) from two or more samples. In such cases, the null hypothesis almost always is that the samples come from a common population or from populations that have equal parameters. That is, the null hypothesis is of the general form µ A = µ B = µ C … The details of how hypotheses like this are tested can be found in the texts listed in the Further Read- ings, but the underlying logic remains exactly as out- lined. An appropriate test statistic is computed and the researcher determines how likely it is that he or she would get a value that large if the null hypothesis is true. If the value of the test statistic is too large, the researcher concludes that the null hypothesis is false. Alternatives to Testing the Null Hypothesis Confidence Intervals Several alternatives to null hypothesis testing have been proposed as ways to draw valid inferences from research data. The most widely used is called the con- ﬁdence interval approach. The essence of conﬁdence intervals is that they approach the issue of inference in a positive way, seeking what is true, rather than in a negative way, rejecting what is shown to be false. Continuing with the thought experiment in which a large number of samples from a population were drawn and the mean for each sample was computed, when a null hypothesis with a 5 % chance of making a mistake in rejecting it was tested, a Z score or a t score for the sample result in the distribution speciﬁed by the null hypothesis was computed. Alternatively, critical values ( CV s) for the mean in the null distribu- tion could have been computed by using the value of Z or t that includes 95 % of the normal or t distribu- tion. The formula for the normal distribution is CV NH = ð + Z : 5 α Þð σ M Þ + µ NH , where CV NH is a critical value for rejecting the null hypothesis, + Z : 5 α is the lower ( − ) or upper ( + ) Z for an interval that has the speciﬁed probability α in the normal distri- bution (note that half of alpha is in each end of the distribution), 528 Inferential Statistics σ M is the standard error of the mean, µ NH is the mean of the null hypothesis distribution. An identical formula with appropriate changes for the t distribution can be used when the population standard deviation is not known, and comparable for- mulas exist for pairs of means and for other statistics such as the correlation coefﬁcient. The result is an interval that includes the speciﬁed proportion of the null distribution. If the sample result, M,fallswithin that interval (called the interval of nonrejection ), researchers would not reject the null hypothesis, but if M falls outside that interval (called the rejection region ), researchers would reject the null. Now suppose that instead of placing the probability interval around the null hypothesis value, researchers place it around the sample result, M : The sample has been drawn from a population that has a mean µ Pop : Means of samples drawn from this population will form a sampling distribution that is a normal distribu- tion with a mean equal to µ Pop and a standard devia- tion of σ M : When researchers test a null hypothesis, they are asking whether it is reasonable to believe that µ Pop = µ NH : If M falls in the interval of nonrejection, then the probability interval that researchers compute around M will include µ NH : However, if the sample mean falls outside the interval of nonrejection, then the interval around M will not include µ NH : The formula is CV CI = ð + Z : 5 α Þð σ M Þ + M , and the resulting interval is called a conﬁdence inter- val. Comparable formulas are available when σ M is not known and for statistics other than the mean. The essence of conﬁdence interval thinking is that researchers now have an interval that has a speciﬁed probability of including the unknown parameter µ Pop : If a researcher draws a very large number of samples from the population and com putes a conﬁdence inter- val around each sample statistic (e.g., M), 95 % (or whatever probability the researcher has speciﬁed by his or her choice of values for Z) of those conﬁdence inter- vals will include the population parameter. Thus, whereas with null hypothesis testing, researchers rule out the region of nonrejection for µ Pop when they reject the null, but leave the rest of the possible values as a region of uncertainty, w ith conﬁdence intervals, the researchers rule out all of the possible values except those within the conﬁdence interval. In this sense, conﬁdence intervals give re searchers information about where the parameter is, rather than where it is not. Another advantage claimed for conﬁdence inter- vals is that they subsume the null hypothesis test. That is, if the conﬁdence interval includes the null value, a conventional null hypothesis test would have led to nonrejection. If the conﬁdence interval does not include the null value, the null hypothesis would have been rejected. A ﬁnal advantage of conﬁdence intervals is that as sample size increases, the conﬁdence interval gets nar- rower, yielding a more precise estimate of the popula- tion parameter. On the other hand, as sample size increases, the area of nonrejection of the null hypothe- sis gets smaller. This makes it easier to reject the null hypothesis, but if researchers do reject the null, the area of uncertainty is larger, prod ucing the paradoxical situ- ation that a larger study yields less information. Bayesian Inference A third way in which probability can be used to help researchers decide about the truth of hypotheses is known as Bayesian statistics. As Cohen pointed out, the Bayesian approach asks about the probability that a particular hypothesis is correct, given the data observed, whereas traditional hypothesis testing asks what is the probability of the data, given the null hypothesis. Cohen argues that the former question is the question of interest. Bayesian analysis requires that researchers specify a probability distribution under the research hypothesis, known as the prior probability distribution. They then collect some data and use the observed data distribution to calculate a revised probability distribution known as the posterior probability distribution. That is, they use the data to modify their belief about the true state of affairs. As Cohen also points o ut, this allows researchers to use successive studies to reﬁne both their hypotheses and their parameter estimates. Unfortunately, Bayesian analysis is computationally more difﬁcult and requires a better grasp of mathematics than do null hypothesis testing or conﬁdence intervals, so these methods are not as widely used in the behavioral sciences. Model Fitting Another attractive alternative to null hypothesis testing is the ﬁtting of speciﬁc models to the observed data. The question being asked is how well the model Inferential Statistics 529 ﬁts the data, and alternative models can be compared by contrasting the degree of lack of ﬁt. The entire area of structural equation modeling uses this approach. The mean speciﬁed in a null hypothesis ( µ NH )postu- lates a speciﬁc model for each observation in a set of data. The lack of ﬁt of the null mean for any given obser- vation is given by the distance of the data point from µ NH : The sum of squares for these lack-of-ﬁt values ( SS NH ) is the lack of ﬁt for the null hypothesis model: SS NH = X ( X − µ NH ) 2 : Likewise, the lack of ﬁt for the sample mean is given by its sum-of-squared deviations: SS Data = X ( X − M ) 2 : From the principle of least squares, SS Data must be less than SS NH unless the two means are equal. The model-ﬁtting approach asks whether the reduction in lack of ﬁt is larger than researchers would expect to have occurred by chance. The approach generalizes to thecaseoftwoormoremeans,aregressionline,or any other statistical model that one might want to ﬁt. In each case, the question being asked is whether the more complex model (for example, using two group means rather than one combined mean) provides a reduction in lack of ﬁt that is greater than a chance amount. Robert M. Thorndike See also Descriptive Statistics; Statistical Signiﬁcance; T Scores Further Readings Cohen, J. (1994). The earth is round ( p < : 05). American Psychologist , 49 , 997–1003. Kaplan, D. (2000). Structural equation modeling: Foundations and extensions. Thousand Oaks, CA: Sage. Meehl, P. (1978). Theoretical risks and tabular asterisks: Sir Karl, Sir Ronald, and the slow progress of soft psychology. Journal of Consulting and Clinical Psychology , 46 (4), 806–834. Meehl, P. E. (1997). The problem is epistemology, not statistics: Replace signiﬁcance tests by conﬁdence intervals and quantify accuracy of risky numerical predictions. In L. Harlow, S. A. Mulaik, & J. H. Steiger (Eds.), What if there were no signiﬁcance tests? (pp. 393–425). Mahwah, NJ: Lawrence Erlbaum. Salkind, N. J. (2004). Statistics for people who (think they) hate statistics (2nd ed.). Thousand Oaks, CA: Sage. Schmidt, F. L. (1996). Statistical signiﬁcance testing and cumulative knowledge in psychology: Implications for the training of researchers. Psychological Methods , 1 , 115–129. Sirkin, R. M. (2006). Statistics for the social sciences (3rd ed.). Thousand Oaks, CA: Sage. Skorupski, W. P. (2007). Bayesian statistics. In N. J. Salkind (Ed.), Encyclopedia of measurement and statistics. Thousand Oaks, CA: Sage. Thompson, B. (2006). Foundations of behavioral statistics. New York: Guilford. I NSTITUTIONAL R EVIEW B OARDS Institutional review boards (IRBs) are local universi- ties given the responsibility of oversight for federally funded research involving human subjects conducted by members of the organization. Researchers from the local organization serve on the IRBs and often employ one or more staff members, including a com- pliance ofﬁcer, to help facilitate the review process. IRBs’ responsibilities include providing training so that researchers conduct research safely, approving research protocols that are designed to protect partici- pants from harm, and making sure that potential human subjects are adequately informed of the risks and beneﬁts of their participation so that they can give informed consent. IRBs must balance the risk of the research against its potential beneﬁts in approving research protocols. IRBs have become an entrenched partoftheresearchprocessintheUnitedStates. There is near-unanimous support for the overall goal of IRBs protecting human subjects from unnecessary harm. Because the policies based on medical research are also applied to social science and humanities research, a number of controversies and problems resulting from unclear or changing deﬁnitions or from mission creep and excessive regulation of research have emerged as IRBs have increased in number, size, and scope. IRB critics believe that these concerns have a chilling effect on research; reduce and slow research productivity unnecessarily; and seem to give IRBs powerful control over researchers who feel they have little recourse. Supporters of IRBs see increasing reg- ulation and enforcement as positive steps in further protecting human subjects from harm by overzealous researchers. Nevertheless, it is clear that scholars will continue to have to interact with IRBs as part of the research process. 530 Institutional Review Boards A Brief History IRBs grew out of concerns over serious human subject research abuses such as the experiments conducted by Nazi scientists during Wo rld War II. The Nuremberg Code issued by the War Crimes Tribunal in response to these atrocities is generally recognized as the ﬁrst international code of medical research ethics. However, research abuses continued, such as in Tuskegee, Alabama, in the late 1940s when poor, Black men were denied available medical treatment for syphilis so that researchers could study the effects of the disease on them. Although most of the commonly cited research abuses involve medical research, behavioral or social science researchers also committed abuses. Perhaps the most commonly cited example is a set of psychological studies conducted by Stanley Milgram in the early 1960s. These studies would b e considered unethical by today’s research standards because the participants had been deceived into believing that they were giving excessive electric shock s to other people and were inadequately debriefed about the experiment. In 1974, the National Commission for the Protec- tion of Human Subjects of Biomedical and Behavioral Research was created and the National Research Act was passed by Congress. This law required the estab- lishment of local IRBs to review and approve all human subject research that is federally funded. In 1979, the commission approved a guide for research with human subjects: The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research . Enforcement of the human research guidelines has fallen on a number of different federal agencies over the years. Currently, the Ofﬁce of Human Research Protection (OHRP) in the U.S. Department of Health and Human Services oversees an estimated 3,000– 5,000 IRBs across the nation that regulate medical, social, and behavioral science research. Defining Human Subject Research Because IRBs have responsibility over research involving human subjects, deﬁning human subject research becomes important. In general, the researcher must directly interact with the human subjects for research to need IRB approval. Therefore, publicly available information, such as political speeches, pub- lished or broadcast information, or observations of public behavior in which individuals are not identiﬁed (e.g., percentage of people using cell phones while walking at the mall), is generally not considered the purview of IRBs. However, IRBs may differ on how they classify visiting Internet chat rooms or other pub- licly available online forums. The guidelines deﬁne research somewhat ambigu- ously as contributing to the body of knowledge. This has generally been interpreted to mean that class assignments do not need IRB approval because they contribute to individual learning and not to the gen- eral body of knowledge. Research intended for pre- sentations at conferences or distribution in academic journals or books generally must be reviewed by IRBs. The ambiguous deﬁnition of research has cre- ated problems, particularly in journalism departments and departments involved in historical documentation such as oral histories. Most IRBs have concluded that journalistic activities do not need IRB approval, but practice may vary for oral histories. Informed Consent A critical part of the approv al process for research pro- tocols concerns informed consent. The IRB guidelines indicate that potential par ticipants in a study must be informed of the purpose of the study, the procedures, potential risks and beneﬁts of participation, and the consequences (if any) of withdrawing from the study prior to its completion. Potential participants must be allowed to ask questions about their participation before agreeing to particip ate and be made aware that their participation is voluntary and that they do not have to answer any questions if they do not want to and can withdraw from the study at any time. Informed consent can be accomplished through an oral or written presentation of the information. Oral consent is recommended when a written consent form provides the only link between the human subject and the study. Two common problems with informed consent involve complexity and length. It is recommended that consent be worded using eighth-grade reading levels in order to reasonably ensure comprehension. In an effort to be comprehensive, some consent forms become so long that human subjects fail to read them completely, in effect nullifying the beneﬁt of being informed. Finally, minors (under 18) cannot legally give informed consent under most circumstances; they can only assent to participate. A parent or guardian must give informed consent for the minor to participate, Institutional Review Boards 531 and the minor then assents or agrees to participate. In the past, researchers could notify parents in a letter explaining that their children would be involved in a study and that they only needed to reply if they did not want their children to participate. Current policy does not allow for this form of ‘‘passive consent.’’ Now, active consent must occur in which signed doc- umentation from the parent must be received before a minor can assent to participate. This need for active consent has made it more difﬁcult to conduct research on minors. Levels of Review In reviewing and approving research protocols, IRBs classify research into one of three levels of review. In general, studies in educational settings or using gener- ally accepted research methods such as interviews and questionnaires are classiﬁed as exempt and can be reviewed and approved by an IRB staff member. Exempt protocols must still make sure that human subjects receive adequate information concerning risks and beneﬁts in order to give informed consent, record responses in such a way that the names of par- ticipants are not identiﬁable, and expect that disclo- sure of the participants’ responses would not put them at risk for civil or criminal liability. Studies classiﬁed as expedited must still involve only minimal risk to human subjects, but may involve more sensitive topics or unusual research methods. Expedited studies are reviewed and approved by one or more members of the board rather than staff members. Studies that involve more than minimal risk to human subjects or involve human subjects in a protected group, such as minors or incarcerated individuals, must receive a full board review. The majority of the board members present must approve the study. Issues and Concerns Although nearly all researchers agree with the general mission of IRBs of protecting human subjects, a num- ber of issues or concerns have risen around IRBs. Each potentially hinders legitimate research. Administrative Problems A collection of narratives from active researchers published in a special issue of the Journal of Applied Communication Research in August 2005 edited by Michael W. Kramer and Debbie S. Dougherty pro- vides a sample of the administrative problems per- ceived by researchers. Among the most prominent problems are slow responses or delays from IRBs and changing or incorrect guidelines being applied to protocols. Often, these are the result of applying poli- cies and practices appropriate for high-risk medical research and drug testing to low-risk research in the social and behavioral sciences. This often results in multiple submissions to IRBs and further delays. These processing delays can frustrate faculty mem- bers, particularly those under the pressure of gaining promotion and tenure in a limited time frame. Com- pliance ofﬁcers and board member cite changing rul- ings from OHRP, insufﬁcient staff, and poorly written protocols as reasons for most delays. ‘‘Mission Creep’’ or Excessive Regulatory Control A number of concerns with IRBs were cogently summarized and addressed in The Illinois White Paper , Improving the System for Protecting Human Subjects: Counteracting IRB ‘‘Mission Creep,’’ pub- lished by The Center for Advanced Studies at the University of Illinois in Champaign. Mission creep occurs when IRBs begin to oversee activity that previ- ously was not considered human subject research. Efforts by some IRBs to include journalistic activities would be a clear example of this. Mission creep also occurs through excessive regulation when research that should be classiﬁed as exempt becomes classiﬁed as expedited or full board, or when unnecessary pre- cautions, those typically needed for high-risk medical research, are imposed on low-risk social science research. Those disturbed by mission creep point to prob- lems with IRBs interpreting the guidelines through changing or expanding deﬁnitions. For example, some IRBs consider typical research incentives (e.g., pay or extra credit for a class) as undue coercion. IRBs sometimes broaden the deﬁnition of risk to include experiences that are common to everyday life, such as recalling a mildly unpleasant event from the past. Excessive regulation occurs when all students are considered to belong to a protected group or when psychological counseling must be made available for research involving minimal risk. Supporters of the expanding purview of IRBs indicate that their thor- oughness and care prevent potential problems. 532 Institutional Review Boards Loss of Purpose As IRBs have become more established, they have grown in size and budgets. A number of critics, such as medical researchers George Annas and Elliot Fou- car, have suggested that IRBs no longer focus atten- tion on protecting human subjects, and instead focus on protecting universities from lawsuits by disgruntled research subjects. Others suggest that the OHRP is more concerned with maintaining bureaucratic rules and regulations than protecting human subjects. As evidence of this, critics point out that in the vast majority of cases in which universities have had their research programs temporarily shut down for viola- tions, there is no evidence that human subjects were harmed in any manner. In most instances, the shut- down is caused by missing or poor standard operating procedures, poor minute keeping, or lack of a quorum at the approval meeting. Supporters point to the lack of harm to human subjects as evidence of the success of the procedures. Michael W. Kramer See also Ethics and Research; Ethnography; Experimental Design; Field Experiments; Naturalistic Observation; Qualitative Research Methods; Quantitative Research Methods Further Readings Annas, G. J. (2001). Reforming informed consent to genetic research. Journal of the American Medical Association , 286 , 2326–2328. Center for Advanced Studies. (2005, November). The Illinois white paper: Improving the system for protecting human subjects: Counteracting IRB ‘‘mission creep.’’ Champaign: University of Illinois. Dougherty, D. S., & Kramer, M. W. (2005). Organizational power and the institutional review board. Journal of Applied Communication Research , 33 , 277–284. Dougherty, D. S., & Kramer, M. W. (2005). A rationale for scholarly examination of institutional review boards: A case study. Journal of Applied Communication Research , 33 , 183–188. Foucar, E. (2002). How much oversight is necessary to protect human subjects? Journal of the American Medical Association , 287 , 716–717. Hamilton, A. (2005). The development and operation of IRBs: Medical regulations and social science. Journal of Applied Communication Research , 33 , 189–293. Milgram, S. (1974). Obedience to authority: An experimental view. New York: HarperPerennial. Web Sites Ofﬁce of Human Research Protections: http://www.hhs.gov/ohrp I NSTRUCTIONAL O BJECTIVES Instructional objectives are written by teachers or instructional designers and express the intended out- come of instruction. They specify what the learners will be able to do as a result of instruction by stating the performance standards that learners are expected to achieve. Instructional objectives also go by other labels, such as behavioral objectives, performance objectives , or simply objectives. Two examples of instructional objectives are as follows: (1) given two single-digit numbers, the students will be able to add the numbers without the aid of a calculator, and (2) given a diagram of an amoeba, the seventh-grade science students will be able to label the protoplasm, nucleus, cytoplasm, and the pseudopodia. Instructional objectives do n ot state instructional pro- cedures. Instructional proce dures refer to what teachers do during instruction, including the media used and instructional activities. Instructional objectives, by con- trast, are student-focused; these objectives center on what students should be able to do at the end of instruc- tion. Thus, although instructional procedures are deter- mined in light of objectives, the objectives state what the students should do rather than what the teacher should do. Instructional objectives are also different from instructional goals. A goal is a broad, general statement regarding the intended beneﬁts of instruction. Typically, goals are long term, spanning entire units of instruction. Objectives, by contrast, are speciﬁc and proximal. Thus, the instructional objectives provide the level of speciﬁcity needed to guide daily activities and allow for monitoring of goal progress. A single instruc- tional goal may be achieved after students have met a series of progressive in structional objectives. How instructional objectives are conceptualized and written has been the focus of much work in edu- cational psychology. The work of several researchers, such as Robert Gagne, Benjamin Bloom, and Robert Mager, has converged on a set of recommendations for writing objectives, and both behavioral and cogni- tive learning theories have established principles for the use of objectives in instruction. Instructional Objectives 533 Characteristics of Written Objectives Instructional objectives are a critical piece of effective instruction because they clearly identify and communi- cate the intended outcomes for instructional events. When a teacher communicate s the instructional objec- tives to students, he or she serves to clarify expectations and allows students to attend to the important aspects of instruction. Instructiona l objectives also provide the standards that students can use to evaluate their prog- ress toward desired outcomes. For the teacher, instruc- tional objectives inform the in structional activities that occur and the assessments used to evaluate student progress. Accordingly, instru ctional objectives are writ- ten prior to instruction during the planning stage, are used to determine the instructional procedures used dur- ing instruction, and provide the guidelines for practice- and assessment-related activities. During the planning phase, teachers write the instructional objectives that will determine the course of instructional events. Good instructional objectives are characterized as speciﬁ c, measurable, observable, and short term. Speciﬁc objectives are unambiguous and clearly state what the learner should be able to do. In writing objectives, teachers should avoid the use of words such as know, understand,and analyze because these words do not convey a speciﬁc, unambiguous action. Instead, objectives s hould be written wit

bundet

h verbs that clearly state the observable action that students will be able to carry out at the end of instruction. For example, rather than stating that a student should ‘‘know’’ certain material, a good instructional objective may state that the student should be able to deﬁne, list, label, or match speciﬁc concepts from the material. Rather than stating that a student should be able to ‘‘analyze’’ the material, a good instructional objective may state that the student should be able to diagram, differentiate, question, or summarize the material. In the study of instructional objectives, several lists of verbs have been developed that teachers can use when writing objectives. The most widely known of these is the list of verbs that accompanies Bloom’s Tax- onomy of Instructional Objectives. Bloom and col- leagues wrote the taxonomy in the late 1950s as a means to organize objective s according to their cogni- tive complexity. The six levels of this taxonomy, ordered from the least complex to the most complex, are to know, comprehend, apply, analyze, synthesize, and evaluate. The lowest level of this taxonomy, to know, refers simply to the memorization and recitation of facts. Evaluation, however, not only requires factual knowledge but also the ability to make judgments about that knowledge. Each level can be translated into instructional objectives by using a set of verbs that cor- responds to each level. The verbs describing knowl- edge and analysis that were listed earlier in this entry are two examples of these verb sets. Other examples include the verbs compose, design,and expand to iden- tify clearly what is meant by synthesis. Compare, con- trast ,and judge are three verbs that can turn evaluate into a clear and unambiguous objective. Complete lists of these verb sets can be found in most references on instructional objectives. In addition to the use of speciﬁc verbs, teachers must give consideration to several other factors when writing instructional objectives. For example, instruc- tional objectives must be close in time and reference outcomes that can be both observed and measured. This means that objectives must describe a behavior that the teacher can either see or hear. A teacher can see, for example, students’ written responses or answers to mathematics problems. The teacher can also see the results of students’ science experiments or created representations. Musical performances, such as playing a scale or singing a particular note, provide good opportunities for objectives that are heard. Each of these shares the characteristic of speci- fying a behavior that the student must exhibit in some way to the external world. The ABCDs of Written Objectives Suitably written instructional objectives contain four elements that can be expressed as the ABCDs of objectives. In this system, A stands for audience,B for behavior , C for conditions ,andDfor degree.The audience is the intended group of learners to which the objective applies. The audience may be seventh- grade science students, ﬁrst-grade readers, or fourth- grade learning disabled students. The beneﬁt of attending to the audience when writing instructional objectives is that, by stating the learners to whom the objective applies, it structures the objective so that it is focused on the learner. Behavior refers to what the learners are to do and is indicated by the verb of the objective. Conditions indicate the exact circumstances under which the behavior is to occur. Conditions clarify the resources that learners will have available at the time of the performance and the circumstances under 534 Instructional Objectives which the behavior must be performed. It is the condi- tion element of the objective that speciﬁes, for exam- ple, if the behavior must be exhibited independently or with the aid of a partner, if students will be permit- ted to use notes or books, or if mathematics problems can be solved with the aid of a calculator. The ﬁnal element, the degree, identiﬁes the standards for an acceptable performance. It is the degree element of the objective that clariﬁes the quality or quantity of the behavior necessary to achieve a level that is deemed to be sufﬁcient. Degrees might be expressed as the percentage of problems to be answered correctly, the number of errors that can be made, or the amount of time that a behavior should require. The degree element is most difﬁcult to express when the objectives corre- spond to subjective performances such as writing a good summary or creating a novel design. In these cases, it is important that the objectives are tied to clearly stated scoring criteria. A good summary, for example, may contain the main idea, three supporting details, and a synthesis statement. Sharing these criteria with learn- ers may clarify expectations and further specify the stated objective. Categories of Written Objectives To write instructional objectives, teachers must also consider the type of behavior that is the target of the objective. Objectives can be categorized as psycho- motor, cognitive, or affective. Psychomotor objectives refer to behaviors that the learner is to perform. Play- ing a musical instrument or executing an athletic move is an example of a psychomotor objective. In academic domains, psychomotor objectives can refer to the execution of activities such as properly arran- ging manipulatives, operating equipment, or manag- ing behavior. Cognitive objectives refer to what the students will learn or the intellectual capacity they will acquire through instruction. There are various types of cognitive objectives, such as comprehension or problem solving. Most resources on instructional objectives recommend using the levels of Bloom’s taxonomy to conceptualize cognitive objectives. Cor- responding verbs can, of course, be used when writing these objectives. Finally, instructional objectives can refer to affec- tive outcomes. Affective objectives include students’ attitudes, values, and expectancies. An affective objective may state, for example, that students in a mathematics class will have conﬁdence in their ability to complete mathematics problems or that the students will understand the value of a particular mathematical procedure to their everyday lives. In general, affective objectives are the most difﬁcult to both write and assess because it is difﬁcult to deter- mine the objective and observable behaviors that cor- respond to the desired outcome. The Role of Objectives in Instruction and Assessment Once written, objectives prov ide guidance for the design of instructional materials an d activities. Although there is not a one-to-one correspondence between a written objective and the instructional methods that can be used to achieve the objective, a well-written objec- tive does guide the teacher’s decision-making pro- cess. When selecting instructional materials, for example, the teacher must choose materials that con- tain the necessary information and that present the information in a manner consistent with the objec- tives. In class, instructional activities should also provide opportunities for practice and feedback on the desired behaviors. Scaffolds to support student progress toward the intended outcomes may also be included in the materials. Furthermore, because instru ctional objectives require an observable behavior, they are also used to select the assessments that are used to determine if the objective has been achieved. The in-class activities, take-home assignments, and class tests are all consis- tent with the stated objective so that the perfor- mances for which students are held accountable are the same as those communicated to them through the objective. Student performance on these assessments provides the opportunity for teachers and students alike to assess progress toward the intended out- comes. When ﬁnal assessments, such as a unit exam, are given, the assessment should align with the objec- tives that preceded it. The conditions of the exam, for example, should be consistent with the conditions of earlier objectives. In short, students should be assessed only on behaviors and abilities that were part of earlier objectives. Theories of Learning and Instruction Several theories of learning and instruction discuss the role of instructional objectives in student learning. Instructional Objectives 535 Contingency contracts are recommended by behav- ioral learning theories as a method for the teacher and student to discuss and establish the goals and con- ditions for learning-related outcomes. Contingency contracts are used with individual students. In these contracts, the instructional objective serves as the statement of the desired terminal behavior. Conditions for the behavior and consequences for the outcome are also included. Gagne recommends that objectives be incorporated into the instructional design process. In this theory, objectives are classiﬁed according to one of ﬁve cate- gories of learning outcomes (e.g., verbal information, cognitive strategies). Each category corresponds to a set of critical learning conditions, and the conditions specify the environmental conditions the learner needs in order to achieve the goal. By considering the condi- tions of learning alongside the categories of objec- tives, the instructional designer can translate the instructional objective into the instructional design. Peggy N. Van Meter See also Bloom’s Taxonomy of Educational Objectives; Contingency Contracts; Goals; Grading Further Readings Bloom, B. S., Engelhart, M. D., Furst, E. J., Hill, W. H., & Krathwohl, D. R. (1956). Taxonomy of educational objectives: Handbook I, cognitive domain. New York: McKay. Gagne, R. M., Briggs, L. J., & Wager, W. (1992). Principles of instructional design (4th ed.). Fort Worth, TX: Harcourt Brace. Mager, R. F. (1984). Preparing instructional objectives (2nd ed.). Belmont, CA: Lake. I NTELLIGENCE AND I NTELLECTUAL D EVELOPMENT Modern Conceptions of Intelligence The modern history of theory and research on intelli- gence has its twists and turns largely due to the unwieldy nature of the concept. On one hand, intelli- gence is a concept that has high currency, a valued human resource that people try to cultivate or harness for advancing their causes and agendas, individually or collectively. In the meantime, it is also an abstract, elusive concept with many faces. When a task force put together by the American Psychological Associa- tion (APA) reported ‘‘knowns and unknowns’’ about intelligence, it was not able to come up with a uniformly agreed-upon deﬁnition of what intelligence is, other than the following statements by Ulrich Neisser: Individuals differ from each other in their ability to understand complex ideas, to adapt effectively to the environment, to learn from experience, to engage in various forms of reasoning, to overcome obstacles by taking thought. Although these indi- vidual differences can be substantial, they are never entirely consistent: A given person’s intellectual performance will vary on different occasions, in different domains, as judged by different criteria. Concepts of ‘‘intelligence’’ are attempts to clarify and organize this complex set of phenomena. (Neisser et al., 1996, p. 77) Three observations can be made about the state- ments. First, there is some consensus in the research community on what are typically seen as constituents of intelligence or intelligent behaviors; however, there is no clear answer as to whether they are closely connected facets or just concepts loosely coupled together. Second, the statements treat intelligence squarely as a differential or individual difference con- cept, yet give much leeway for intraindividual and contextual variability. Finally, the statements high- light the term intelligence as a psychological con- struct, a conceptual tool conjured up by scientists to sort and organize observations at an abstract level, not a physical reality like height or weight. Although Francis Galton started the tradition of research on psychometric intelligence, Alfred Binet and Charles Spearman were the two most prominent early pioneers of modern theory and measurement of intelligence. Binet, in collaboration with his doctoral student Theodore Simon, developed the ﬁrst modern intelligence test for the purposes of identifying and helping children with severe learning difﬁculties in school. Although the purposes were to develop a more reliable and objective assessment than informal clini- cal observations could offer, Binet saw the instrument as a clinical tool for diagnostic and instructional pur- poses. He also believed that intelligent performance and behavior involves a set of processes that can be 536 Intelligence and Intellectual Development identiﬁed in children’s performance and targeted for intervention. In comparison with Binet, Spearman was a different kind of researcher. He ﬁrmly believed that human intelligence can be clearly deﬁned and measured with accuracy. He used simple measures of sensory discrimination as an indicator of intelligence, believing that sensitivity to subtle differences and relations best characterizes intelligent persons. In con- trast to Binet’s inclinations as a clinician, Spearman was a mathematician, who preferred numbers to immediate observations. By parsing performance indi- ces into a shared variance and residue variance, he formulated a two-factor theory: individual differences in intelligence can be represented as consisting of a general factor ( g ) and speciﬁc factor ( s ). When Spearman made a bold claim with a title of ‘‘‘General Intelligence’ Objectively Determined and Measured’’ for his famous 1904 article, Binet was not convinced; he doubted whether a phenomenon as complex as intelligence can be reduced to a single number or a set of numbers. Indeed, he made the counterargu- ment that two individuals who obtain the same score might well use quite different skill sets. Binet was more intrigued by subtle individual differences observed during performance than the apparent sim- plicity of the mathematical solution offered by Spear- man. This tension, revealed in exchanges between Spearman and Binet and alluded to in the quoted statements above by the APA task force, has lingered to date. Intelligence: Structural or Functional? Conceptualizing intelligence as a structural feature of mind starts with Galton, who, along with many of his contemporaries, viewed intelligence as a heritable mental faculty. The structural view of intelligence was reinforced by massive intelligence testing and consolidated by the then-newly invented correlation and factor analytical technique. Efforts to delineate psychometric intelligence culminated with J. Paul Guilford’s mapping of various conﬁgurations of abili- ties based on content, process, and product, and John Carroll’s reanalysis of hundreds of psychometric stud- ies of human abilities. Although the structural view of intelligence implicitly assumes that intelligence is a capacity of some sort, some voices resist such reiﬁ- cation. Some scholars argued that psychometrically measured intellectual performance is better seen as an index of the effectiveness and efﬁciency of mind vis-a ` -vis an array of task conditions rather than mental entities. An alternative, functional account involves an understanding of the context in which performance is observed and assessed, as well as how the person carries out the task. Whereas the structural view deﬁnes intelligence squarely as a person charac- teristic, a functional view deﬁnes intelligence at the interface of an enactive person and an impinging environment, as ﬁt execution of behavior or perfor- mance in that context. In short, when the structural view sees competence as a personal trait, the func- tional view sees competence as context dependent. Intelligence: Nature or Nurture? The early pioneers of intelligence research differed with respect to whether mental capacity is heritable. Whereas Galton ﬁrmly believed that intelligence is largely a heritable quality, Binet considered it a human condition that can be modiﬁed through education and social interventions. Whether the quality called ‘‘intel- ligence’’ can be improved through education is still controversial today. There is a pervasive pessimistic belief that people cannot do much when it comes to intelligence. Supporting evidence comes from twin studies that show that as one reaches adolescence and adulthood, genetics seems to play an even more sig- niﬁcant role in one’s intellectual performance than when one is younger. On the other hand, evidence also indicates that intelligence is a malleable quality, and schooling and effective instruction make a differ- ence in one’s intellectual performance and cognitive organization of personal experience. Raymond Cattell developed a more differentiated scheme in which ﬂuid intelligence (Gf), the ability to manipulate complex information and detect patterns and relations, is interpreted to have direct biological underpinnings, whereas crystallized intelligence (Gc) reﬂects the cumulative effect of experience and edu- cation. Thus, Gf represents true genetically based intelligence and Gc is simply a derivative outcome of Gf acting upon experience and knowledge. There is a body of research on cognitive aging that seems to support the notion that Gf tends to decline with aging, but this trend is compensated for by the incremental changes in Gc. However, research also demonstrates that environment (likely including education) pro- duces greater effects on Gf than Gc, afﬁrming an opposite contention that ﬂuid abilities are among the most important products of education and experience. Intelligence and Intellectual Development 537 Indeed, an important task not fulﬁlled by the tradi- tional IQ tests is how to design instructionally rele- vant useful tests of ﬂuid abilities, showing how well students can transfer their learning to novel problems. However, when one examines what carries the most weight in general intelligence and in predicting future performance, crystallized intelligence seems to be a winner. Thus, the contention that Gf represents ‘‘natural ability’’ and Gc is simply ‘‘achieved compe- tence’’ is a controversial one. The nature-nurture issue is also related to whether one views intelligence as ultimately a structural fea- ture of the person, reﬂecting a capacity, or a relational and functional property, reﬂecting the person–task interaction. At least three major sources of intelli- gence can be identiﬁed: biological, domain experi- ence, and reﬂective. Biologically based intelligence (e.g., neural efﬁciency) may be difﬁcult to modify, but both domain experience and reﬂective thinking are subject to signiﬁcant environmental inﬂuences. Intelligence: General or Specific? At face value, standard intelligence tests are a com- posite measure of performance on a variety of tasks, originally intended to be th iswaybyBinet.Thepracti- cal utility of such a test may lie in the very fact that it is a gross measurement, like the Dow Jones Index, sam- pling as many ‘‘active performers’’ as possible to obtain an overall estimate. The paradox is that the broader the range of tasks a test covers , the less psychologically meaningful the test becomes. The trade-off seems to be between speciﬁcation of cognitive (and possibly motivational) proc esses given a cognitive task, and the breadth of representations of task conditions that affords performance consis tency, stability over time, and predictive validity. Practicality aside, it becomes theoretically problematic when one attempts to interpret various factor structures as indicative of how the mind is structured. Spearman, for one, ventured into such a speculation when he interpreted the general factor as ‘‘mental energy.’’ The three-stratum theory proposed by Carroll in 1993 represents an integration whereby cog- nitive abilities are represented as a continuum from the most general to the most speciﬁc. Yet many researchers questioned whether the factor structures can truly afford a theory of the structure of human abilities, or they just reﬂect a statistical artifact. Nevertheless, there is still a strong belief among many students of intelligence that general intelligence exists and that standard intelligence (IQ) tests mea- sure it well. When it comes to understanding the nature of general intelligence, some take a more reductionistic view by tracking it down to its neurobi- ological roots. Recent research of developmental biol- ogy also seems to suggest that prolonged thickening of cortices may underlie more advanced cognitive development of individuals whose IQ scores place them roughly at the top 3 % of the population. Others characterize general intelligence as the ability to deal with cognitive complexity; that is, more complex tasks require more mental manipulations of informa- tion, hence higher demands on mental capacity. Intelli- gence tests have been found to be correlated with high- complexity tasks more than low-complexity tasks, although exceptions also exist. Research on mental retardation also suggests deﬁcits in extracting abstract relations and principles. Although psychometricians have tried to map out human abilities, general or speciﬁc, for a long time, Howard Gardner and Robert Sternberg brought to the ﬁeld different breeds of theoretical perspectives and research evidence. Gardner views psychometrically measured IQ as representing a culturally narrow view of what is important for effective intellectual func- tioning. His theory of multiple intelligences, for good or ill, has successfully pluralized the concept of intel- ligence. Moreover, his argument that the mind is not an all-purpose information-processing device but is composed of many specialized modules dedicated to processing speciﬁc types of information was based on a set of neuropsychological evidence. The underlying argument that processes are always sensitive to content is now widely supported. In comparison, Sternberg’s triarchic theory represents a more complex system of theoretical propositions, encompassing cognitive, experiential, and contextual dimensions. The question of whether intelligence is unitary and general or pluralistic and dom ain-speciﬁc often carries a structuralist overtone concerning how the mind is innately structured and organized, regardless of envi- ronmental experiences. It does not take into account the possibility that mental capacities can also be shaped by task environments over time. Gardner expressed a structural view of intelligence when he conceptual- ized the mind as innately modular. His recent concep- tions seem to move toward a more functional view. A functional view of intelligence is by nature more domain-speciﬁc and context-bound. It does not presume, however, that the mind is innately domain-speciﬁc. 538 Intelligence and Intellectual Development Rather, domain-speciﬁc competence could be the prod- uct of adaptive efforts that might involve both domain- general and domain-relevant resources. Early Conceptions of Intellectual Development From the onset, developmental approaches to intelli- gence take a different approach from differential or individual difference approaches, assuming that it is a function of age-graded development; that is, sooner or later a person will go through similar developmen- tal changes in his or her intellectual structures and functions. Although working in Simon’s laboratory on child psychology, Jean Piaget saw something that was apparently missed by psychologists of his time; that is, children’s thinking has their own ‘‘logic’’ that is dif- ferent from adults’. Piaget envisioned a developmental process that is completely outside the radar of differen- tial psychology: At different developmental levels, intelligence has its different structures and functions, showing different organizational principles in cognitive functions. Such a developmental orientation has turned out to be extremely fruitful in understanding how chil- dren act on and represent their world, and how educa- tors can facilitate children’s intellectual development. Oddly enough, although Piaget differed signiﬁcantly from differential psychologists in his theoretical inter- ests, his assumption about the existence of deep struc- tures of intellect (schemes and operations) reﬂects a conviction similar to that of differential psychology; that is, mapping the structure of intellect is possible just as is mapping the physical structures of an organism. However, Piaget also started off a tradition of devel- opmental research that shows little interest in individ- ual differences. The separate paths of developmental and differential psycholog y have yielded ﬁndings and theories that are virtually noncommunicative with each other. On one hand, developmental researchers are seeking an understanding of how an ‘‘average’’ child develops intellectually or cognitively over time, while regarding individual differe nces as trivial or ‘‘noises’’ to be dismissed. On the other hand, consistent differ- ences in intellectual perfo rmance observed within any age have prompted differential psychologists to develop age norms for individual differences in academic and intellectual performance calibrated to months of age. Because IQ scores are age-normed, intellectual develop- ment from such a different ial point of view is simply how stable these age-normed individual differences are over time. Both traditions can be criticized as missing an important part of intellectual development. On the differential side, age-normed rank order scores mask incremental changes in intellectual functioning and qual- itative changes or reorganization of cognitive functions. Indeed, age-normed standard scores help perpetuate the notion that intelligence is a ﬁxed quality, and by a further leap of faith, a genetically ba sed individual difference. On the normative developmental side, in seeking to understand a typical or ‘‘average’’ child at a given age, the normative developmental psychology is also guilty of neglecting vast individual differences in intellectual functioning and development, not only in terms of psy- chometrically measured indi vidual differences, but also in terms of different pathways and trajectories. Metatheoretical Assumptions Guiding Theory Building There are three main metatheoretical frameworks guiding research and theorizing: mechanistic, organis- mic, and contextualist. Mechanistic perspectives would simply see intellectual development as a deriva- tive outcome of changes in cognitive machinery (e.g., working memory capacity). Organismic per- spectives would view intellectual development as reg- ulated by internal rules of growth (e.g., disequilibrium or innate skeletal principles). Contextual perspectives would view intellectual development as fundamen- tally embedded in the broader sociocultural context and situated at the interface of person–environment interactions (e.g., zone of proximal development). Whether development of intelligence should be considered differential, quantitatively or qualitatively, is a matter of whether the core ontological commitments endorse a differential provision. Mechanistic approaches can lead to theories of differential intellectual develop- ment, in reading or other cognitive development, as long as the development is traced to the basic cognitive struc- tures and operations. Organismic principles can also yield insights into differential development of intellec- tual competence by specifying when individuals diverge in how they negotiate the max imizing of mastery of the environment on one hand, and the optimizing of one’s affect on the other. The contextualist doctrine, by far, offers the most unconstrain ed versions of intellectual development. On one hand, it is liberating by permitting various environmental forces (including fortuitous events and life-changing encounters) and individuality (with all its idiosyncrasies) to play a role in intellectual Intelligence and Intellectual Development 539 development. On the other hand, it opens up a multitude of developmental possibili ties that may prove unwieldy for scientiﬁc research or ma yfallshortofparsimony. Again, there is a tension on the nomothetic or universal versus ideographic or parti cularistic continuum where cognitive and intellectual development are concerned. Notwithstanding the challenges of integrating differential and developmental theories and research in a theoretically coherent manner, developmental approaches stand the best chance to address the three basic questions regarding intelligence discussed earlier. Traditional intelligence testing, and its meth- odological backbone, factor analysis, represent a static method and therefore cannot afford insights into dynamic person–environment interactions and changes of intellectual functions over time. By studying the tem- poral sequences of person–t ask or person–environment transactions, the structural-functional tension can be addressed as an issue of how initial structure facilitates functionality and how functionality leads to new struc- tures. By studying nature and nurture in reciprocation and interaction, how biological preparedness and envi- ronmental affordances and constraints work in concert as a system can be further explicated (e.g., passive, evocative, and active correlations between the person and the environment). By studying intellect in the making, it is possible to show contributions of both domain-speciﬁc and domain-ge neral cognitive resources. Because individual differences in intellectual develop- ment are one of the most important educational consid- erations, it is imperative from an educational point of view to incorporate a differen tial provision of how indivi- duals differ in their intell ectual development, and how learning and instructional act ivities enhance intellectual functioning and development. New developments in the ﬁelds of developmental and differential psychology show good promise of such a theoretical integration and consilience. New Developments on Intellectual Functioning and Development There are ﬁve aspects of current thinking that facili- tate an integrative approach to intellectual functioning and development: 1. The importance of functional and social contexts, 2. The reciprocal relationship between functioning (including learning) and development, 3. The role of nonintellective factors in intellectual functioning and development, 4. A life span perspective on intellectual development, and 5. Developmental processes and mechanisms leading to differential intellectual development. The Role of Functional and Social Contexts In a narrow sense of the term, context refers to a set of immediate conditions where speciﬁc intellec- tual performance is observed. It can include the nature of tasks involved, situations in which the performers ﬁnd themselves, and instruments and criteria used to make observations and assessment as well as tem- poral changes of strategies and performance. This emphasis shifts from an exclusive focus on intellectual performance as a person variable (i.e., competence) to a focus on intellectual performance as relational property of person–task interaction in real time (i.e., perfor- mance). Microgenetic methods, which follow indivi- duals’ interactions with a speciﬁc task environment intensively for days and weeks, uncover intraindividual as well as interindividual variability in children’s intel- lectual performance. Such an idiographic approach to studying real-time, micro-level developmental pro- cesses has signiﬁcantly changed the way development is understood, from a view of monotonic incremental or structural changes to that of variation, selection, and optimization, not unlike the process of biological evo- lution. Such a micro-level process view of development integrates functioning and development. At a more abstract level, context can mean differ- ent functional contexts where intelligent behavior may entail different sets of capabilities and propensi- ties, such as academic versus practical settings. Thus, Brazilian children can perform well on ‘‘street math’’ but cannot do as well on equivalent ‘‘school math.’’ Professional racetrack gamblers are capable of sophis- ticated reasoning in their domain of expertise but do not perform superbly on standard intelligence tests. Each domain or ﬁeld may have different sets of abilities and propensities and differen t threshold requirements. Even within academics, requirements for successful adaptations may be quite different. An emphasis on context also implies a speciﬁc social milieu or cultural context where intellectual f unctioning and development take place. Lev Vygotsky insis ted that all higher mental functions initially occur at the social level and then 540 Intelligence and Intellectual Development are gradually internalized with practice. The role of more competent others is indispensable. The notion of zone of proximal development treats intellec- tual development as fundamentally mediated by socializing agents. The new movement on ‘‘situated cognition’’ also stresses the distributed nature of intellectual functioning. The Relationship Between Functioning and Development Current thinking and research involve a new under- standing of relationships between functioning and development, and learning and development. Develop- ment is traditionally conside red structural organization and reorganization of experience and cognitive and behavioral functions that evolve over time. Experi- ence and knowledge facilitate this process but do not change its nature. Learning, on the other hand, is seen as a process of the acquisition of new knowl- edge and skills that has little bearing on the cognitive infrastructure. This view has been changed signiﬁ- cantly. Research shows, for example, that children who are chess players performed meaningful chess memory tasks at a much higher level than adults who had not learned chess. The study demonstrates that knowledge signiﬁcantly alters basic cognitive functioning. The emergent expertise literature pro- vides compelling evidence that knowledge enables sophisticated reasoning and problem solving. Now, learning is seen an integral part of development. Integrating learning and knowledge into develop- mental theory entails a more functional, rather than structural, view of intellectual development; in other words, intellectual development is likely more con- textual than organismic. It opens a new way of looking at intellectual development that stresses the importance of contextual experiences (including educational experience) and the facilitative role of others while incorporating internal principles gov- erning the behavior of the organism at a speciﬁc developmental level. Urie Bronfenbrenner and Stephen Ceci argued that biological potential for learning and intellectual devel- opment can be realized only through proximal pro- cesses, that is, transactional experiences with speciﬁc environmental contexts both immediate and mediated. Accordingly, learning potential in a given situation, which is typically considered an important aspect of intelligence, can be built into a differential theory of intellectual development. This is in keeping with Vygotsky’s notion of zone of proximal development and its practical application, dynamic testing, which is not assessing what the person already knows, but how well the person learns in real time when given instruc- tional hints and prompts. Dynamic testing helps diag- nose a child’s ability to learn at a speciﬁc level of competence, and it informs educational interventions aimed at assisting children with learning difﬁculties, reminiscent of what Binet was doing during the incep- tion of intelligence theory and measurement. The Role of Nonintellective Factors Both psychometric theorie s of intelligence and age- graded normative theories of intellectual development tend to be ability-centric theories; that is, the kind of abilities a person can display in a perform-on-demand condition. Such ability-centric approaches elicit maxi- mal performance at the expense of neglecting a per- son’s typical engagement, which has much to do with personal dispositions, such as openness to experience. Kurt Fischer made a related d istinction in developmen- tal theory between functiona l-level and optimal-level development, an important revision of Piaget’s theory, which assumes by default an optimal-level develop- ment (e.g., emphasizing what adolescents potentially can do, rather than what they are actually able to do). Nonintellective factors m ay play what Bronfenbren- ner called a development-instigative role. Intellectual dispositions, such as risk-taking and open-mindedness, also play an important role in how one approaches an intellectual challenge. Intel lectual dispositions may lie in the intersection of personality and intelligence. It may be argued that nonintellective factors such as interest and persistence may be particularly important in differential intellectual development, as much knowledge and skill building entails sustained engage- ment and deliberate practice. Personal agency in intellectual development can also take a more conscious form. Development is now seen as a self-modifying process, part of intellectual development as self-engende red changes. All of these conceptions involve a distinct role of the consciousness and self. One example is what Annette Karmiloff- Smith called representational redescription made by children to articulate otherwise implicit knowledge. Rather than seeing intellec tual development as funda- mentally cognitive reorganization due to changes in cog- nitive infrastructure or architecture, Karmiloff-Smith Intelligence and Intellectual Development 541 sees the role of consciousness as reﬂecting a developing human tendency to exercise cognitive and metacognitive control over aspects of their environment. This is in line with a renewed emphasis on reﬂective thinking, which was stressed decades ago by Dewey as quintes- sential for effective intellectual functioning. Education seems to play an important role to enculturate such habits of mind. A Life Span Perspective Current thinking goes beyond adolescence to embrace a life span perspective on intellectual func- tioning and development. Biologically, the Gc and Gf distinction provides a useful perspective on when one reaches cognitive maturity and when cognitive or neu- ral efﬁciency starts to decline with the aging process. However, intellectual development is constrained, but not dictated, by biological differences and changes; in other words, intellectual development is not as canalized as other aspects of development, such as language development. This point becomes clearer when one views intellectual development (Gf or Gc) as often enabled by learning experiences and educa- tion. It can also be argued that much of intellectual development occurs beyond adolescence. There is a growing body of research on adults’ differential intellectual development. The ﬁndings of behavioral genetics research that identical twins are more alike when they reach and go beyond adolescence, and that nonshared environmental factors play a more distinct role in later development, can be incorporated and further tested in such research. The topic of intellec- tual functioning and development also overlaps with expert performance and development of expertise. Coherent Accounts of Processes and Mechanisms Attempts at integrating differential and develop- mental approaches have been made at the level of childhood, adolescence, and adulthood. However, an important task is to provide processes and mechan- isms that account for differential development in a coherent manner. Some scholars suggest that mental development in the ﬁrst 6 years is similar for every- one and then diverges due to both environmental and genetic forces. Others provide motivational principles explaining differential pathways and trajectories. Robert Siegler’s overlapping-wave theory of cognitive development also potentially explains how differen- tial development occurs. Studies that incorporate both comparative and microgenetic methods may ulti- mately be capable of addressing the two central issues of intellectual functioning and development: a process account of development that entails the role of adap- tive functioning, and a developmental theory that is capable of addressing differential pathways and tra- jectories. In this way, differences between nomothetic and idiographic approaches to understanding intelli- gence and intellectual development can be eventually resolved or reconciled. David Yun Dai See also Bell Curve; Emotional Intelligence; Fluid Intelligence; Gifted and Talented Students; Metacognition and Learning; Multiple Intelligences; Triarchic Theory of Intelligence; Vygotsky’s Cultural-Historical Theory of Development Further Readings Carroll, J. B. (1993). Human cognitive abilities: A survey of factor-analytic studies. Cambridge, UK: Cambridge University Press. Ceci, S. J. (1996). On intelligence: A bio-ecological treatise on intellectual development (2nd ed.). Cambridge, MA: Harvard University Press. Ericsson, K. A., Charness, N., Feltovich, P. J., & Hoffman, R. R. (Eds.). (2006). The Cambridge handbook of expertise and expert performance. New York: Cambridge University Press. Gardner, H. (2003). Three distinct meanings of intelligence. In R. J. Sternberg, J. Lautrey, & T. I. Lubert (Eds.), Models of intelligence: International perspectives (pp. 43–54). Washington, DC: American Psychological Association. Guilford, J. P. (1967). The nature of human intelligence. New York: McGraw-Hill. Lohman, D. F., & Rocklin, T. (1995). Current and recurrent issues in the assessment of intelligence and personality. In D. H. Saklofske & M. Zeidner (Eds.), International handbook of personality and intelligence (pp. 447–474). New York: Plenum. Messick, S. (1992). Multiple intelligences or multilevel intelligence? Selective emphasis on distinctive properties of hierarchy: On Gardner’s Frames of Mind and Sternberg’s Beyond IQ in the context of theory and research on the structure of human abilities. Psychological Inquiry , 3 , 365–384. Neisser, U., Boodoo, G., Bouchard, T. J., Boykin, A. W., Brody, N., Ceci, S. J., et al. (1996). Intelligence: Knowns and unknowns. American Psychologist , 51 , 77–101. 542 Intelligence and Intellectual Development Perkins, D., & Ritchhart, R. (2004). When is good thinking? In D. Y. Dai & R. J. Sternberg (Eds.), Motivation, emotion, and cognition: Integrative perspectives on intellectual functioning and development (pp. 351–384). Mahwah, NJ: Lawrence Erlbaum. Sternberg, R. J. (Ed.). (2000). Handbook of intelligence. Cambridge, UK: Cambridge University Press. I NTELLIGENCE Q UOTIENT (IQ) Intelligence is a construct that has been proposed by psychologists to underlie much of human behavior and is a signiﬁcant factor contributing to an individ- ual’s ability to do some things more or less well. Most would agree that some children are better at math or language arts than others, or that some hockey players or musicians are gifted in compari- son to their peers. It might be argued that some indi- viduals are born that way, whereas others have the beneﬁt of good environments and learning opportu- nities that can build on their basic abilities. The intelligence test, and resulting intelligence quotient or IQ, is a means for assessing and measuring intel- ligence, with the results often used to classify or select persons or predict such outcomes as school achievement. Both the construct of intelligence and its measure- ment are not new, and both existed well before the advent of psychological science. Historians have traced the forerunner of current cognitive ability and achievement assessment to more than 2000 years B . C . Although intelligence has been studied in a number of ways, from an early emphasis on sensory processes to the more current attention given to brain-imaging techniques, the mainstay in the study and assessment of intelligence has been the IQ test. Psychologists not only assess intelligence but also study how intelli- gence is expressed; what causes it; and how it contri- butes to understanding, explaining, predicting, and even changing human behavior. Despite intelligence being a much studied area of psychology, there is still considerable controversy and emotion regarding the use of the IQ and intelligence tests and the results gleaned from them in such contexts as schools and industry to describe both individuals and groups. Given continued advances in the theories of intelligence and cognitive assessment instruments, the issue appears to be less with the con- structs and the tests used to measure it, and more with how this information is or can be used. Theories of Intelligence Psychology joined the scientiﬁc community in the late 1800s, and since then, a number of theories outlining human intelligence, accompanied by a huge body of research, have emerged. The hallmark of science and scientiﬁc inquiry is the creation of theories and the pursuit of empirical support for the hypotheses that are generated by and from a particular theory. The current theories of intelligence attempt to explain what it is, what causes it, and what intelligence tells us about other human behaviors. Although research has demonstrated that there is a considerable genetic component to intelligence, it is also recognized that intelligence is an acquired ability that reﬂects opportunity and experience such as comes from effective schooling and home environ- ments. Studies showing the remarkable similarity in measured ability between twins, whether reared together or apart, provide much evidence for a genetic foundation to intelligence. However, intelligence appears to be polygenic rather than located on a spe- ciﬁc gene. Studies have also shown that animals raised in very restricted in contrast to ‘‘rich’’ environ- ments not only show considerable differences in, say, their capacity to solve problems, but also show an impact on their brain structures (e.g., number of neural connections). As well, research has shown how the effects of poverty and restricted educational opportunities can negatively inﬂuence human devel- opment, including intelligence. Among the environmental factors that are known to directly inﬂuence brain functioning and thus intellectual development and expression are various medical conditions, neurotoxins, drugs such as alco- hol (certainly during pregnancy, as observed in chil- dren diagnosed with fetal alcohol syndrome), and chemical pollutants such as lead and mercury. Almost anything that negatively affects the brain, such as head injury and oxygen deprivation, will have small or large observed effects on intelligence and its expression. Less obvious but just as impor- tant are such additional factors as motivation, self- concept, and anxiety, all of which can inﬂuence a person’s score on an IQ or intelligence test and their everyday functioning at work or school. Intelligence Quotient (IQ) 543 Culture also affects the expression of intelligence. Although there is a universal ability related to the capacity to acquire, store, retrieve, and use information from everyday experiences as well as from direct teaching (e.g., school), how this is expressed, the con- tent of a person’s response to a question, and the lan- guage used in providing an answer to a question all reﬂect the interaction between the person’s genetic capacities and the environmental opportunities for intelligence. In addition, arguments have been made that what constitutes intelligence may vary across cul- tures and that different ethnic groups may have differ- ing, but equally intelligent, reasoning strategies. On the other hand, the successful ad aptation of contemporary assessment instruments for use in a large number of countries suggests that centr al abilities and capacities comprising intelligence ma ybesharedacrosscultures. It should also be pointed out that intelligence is also a developmental construc t. A 5-year-old child has a very different view of, say, cause–effect relationships or the understanding of number concepts than does a 15-year-old in Grade 10 or a 35-year-old with a uni- versity degree or a 50-year-old working in a factory. Brainmaturationverymuchinﬂuencesthequalitative description of intelligence. At the same time, it has been demonstrated that inte lligence does change across the life span, with some kinds of intelligence referred to as crystallized intelligence (e.g., a person’s knowl- edge of words and language, learned skills such as solving arithmetic problems) more likely to remain unaffected and possibly continue to improve with age than are abilities reﬂecting ﬂuid intelligence and speed of processing information (reﬂecting neural efﬁciency), barring, of course, dementia and other diseases under- lying cognitive decline. Another debate found in theoretical discussions and observed in models of intelligence is centered on whether intelligence is a single characteristic or is composed of several or even multiple abilities. These views can be traced back to the turn of the previous century, when psychologists such as Spearman argued that intelligence was a set of speciﬁc but related fac- tors that resulted in an overarching general factor (essentially similar to the current full-scale IQ [FSIQ] score found on many tests). In contrast, Thurstone proposed that intelligence was made up of a number of primary mental abilities that could not be captured in a single summary score or an FSIQ. Today’s tests and models continue to reﬂect these divergent viewpoints. For example, psychologists such as Guilford have proposed that intelligence may have 120 or more facets, while Wechsler has argued for the relevance of the FSIQ (but also the importance of look- ing at both verbal and nonverbal performance). Other current models proposed by St ernberg, describe intelli- gence along the lines of practical, analytical, and crea- tive abilities, whereas Gardner suggests that there are likely eight to nine core kinds of intelligence reﬂecting, for example, interpersonal intelligence (required for effective social interaction and communication), kines- thetic intelligence (obser ved in athletes who excel in their sport), musical ability (found in performers and composers), and logical-m athematical i ntelligence (reﬂectingthecapacitytoreasonlogicallyinmathe- matics and science such as physics). Other views, drawing from the work of Piaget, focus more on how intelligence develops (in stages) and how it can be encouraged through direct in struction and supportive learning environments (e.g., instrumental enrichment). Thus, there is quite some diversity in how intelli- gence is deﬁned, determining the key factors that affect its development and expression, and how it is best measured. Although this may be perplexing to some, it does show how complex intelligence is and, even more so, how very complex human behavior is. At the same time, a great deal is known about intelligence and what it tells us about human behavior. For example, intelligence tests, yielding a measure of general men- tal ability, are one of the best predictors of student achievement and success in elementary schools. On the other hand, and as expected, intelligence tests have been found to be more limited in predicting achieve- ment among intellectually h omogeneous populations. For example, university s tudents generally possess average or above-average levels of intelligence such that divergent performance in this group appears to be more highly related to speciﬁc cognitive competencies (e.g., high aptitude in math ) and personal attributes (e.g., motivation, study skills). Intelligence is addition- ally considered a key factor in understanding human capacity to manage stress and develop resiliency, psy- chological well-being, and even longevity. History of Intelligence Testing The very earliest tests of intelligence were not based on any particular scientiﬁc views and in many instances simply showed the wide or narrow range of perfor- mance on such tasks as strength of grip or pitch dis- crimination. More to the point, these tests did not tell us 544 Intelligence Quotient (IQ) about other human characteris tics that, by expectation, they should. If intelligence is an underlying capacity that inﬂuences how well a person does in school, or a person’s accuracy at solving arithmetic problems, or the speed at which he or she can perform other mental tasks, then the tests should be correlated with those behaviors and be able to predict how well a person may perform on those tasks requiring intelligence. In contrast to the earliest tests of intelligence, more recent intelligence tests ha ve resulted from extensive research efforts, while still garnering a great deal of misunderstanding from the general public. The ﬁrst suc- cessful intelligence tests were developed by Binet and Simonattheturnofthelastcenturyandusedinthe schools of Paris, France, to help identify and classify schoolchildren according to t heir ability to learn and whether they would beneﬁt from regular or special school programs. A short time later, these tests were introduced in the United States. Along with the Army Alpha and Beta intelligence te stsusedtoscreenmilitary recruits during World War I, there was growing opinion that intelligence tests had considerable value for pur- poses ranging from personnel selection to identifying children who were intellectually gifted or retarded. These early landmarks in the history of testing laid the foundation for the advancement and proliferation of subsequent intelligence tests. For example, the ﬁrst intelligence test created by Wechsler, in 1939, has evolved into several recent ly published tests for asses- sing intelligence from pres chool years to age 89, and these tests have now been adapted for use in a large number of countries. The number of tests available to psychologists for assessing cognitive abilities has grown considerably over the past 60 to 70 years. Current Intelligence Tests Today’s intelligence tests vary from very brief mea- sures that assess only a limited or narrow part of the broader intelligence framework (e.g., Raven’s Matri- ces) to large comprehensive batteries that tap many dif- ferent aspects of intelli gence ranging from verbal comprehension and spatial reasoning ability to memory and processing speed (e.g., Woodcock-Johnson Cogni- tive). The large number of tests available also includes tests speciﬁc to various age ranges, both group and individually administered tests, brief and comprehen- sive batteries, and modiﬁed tests for use with, for example, hearing-impaired clients, or clients who are nonverbal or who are less proﬁcient in English. The majority of intelligence tests require a well- trained psychologist administering subtests that require the client to complete a range of tasks. Two broad types of tasks are used on intelligence tests—verbal and nonverbal. Verbal tasks generally entail a verbally presented prompt or question and require an oral response such as deﬁning words (What is a hammer?), responding to general information questions (What is the distance between the earth and the moon?), or iden- tifying similarities between two words (How are con- vention and meeting alike?). Nonverbal tasks usually involve visual stimuli or materials and/or require a psy- chomotor response such as copying geometric patterns using blocks, identifying important parts that are miss- ing from both common and uncommon objects, or identifying patterns within a visual array. Although instructions and prompts to nonverbal tasks are some- times given orally, verbal r equirements are minimized within some tests through the use of gestures, model- ing, or pictorial directions. Both verbal and nonverbal tasks can be employed to measure a wide range of cognitive abilities and capabilities. For example, short-term memory may be assessed through a task requiring a student to repeat a string of presented numbers (verbal task) or to touch blocks in a previously observed order (nonverbal task). Regardless of the types of questions used, the psychologist is careful to ensure that administration and nonintellective factors do not confound the infor- mation gleaned from these tests. For example, it is necessary to make accommodations for persons with, for example, visual, auditory, or motor problems, lest these interfere with their performance on tests that should be speciﬁcally tapping intelligence. The raw scores obtained on intelligence tests are given meaning by comparing them to the performance of large and appropriate reference groups. These group performance indicators, called norms , are based on extensive standardization studies whereby the test is administered to large numbers of examinees to both ensure that the test is working well and to build a com- parison group that is similar on key characteristics such as age and that reﬂects the composition of the larger community (ethnicity, sex, socioeconomic sta- tus, etc.). An individual’s raw scores on the parts and the whole test are then converted to standard scores through the use of tables; these standard scores are often referred to as IQ scores, and in the case of, say, the Wechsler Intelligence Scale for Children–Fourth Edition (WISC-IV), four index scores assessing verbal Intelligence Quotient (IQ) 545 comprehension, perceptual reasoning, working mem- ory, and processing speed, together with a full-scale IQ, are reported with the average score set at 100. Furthermore, based on the use of normal curve pro- portions, scores of 130 would represent intellectually gifted persons; this score is obtained by less than 2 % of the population. Scores of 115 suggest high average ability that exceeds the scores obtained by about 84 % of the population. In contrast, scores of 85 are seen as low average; a person with this score would have scored higher than 16 % of the population, but some 84 % of the population scored higher than him or her. For an intelligence test to be truly useful, it must demonstrate sound psychometric properties that include reliability and validity. Fo r a test to be reliable, it should have a minimum of measurement error, thereby measuring with consistency and precision. Thus, a FSIQ score will have some error associated with it and should never been taken as an exact measure but rather one that reﬂects a range wherein the person’s true score is likely to be. Validity means that the test in fact measures what it is intended to measure. If the test is supposed to measure acqui red knowledge or crystal- lized intelligence, then it should do just that and not do something else. Although it can be said that current intelligence tests are among the very best measures used by psychologists, certa in caveats still apply. For example, no one test tells everything about a person’s full intellectual ability, because other factors, such as depression, low motivation, test anxiety, or cultural factors, can inﬂuence intelligence test scores. Current Uses of Intelligence Tests The use of IQ and other intelligence tests is a complex process that requires a comprehensive understanding and training in such areas as test principles (reliabil- ity, validity, test construction, norm groups, types of scores); human development; and test administration and interpretation. As such, certain state and provin- cial restrictions exist that limit who is permitted to administer and interpret the r esults. In general, the use of intelligence tests is limited to psychologists or other such individuals who have a minimum of graduate- level training in psychology and assessment. Although most commonly used by school or clinical psychologists within school a nd clinical settings, intel- ligence tests may also be used by psychologists within other specializations (e.g., counseling, industrial organi- zation, research) and in such additional settings as community and state agencies, workplaces, universi- ties, and private practices. In part, the purpose for administering an intelli gence test may vary to some extent depending on the reason for referral and who is administering it and in which setting. A school psy- chologist may use the results of an intelligence test to help decide which students should be selected for a gifted program, whereas a neuropsychologist may use the results to assist wit h determining the location and extent of a brain injury. In general, intelligence tests provide information that can inform a wide range of diagnostic and decision-making processes. Among the most common uses of in

bundet

tel ligence tests are to assist with diagnostic and eligibili ty decisions, intervention planning, progress monitor ing, and research into cogni- tive functioning. Diagnostic and Eligibility Decisions Originating with Binet and Simon’s development of an intelligence scale to identify children who would bene- ﬁt from regular and special education, one of the pri- mary uses of intelligence tests has been and continues to be in making diagnostic and eligibility decisions. In particular, various classiﬁcation systems, laws, and pol- icies use an individual’s leve l of intellectual function- ing (IQ or equivalent) for the purposes of identifying particular groups of individuals and for determining eli- gibility for services. An example of this is the diagnos- tic criteria for mental retardation (MR) outlined within the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition, Text Revision) (DSM–IV–TR), which assigns the sever- ity of MR based on obtained IQ scores (see Table 1). Such diagnoses or classiﬁca tions often assist individ- uals in accessing appropriate assistance from medical, educational, or mental heal th specialists and support an individual’s eligibility for particular programs or services. Certain government agencies, for example, provide adults with IQs below a certain level with addi- tional ﬁnancial and individualized supports to participate more fully within society. Whereas a global assessment of intellectual func- tioning is most commonly used for classiﬁcations requiring either low or high levels of cognitive func- tioning (e.g., mental retardation, giftedness), IQ scores within the normal ranges can also provide useful information for diagnostic and eligibility decisions. For example, the assessment of at least average cogni- tive abilities can support a diagnosis of a learning 546 Intelligence Quotient (IQ) disability when found in conjunction with particular cognitive processes that are compromised such that the child has considerable difﬁculty in the acquisition and use of certain skills (e.g., reading). Alternatively, an assessed absence of cognitive deﬁcits can assist a psychologist in determining if noncognitive factors are contributing to an individual’s poor performance in the classroom. By ruling out cognitive factors, the contributions of such additional factors as motivation or emotional problems can be explored further. Intervention Planning As with diagnostic and eligibility decisions, intelli- gence tests have a long history of being used to inform intervention planning. At the broadest level, identiﬁcation of causes or particular diagnoses pro- vides valuable information for treatment and interven- tion. For example, an individual identiﬁed as having severe mental retardation will likely require a program that provides extensive training in elementary self- care and social skills. Evaluation of intellectual func- tioning can also provide information about the general strategies and approaches to intervention. Individuals with higher levels of intelligence do well with indirect or inquiry-based instruction, whereas studies demon- strate the beneﬁts of functional, systematic, carefully sequenced, direct instruction for individuals assessed with lower levels of general intelligence. The link between a child’s global IQ score and intervention is often indirect, however. For example, by knowing that a ﬁrst-grade child is struggling in mathematics due to an ability problem, interventions may focus on practical activities to increase that child’s counting skills and understanding of number concepts. In contrast, a child with motivational pro- blems may demonstrate improvements if incentives for work completion are provided. It is also possible to more acutely identify appropriate intervention prac- tices and remedial activities when speciﬁc cognitive abilities and processes are examined. In particular, the increased range of narrow and broad abilities assessed by contemporary intelligence tests provides an avenue for parsing out speciﬁc abilities, experiences, or learn- ings that may be contributing to depressed achievement or impaired cognitive performance. Results from cur- rent tests may suggest the need to target such areas as an individual’s visual and auditory processing skills, short- and long-term memory stores, or processing speed. The greater speciﬁcity that is available from contemporary assessments is akin to the doctor who provides speciﬁc feedback regarding a patient’s blood pressure, cholesterol level, and heart rate in the assessment of physical health. Signiﬁcant problems identiﬁed in any one of these areas are associated with a prescribed course of action. Similarly, by identify- ing a particular cognitive ability that is underdevel- oped, speciﬁc interventions can be implemented or compensatory supports put in place to ameliorate the identiﬁed weakness. Monitoring Progress Although an individual’s cognitive functioning is con- sidered to be fairly stable over time, changes in IQ scores can result from a number of factors, such as development, intervention, environment, and injury. As such, intelligence tests can be used to monitor an indi- vidual’s intellectual abilities over time. For example, cognitive scores can be compared to determine the effectiveness of special e ducational programs, treat- ment (medications), and training. For individuals sus- taining cognitive injuries, availability of assessment of premorbid cognitive functioning can assist with deter- mining the extent of such injury and also monitoring the rate at which cognitive abilities are recovered. Research Of course, what is known about intelligence, in terms of what it is and how it contributes to understanding, explaining, predicting, and even changing human behavior has been heavily informed by intelligence tests. By studying the results of intelligence testing Table 1 Diagnostic and Statistical Manual of Mental Disorder (Fourth Edition, Text Revision) (DSM–IV–TR): Speciﬁed Intellectual Impairment (IQ) Required for Assessing Severity of Mental Retardation (MR) Severity of Mental Retardation IQ Level Mild 50–55 to approximately 70 Moderate 35–40 to 50–55 Severe 20–25 to 35–40 Profound Below 20 or 25 Intelligence Quotient (IQ) 547 within particular groups of individuals or across time, psychologists’ understanding of what constitutes intelli- gence and the processes underpinning cognition deep- ens. For example, the use of intelligence tests with elderly populations has provided insight into how the brain ages and processes spared from more rapid cog- nitive decline. Similarly, psychologists’ understanding of the basic psychological fun ctions that are affected in certain disabilities and diso rdersisdirectlyinformed through the administration of cross-battery assessment approaches, in which tasks from a number of IQ tests are administered. Intelligen ce tests also feature promi- nently in research into the effectiveness of certain treat- ments or interventions. Summary Among the limitations cited against the use of intelli- gence tests includes possible error in measurement, curriculum insensitivity, bias against students of diverse linguistic and cultural backgrounds, and lim- ited instructional applicability. Although merit and support can be found for arguments on both sides of such debate, recent research advancements and cur- rent ‘‘best’’ professional practices within the ﬁeld of assessment serve to bolster arguments in favor of using intelligence tests in describing individual differ- ences. First, the more comprehensive structure of human cognitive competencies and abilities assessed by contemporary intelligence tests greatly increases the utility of such instruments to better inform a wide range of diagnostic and decision-making processes. Second, it was once believed that intelligence is what the test measures, but it is currently acknowledged that contemporary intelligence tests yield information that only partially explains or accounts for ‘‘intelli- gence.’’ There is additionally the recognition that test results need to be viewed in consideration of a myriad of additional factors (social, emotional, behavioral) and other information and knowledge about an individual (family structure). Psychologists’ under- standing of what constitutes intelligence and how it is developed and expressed is additionally enhanced through complementary avenues of investigation that use alternative methods (e.g., brain imaging techniques) or investigate aspects of intelligence pre- viously unattended to (e.g., emotional intelligence). When contemporary assessment instruments are paired with multidimensional approaches to assessment, the test results, including thos efromreliableandvalid intelligence tests, serve as invaluable components in many diagnostic and decision-making processes. Michelle A. Drefs and Donald H. Saklofske See also Creativity; Individual Differences; Intelligence and Intellectual Development; Intelligence Tests; Multiple Intelligences Further Readings Birney, D. P., & Sternberg, R. J. (2006). Intelligence and cognitive abilities as competencies in development. In E. Bialystok & I. M. F. Craik (Eds.), Lifespan cognition: Mechanisms of change (pp. 315–330). New York: Oxford University Press. Carroll, J. B. (1993). Human cognitive abilities: A survey of factor-analytic studies. New York: Cambridge University Press. Deary, I. (2001). Intelligence: A very short introduction. Oxford, UK: Oxford University Press. Flanagan, D. P., & Harrison, P. L. (2005). Contemporary intellectual assessment: Theories, tests, and issues (2nd ed.). New York: Guilford. Gardner, H. (2006). Multiple intelligences: New horizons in theory and practice. New York: Basic Books. Gregory, J. R. (2007). Psychological testing: History, principles, and applications (5th ed.). New York: Pearson. Kamphaus, R. W. (2005). Clinical assessment of child and adolescent intelligence (2nd ed.). New York: Springer. Matthews, G., Zeiner, M. T., & Roberts, R. D. (2002). Emotional intelligence and myth. Cambridge: MIT Press. Priﬁtera, A., Saklofske, D. H., & Weiss, L. G. (2005). WISC-IV clinical use and interpretation: Scientist- practitioner perspectives. New York: Academic Press. Sattler, J. M. (2001). Assessment of children: Cognitive applications (4th ed.). San Diego, CA: Author. I NTELLIGENCE T ESTS An intelligence test is a structured situation designed to elicit information about the cognitive abilities of an individual. The test may be administered individually or in a group. There are certain advantages to having a test administered to a person one-on-one with a trained examiner, typically a school psychologist. Noreadingneedberequiredonthepartoftheexam- inee, so it is possible to test young children and people of limited literacy. And an empathic and perceptive examiner can maintain a continuing high level of motivation on the part of the examinee. 548 Intelligence Tests Furthermore, the examiner can observe aspects of the examinee’s behavior in the testing session that are not reﬂected in the numerical score but that would assist in diagnosis. On the other hand, individual testing is expensive—prohibitively so if information is desired on each child in a school or on each applicant for admission to an educational institution or a training program. Primarily because o f cost considerations, most ability testing has become group testing, using paper- and-pencil or computer-admin istered testing instruments that can be scored objectively. Scores on intelligence tests are normally distrib- uted in large populations and are usually reported on a scale in which 100 indicates average intelligence. Scores are scaled so that about the top 16 % of the population will receive scores of 115 or above, the top 2.5 % will receive scores of 130 or above, the bot- tom 16 % will receive scores of 85 or below, and the bottom 2.5 % will receive scores of 70 or below. The majority of people—68 % of the population—will have scores clustered around 100, ranging from 85 to 115. The typical intelligence test will have a variety of items designed to tap different aspects of the person’s cognitive abilities. Some of the items may ask for spe- ciﬁc pieces of information, such as how many years there are in a decade or how much change a person would receive if he or she bought an article of cloth- ing costing $18.67 and gave the clerk a $20 bill. Other questions might ask about objects missing or out of place in a picture; still others would be tests for memory, such as repeating back a list of ﬁve digits that has been read aloud, or tests of reasoning, such as ﬁnding the right pattern piece to complete a design. In an individually administered test, the examiner asks each question, records the answer, and makes a judg- ment as to the answer’s correctness or quality. Testing stops when the examinee has failed to answer a speci- ﬁed number of questions correctly. When the test is administered to a group, the questions are often in multiple-choice format, and responses are usually recorded by ﬁlling in bubbles on an answer sheet. Answers are compared to a key, so judgment as to correctness is avoided. Studies repeatedly reveal a link between scores on tests of intelligence and educational achievement. Those with higher test scores typically receive higher scores on standardized tests of academic achievement, earn higher grades in school, and complete more years of education. Research also shows a moderate association between intelligence test scores and job status and occupational success. History Alfred Binet is generally given credit for creating the ﬁrst modern intelligence test in 1905. He developed it in France, where he was working in the public schools. In the 1908 version of his test, Binet intro- duced the idea of ‘‘mental level’’ as a way to express the cognitive ability of a child. The mental level of an item was the age at which the average child could solve that particular problem. An item that could be solved by the average child age 7 or above, but not by a child of age 6, was given a mental level of 7 years. Items were grouped by mental level, and test- ing ended at the ﬁrst level where a child could not answer any items correctly. Henry Goddard popularized Binet’s 1908 test in the United States. Several English-language versions of the Binet scale were quickly developed by Goddard and others. In 1916, Louis Terman published an American edition that came to be called the Stanford- Binet and soon replaced all competitors. This test popularized the term intelligence quotient or IQ because scores were expressed as the ratio of mental level or mental age, divided by actual or chronologi- cal age. A child who tested ‘‘at age’’ received an IQ of 1.00. This ratio came to be multiplied by 100 to remove the decimal point, resulting in a scale where the average IQ is 100, the reference point still in use. During World War I, American psychologists under the leadership of Robert Yerkes produced two new group-administered tests for screening Army draftees: a verbal form, Form Alpha, for those who could read English, and a nonverbal form, Form Beta, for the illiterate and those who did not speak English. Form Alpha had eight subtests and Form Beta had seven. Subtest scores were combined to produce a total IQ. After the war, the use of intelligence tests in schools, for college admissions, and in industry spread rapidly. Soon, several million tests, mostly of the group variety, were being administered each year. In the 1930s, David Wechsler applied intelligence testing in the psychiatric clinic of Bellevue Hospital. The Stanford-Binet had been developed for use with children, and Wechsler needed a test for adults. He adapted for individual administration the tests from the Army testing program, grouped the tests into a verbal scale and a performance scale, and collected Intelligence Tests 549 norms for adults. The resulting test battery was called the Wechsler-Bellevue Intelligence Scale. The verbal subtests were combined to produce a verbal IQ, or VIQ, and the nonverbal subtests yielded a performance IQ, or PIQ. The verbal and performance scores were combined to produce an overall description of cogni- tive ability called full-scale IQ or FSIQ. At the same time, Wechsler rejected the IQ as the ratio of mental age to chronological age and introduced the scale that all intelligence tests use today, a scale with a mean of 100 and a standard deviation of 15. Theories of Intelligence Binet had viewed intelligence as a unitary concept that he equated with judgment. In 1904, Charles Spearman had suggested that all cognitive perfor- mances depended on a person’s level of general cog- nitive ability, which he labeled g , and a speciﬁc ability required by that task and no other, called s. Spearman’s ‘‘two-factor theory’’ (of g and s)agreed with Binet’s ideas and his test, but it was rejected and harshly criticized by many American psychologists, particularly Edward L. Thorndike and his students. A major factor in the debate was the nature of the battery of tests that each camp analyzed. Spearman usually had a small sample of subjects, each of whom had taken one test for each aspect of mental ability, much like those included in the Army testing program. Thorndike usually had a large sample of subjects who had taken a much larger set of tests, several for each type of ability. When L. L. T hurstone developed the methods of factor analysis in the 1930s and analyzed a very large battery of tests, he found that there were about eight identiﬁable dimensions of cognitive ability that he called the primary mental abilities. However, he also found that these primary mental abilities were all positively correlated. Spe arman interpreted this ﬁnd- ing that all mental abilities were positively correlated as vindication of his theory. The debate over the structure of cognitive abilities continued through World War II. Thurstone’s factor analytic results led to the development of batteries of teststoassesshisabilityfactorsaswellasothers.The most extreme of the multifactor theories was proposed by J. P. Guilford in the 1950s. Guilford’s Structure of Intellect model eventually postulated 120 relatively independent abilities organized along the dimensions of Contents, Products, and Op erations. Today, the main remaining parts of Guilford’s theory are the concepts of divergent and convergent thinking. Divergent thinking is the ability to see nontraditional solutions to problems and is offered as a major component of crea- tivity, whereas convergent thinking, the ability to extract a single correct solution to a problem, is seen as keytoperformanceonmo st intelligence tests. Cattell-Horn-Carroll Theory and Related Tests In the 1940s, Raymond B. Cattell identiﬁed two broad classes of intellectual functioning, the ability to solve new problems, which he labeled ﬂuid intelligence,or Gf , and the fund of knowledge and information one had acquired from experience, which he called crystal- lized intelligence ,or Gc. Starting in the 1960s, Cattell and his student, John Horn, expanded this theory to include eight or nine broad abilities. A massive factor reanalysis of more than 400 previous studies led John Carroll to offer a similar theory in 1993. Carroll’s three-stratum theory postulated about 70 narrow, spe- ciﬁc abilities that could be grouped into 9 broad abili- ties, which in turn gave rise to a single general ability factor at the highest level. Cattell, Carroll, and Horn agreed that they had arrived at essentially the same place, and their uniﬁed theory is now referred to as the Cattell-Horn-Carroll or CHC theory. CHC theory now forms the theoretical basis for almost all of the com- mercially available intelligence tests. Reaction-Time Theory Beginning about 1980, a new line of research into intelligence began in the work of Earl Hunt and Arthur Jensen. A century earlier, Sir Francis Galton had pro- posed that reaction time could be used to measure intelligence, but the primitive instruments available lacked sufﬁcient accuracy to detect differences. With the advent of computers capable of accurately timing both stimulus presentation and response time, speed of neural processing became a p otent area of research on intelligence. As research using these methods pro- gressed, speed of informati on processing (called the chronometric approach to intelligence )andtheextent andefﬁciencyofworkingmemorywerefoundtocor- relate substantially with sco res on traditional intelli- gence tests. The massive body of work by Jensen has been particularly inﬂuential. He and his colleagues have shown that the aspect of traditional intelligence 550 Intelligence Tests tests that correlates most highly with processing speed/ working memory is the single general ability factor most similar to Spearman’s g. One of the characteristics of processing speed that makes it attractive as a measure of intelligence is its relative freedom from cultural inﬂuences. One of the enduring criticisms of intelligence tests is that they are culturally biased, thereby resulting in lower scores for individuals who are not from the culture producing the test (generally upper-middle-class Whites in the United States). Critics of intelligence testing claim that the tests measure exposure to the majority culture and therefore underpredict the suc- cess of members of minority groups. Various lines of evidence do not support these claims, but the claims continue nonetheless. Because information- processing tasks can be designed to separate reaction time and movement speed from decision speed, rela- tively pure measures of an individual’s ability to process information that are largely independent of experience are possible. The fact that these measures correlate most highly with those traditional intelli- gence measures that are least inﬂuenced by culture, such as Raven’s Progressive Matrices, suggests that information processing and ﬂuid intelligence are similar. Contemporary Intelligence Tests A large number of intelligence tests have been devel- oped over the years to measure various aspects of cognitive ability. Those that are offered for sale are reviewed in the Mental Measurements Yearbooks , published about every 2 years by the Buros Institute for Mental Measurements at the University of Nebraska. Some, such as the Stanford-Binet Fifth Edi- tion and the Woodcock-Johnson Third Edition, are intended for use with the full range of ages from 2 years to 80 +, although the speciﬁc tests used at each age may differ. Others, such as the Wechsler Scales, have different tests for young children (the Wechsler Pre-School and Primary Scale of Intelligence), school-age children (the Wechsler Intelligence Scale for Children), and adults (the Wechsler Adult Intelli- gence Scale). Others, such as the Kaufman Ability Battery for Children, are only for a speciﬁc and lim- ited age range, but almost all of them make an effort to measure some or all of the group factors in the CHC model. They all also offer a single index of gen- eral cognitive ability. Educational Decision Making and Intelligence Tests In educational settings, intelligence testing is used primarily to gain information to support a variety of placement decisions. School personnel need informa- tion to help predict how much an individual will learn from a particular educational program or how suc- cessful he or she might be in each of several alterna- tive programs. For example, a high school may have to decide whether a freshman should be in the advanced placement section in mathematics or in the regular section. In addition, schools regularly must determine whether certain students meet eligibility critieria for special education services and, if they do, which services will best serve the needs of those stu- dents. In each of these cases, information helps the person making the classiﬁcation or placement deci- sion to identify the group to which an individual most likely or properly belongs. A great deal of controversy exists surrounding placement decisions in educational settings. However, because of the existence of individual differences in achievement among students and the tendency for these differences to become larger, not smaller, as stu- dents progress through school, it is likely that such decisions will always be necessary. Students differ markedly in how much they have learned from previ- ous instruction, their motivation to learn, the instruc- tional conditions under which they learn best, their rate of learning, and their ability to apply previous experience to new learning situations. Furthermore, an instructional program that facilitates learning and growth for one student may hinder it for others. Class- room teachers and other educators have long recog- nized the need to adapt teaching methods, learning materials, and curricula to meet individual differences among students and to place students in the kind of learning environment that will optimize their educa- tional opportunities. Some students have instructional needs that differ so markedly from those of other students in the class that the teacher is unable to make the necessary adjustments within the classroom. Typically, these students are of two kinds. At one extreme are gifted students whose level of achievement and speed of learning greatly exceed those of other students in the class, and at the other are students who are having great difﬁculty mastering basic educational skills such as reading, are unable to adapt to classroom demands, Intelligence Tests 551 or are so disruptive that they interfere with the learn- ing of other students in the classroom. A teacher may decide to refer these kinds of students to a school psy- chologist or counselor for assessment and possible alternative placement. Assessment of Children With Learning Difficulties Assessment of the construct of intelligence is central to the diagnosis of a number of prominent learning difﬁculties. Whenever a student is identiﬁed as a can- didate for special education services, federal law requires that documentation, in the form of test scores, is collected and used as part of eligibility and placement decisions regarding that student. The type of test evidence needed to support different classiﬁca- tions varies by state and by classiﬁcation, but a standardized test of intelligence is commonly among those administered. This is because intelligence tests are the most powerful tools employed in such evalua- tions. Because such tests are used diagnostically, a wide range of information can be gathered, ranging from the recall of general information to more pro- cesses. Although such tests are used to determine eli- gibility for special education, they are also very useful in the identiﬁcation of the child’s cognitive strengths and weaknesses—all valuable data when it comes time to provide a structured plan for the child’s future learning-focused activities. The classiﬁcation of mental retardation almost always involves administration of an individual intel- ligence test such as the Stanford-Binet or the age- appropriate Wechsler Scale. Because limited intelli- gence is considered a principal feature of mental retardation, comparison of a student’s test score to national norms of children of the same age is a neces- sary step in identifying this disability. An intelligence test score in the 70 to 75 range is one indicator of mental retardation but is not, in and of itself, enough evidence for a diagnosis. The individual must also show substantial deﬁcits in adaptive and social func- tioning to be classiﬁed mentally retarded. Below- average intellectual functioning and an inability to adapt to the demands of normal life constitute mental retardation. A diagnosis of a learning disability requires even more extensive assessments. Because learning disabil- ities can affect a child’s (or an adult’s) use of spoken or written language, performance of simple or complex mathematical calculations, and direct attention, there has to be a clear distinction between mental retarda- tion and a learning disability. Intelligence tests are generally administered to ensure that a particular stu- dent’s intellectual functioning is within the range of normal, and that the student is not delayed or men- tally retarded. Standardized achievement can also be administered to help provide additional information regarding the teacher’s in-classroom observations about a particular area of skill in an effort to more clearly identify speciﬁc deﬁcits. Such observations can help better identify a discrepancy between intel- lectual capacity, as measured by the intelligence test, and level of academic performance. The discrepancy is the key to the identiﬁcation of a learning disability. Intelligence tests can be differentiated from achieve- ment tests chieﬂy in the use to which the scores are put. Achievement tests measure the fund of knowledge one has acquired in a speciﬁc academic or occupational domain, such as language arts or mathematics. Achieve- ment tests are postinstruction measures, and interpreta- tion of the scores should nor mally be restricted to the domain and instructional experiences of interest. Together with intelligence tests, achievement tests help complete the picture of an individual’s speciﬁc learning weaknesses. For children with speech and language disabilities— one of the largest groups of children receiving special education services today—intellectual assessment is less important than for the other high-frequency disability categories, learning disabil ities and mental retardation. The main reason is that it is entirely possible for a child to have difﬁculty in expressive ability, either oral or written, without affecting o ther, higher-order proces- sing skills. However, general tests of intellectual ability can provide useful information, especially for children who are mentally retarded or have a speciﬁc learning disability. Such tools can identify the role of psycho- logical processes and allow the conclusion that multi- ple information-processing pathways are involved. For children with disabilities that are low in fre- quency (e.g., visual or aural), assessment of intellectual abilities presents a challenge. This is because most tests are not suited, and not designed, for children with such disabilities. Consequentl y, testing must be conducted using modiﬁed formats with consideration of the lim- itations that the disability presents. The most signiﬁcant concern when this is the case is that such accommoda- tions are not used during the standardization process 552 Intelligence Tests and are often generated impromptu by the examiner. Not all learning problems are the result of disabilities. Some children are simply slo wer at developing certain skills, and learning difﬁculties may be just the result of maturational delays. Use of standardized tests of intelli- gence and academic achievement provide useful evi- dence to help tease these differences apart so that children receive appropriate educational experiences. Criticism of Intelligence Testing in Education It is often asserted that sta ndardized tests of all types are used to deprive certain groups of access to educa- tional opportunities. When tests are used in a mechani- cal fashion as in selection and placement activities, and due to the recognized fact that some groups in our society have historically performed less well on tests, standardized tests can unfa irly bar members of under- represented groups from educational opportunities. The term for this idea is bias in testing.Whether a test score or its use is judged to be biased depends, to a large degree, on the deﬁnition of bias being used. Given this concern, the question is whether individuals are indeed unfairly barred, and if this is the case, what then consti- tutes fair and equitable use of tests for the selection, placement, and classiﬁcation of individuals? There have been a number of well-documented court cases that have offered highly publicized deci- sions one way or the other on the issue of test bias. One such case speciﬁcally targeted intelligence tests. LarryP.etal.v.WilsonRilesetal. (1979) is the case in which the validity (an essential psychometric com- ponent of any test) of intelligence tests was called into question for Black children, in particular. The presiding judge ruled that standardized , individually administered tests of intelligence were biased against children who were culturally different from those on whom the test was normed. In a similar court case in Illinois, Parents in Action on Special Education v. Hannon (PASE v. Hannon , 1980), a different deci sion, with different implications, was reached: Standardized measures of intelligence taken in the context of a broader assess- ment system were found not to be biased against chil- dren from culturally different backgrounds. The use and interpretation of test scores within groups whose cultures and experiences differ from the general population for which a test was designed have received a great deal of attention during the past 25 years. Although there are many subgroups in American society, ethnic and linguistic minorities are the most clear-cut of these and the ones for whom the appropri- ateness of tests and questionnaires is most open to question. Some critics contend that testing students from linguistic minorities in other than their preferred language produces inaccurately low estimates of true intellectual capacity. In Guadalupe Organization v. Tempe Elementary School District (1972), the court established that when the child’s primary language is not English, the primary language the child does speak must be determined prior to assessment and that formal provisions to ensure accurate assessment must be made. Examples of provisions that can be made and were mentioned in this case include the administration by examiners who are ﬂuent in English as well as the child’s primary language, the presence of an interpreter to assist when necessary, the use of instruments that do not stress spoken language, and an assurance that the test results will be explained to parents or guardians in their primary language. The number of Americans for whom English is not the preferred language has increased such that many ability and achievement tests are available in a Spanish translation. Major test pub- lishers now go to considerab le lengths to ensure that their test items do not present an unfair challenge for students from diffe rent backgrounds. Tracy Thorndike Christ and Robert M. Thorndike See also Intelligence and Intellectual Development; Intelligence Quotient (IQ); Mental Age; Stanford–Binet Test; Working Memory Further Readings Flanagan, D. P., Genshaft, J. L., & Harrison, P. L. (1997). Contemporary intellectual assessment: Theories, tests, and issues. New York: Guilford. Jensen, A. R. (1998). The g factor: The science of mental ability. Westport, CT: Praeger. Thorndike, R. M. (1990). A century of ability testing. Chicago: Riverside. Thorndike, R. M. (2005). Measurement and evaluation in psychology and education (7th ed.). Upper Saddle River, NJ: Prentice Hall. Web Sites Buros Institute for Mental Measurements: http://www.unl.edu/buros/bimm Harcourt Assessment (also PsychCorp): http://www.harcourtassessment.com Intelligence Tests 553 National Association of School Psychologists: http://www.nasponline.org Pearson’s Assessment Group: http://www.pearsonassessments.com Riverside Publishing: http://www.riverpub.com I NTERNAL V ALIDITY In experimental psychology research, it is important to conﬁrm that results of a study are actually due to independent or manipulated variables, rather than extraneous variables. Internal validity is the degree to which the results are associated with the independent variable, and not other, uncontrolled factors. When an experiment is said to be internally valid, a direct causal relationship between the i ndependent and dependent variablesisdemonstratedunequivocally.Thisisincon- trast to external validity, which is the generalizabilty of results across different expe rimental settings. This entry outlines several types of extraneous variables that may jeopardize the internal validity of research in educa- tional psychology. Threats to Internal Validity If an experiment extends over a long period of time, uncontrolled changes may occur in the experimental groups—this confound is known as the history effect. If a difference is observed between the experimental and control groups, it may be due to history effects or the interaction of history with treatment. For example, suppose an educational psychologist were to investi- gate the effectiveness of a reading program across the school year delivered to two groups (experimental and control) by specially trained teachers. Suppose further that for 2 months, a substitute teacher replaced the original teacher in one group—this may pose an uncontrolled effect, as the substitute teacher may have inﬂuenced students’ reading performance. As with history, maturation is also a concern when a study extends over a long period of time. As experi- ments go on, biological or psychological changes may occur in participants. For example, participants become older, and they may change emotionally or physically. For example, suppose an educational psy- chologist were to examine the effectiveness of a read- ing program over one school year. If factors such as physical, social, and intellectual development are not controlled for, these processes, rather than the reading program, may contribute partly or entirely to reading proﬁciency changes. Therefore, the investigator could not claim unequivocally that the reading program is beneﬁcial; this study would lack internal validity. Another threat to internal validity is testing. Many educational experiments use pretest and posttest designs; and often, both testing situations are similar. If so, then the participants may display a practice effect from repeated testing. If participants show an improvement on a particular test, it may be due to the repeat test, rather than the independent variable (such as the reading program). Additionally, subjects might be sensitized to certain aspects of the pretest on which they focus more, and, as a result, do better on the posttest. On the other hand, if different tests are used in pretest-posttest situations, this may cause differ- ences in measurement that may bias the results. For example, if a pretest consists of a more difﬁcult read- ing test and the posttest is easier, the improvement may not be due to the independent variable, the read- ing program, but to the differences in measurement. This is called an instrumentation effect. Attrition occurs when there is loss of participants during the course of an experimental treatment. Par- ticipant dropout can bias the results because there may be important differences between the participants who discontinue in the study and the participants who remain in the study. That is, if certain types of partici- pants are more likely to drop out of the study than others, then the group that concludes in the study con- sists of different types of individuals than the group that started. This can be detrimental to the results of experiments. If a study is measuring improvement to reading after the implementation of a reading pro- gram, and the students who drop out of the study are low-achieving students, then the results would be biased; the concluding sample would consist of high- achieving students and low-achieving students would not be represented. Another threat to the internal validity of an experi- ment is regression toward the mean. When individ- uals score unusually low or high on pretests, posttest scores are likely to be less extreme and closer to the mean; over repeated testing, scores have a tendency to regress toward the mean (i.e., unusual scores are unlikely to occur for the same individual on both test- ing situations). If the scores on the posttest are differ- ent from those on the pretest, it may be due either to the independent variable or to regression effects. 554 Internal Validity When experimental studie s use control and experi- mental groups, ideally the groups are equal on all things except the independent variable. However, if the groups are not equal, then subject selection effects may occur. As an example, suppose that an educational psychologist is interested in studying the effects of instructional technology in a n introductory educational psychology course. The technology is implemented in the early morning class and not in the evening class. If the group that had used instructional technology per- forms better on an exam, the technology may not be the only factor that inﬂuenced this outcome. The two groups might be different in some aspects. Why would some students choose a morning section, whereas others choose an evening sect ion? These potential dif- ferences in groups may contribute to the outcome of an experiment. Subject selec tion effects may interact with other threats to internal validity. For example, one group might mature differently and thus complicate the results; researchers would not be sure if the results of an experiment are due to the independent variable or to maturation effects. This th reat to internal validity is referred to as selection-matura tion interaction. Selec- tion effects may interact wi th other threats to internal validity (e.g., a group might be affected by history or testing in different ways). The ﬁnal threat to internal validity is that of experimental or expectancy bias. When experimenters administrate tests and are in direct contact with the participants, they may inadvertently inﬂuence the results. If there is a change in the dependent variable in one group, it may be due either to the independent variable or to the inﬂuence the experimenter may have had. For example, individuals conducting the experiment may display unintended, differential treat- ment between the experimental and control groups. This may confound results. Enhancing Internal Validity Researchers who are aware of the potential problems that may arise in experimental research can create experiments that avoid or reduce the above-mentioned threats to internal validit y. Randomized experiments are highly recommended as they can eliminate most of these threats (e.g., there is no subject selection bias if subjects are selected random ly, history effects will be reduced). Unfortunately, much of educational research compares treatments given to intact classrooms. Random assignment at the subject level is difﬁcult, as students are members of exi sting classrooms. In other words, classrooms cannot be split up easily, with only some children receiving various treatments. It is also important to note that the presence of a threat to internal validity does not necessarily mean that the internal validity of an experiment is weak- ened. For example, maturation may occur in students between pre- and post-readings tests, yet not enough to invalidate the results of an experiment. It is almost impossible to design the perfect experi- ment that completely safeguards all threats to internal validity. Rather, the internal validity of an experiment varies along a continuum. An awareness of these poten- tial threats and ways of enhancing internal validity may protect an experiment from some of these problems. Jennie K. Gill See also Experimental Design; External Validity; Inferential Statistics; Quantitative Research Methods; Statistical Signiﬁcance; Testing Further Readings Campbell, D. T. (1969). Reforms as experiments. American Psychologist , 24 , 409–429. Campbell, D. T., & Stanley, J. C. (1963). Experimental and quasi-experimental designs for research on teaching. In N. L. Gage (Ed.), Handbook of research on teaching (pp. 171–246). Chicago: Rand McNally. Cook, T. D., & Campbell, D. T. (1979). Quasi- experimentation: Design and analysis for ﬁeld settings. Boston: Houghton Mifﬂin. I NTRINSIC V ERSUS E XTRINSIC M OTIVATION The concepts of intrinsic and extrinsic motivation are used in the ﬁelds of education and social psychology. In everyday language, intri nsic motivation is simply another way of saying that people are interested in and enjoy what they are doing. According to the social psy- chological literature, people are said to be intrinsically motivated when they do an activity for its own sake, not for any extrinsic reward. Extrinsically motivated behaviors are those in which an external controlling factor can be readily identiﬁed. If people solve puzzles, play games, or paint pictures for no obvious external reason, they are said to be in trinsically motivated. On Intrinsic Versus Extrinsic Motivation 555 the other hand, students who study hard to obtain high grades, employees who work extra hours for pay, and children who do their homework to please their parents are said to be extrinsically motivated. This entry addresses the history of these concepts, theoretical per- spectives, research relate d to intrinsic and extrinsic motivation, and the practical implications of the research ﬁndings. Historical Overview One of the earliest uses of the term intrinsic motivation came from primate research conducted by Harry Har- low and his colleagues in the 1950s. Intrinsic motiva- tion was used to refer to the fact that monkeys would solve puzzles when solution to those puzzles did not result in extrinsic reward. Essentially, Harlow found that the animals performed ef ﬁciently ‘‘without resort to special or extrinsic incen tives’’; solving the puzzles seemed to be its own reward. In the 1950s, psychologi- cal accounts of motivation were dominated by Clark Hull’s drive reduction and Sigmund Freud’s psychoan- alytic theories; both of these accounts emphasized pri- mary drives (e.g., thirst, hunger, and sex). These theories could not easily explain Harlow’s ﬁndings; the animals in Harlow’s research were not motivated by primary drives. A review by Robert White in 1959 revealed additional evidence for non-drive-based motives such as exploration, curiosity, spontaneous activity, and manipulation of objects—motives that could account for ﬁndings such as those of Harlow. These new motives, subsumed under the term effec- tance motivation , form part of a process whereby an animal or person learns to deal effectively with the environment. Today, the term intrinsic motivation is used by many psychologists to refer to non-drive-based motivation. One unresolved issue concerns the distinction between intrinsic and extrinsic motivation. The term intrinsic motivation is deﬁned by the absence of obvious external factors such as extrinsic rewards. Although many human behaviors appear to occur in the absence of any obvious or apparent extrinsic con- sequence, they may, in fact, be a result of past conse- quences, or they may be due to anticipated future beneﬁts. From this perspective, intrinsically motivated behavior is simply behavior for which appropriate controlling stimuli have yet to be speciﬁed. In other words, when we do not know why a person engages in a particular activity, we infer intrinsic motivation. Thus, the motives for engaging in many activities get categorized as intrinsic motivation. As a result, behav- ior due to distant, hidden, or obscure external causes gets mistakenly labeled as intrinsically motivated. In spite of these conceptual difﬁculties, the terms intrin- sic motivation and extrinsic motivation are frequently used in education and psychology. Many theorists have been interested in the relation- ship between intrinsic and extrinsic motivation. Some theorists in organizational and industrial psychology assumed that intrinsic and e xtrinsic rewards were addi- tive, combining to increase overall performance. In terms of work, the view was that the highest motiva- tiontoperformwouldoccurwh en jobs were interesting and challenging and when employees received extrin- sic rewards for performance (e.g., pay, recognition, promotion). That is, the belief was that productivity and satisfaction would be highest when intrinsic moti- vation was supplemented with extrinsic incentives. In the 1960s, the psychologist Richard DeCharms challenged this view and suggested that intrinsic and extrinsic motivation may not add together, and that external rewards might actually interfere with or sub- tract from intrinsic motivation. DeCharms speculated that external rewards would change people’s percep- tions about the causes of their behavior. If people were rewarded for engaging in activities, they would begin to see themselves as doing the activity for the reward rather than for interest and enjoyment. In this way, DeCharms suggested, external rewards would undermine intrinsic motivation. Research on Extrinsic Rewards and Intrinsic Motivation The hypothesis that extrinsic rewards could disrupt an individual’s intrinsic motivation has been highly inﬂu- ential in social psychology and education, leading many researchers to investigate the relationship between external rewards and intrinsic motivation. Two early experiments on the topic, one conducted by Edward Deci, the other by Mark Lepper and his associates, are frequently cited as example s of how extrinsic rewards negatively affect people’s intrinsic motivation. Because incentives and reward systems are often used in busi- ness and education, the early ﬁndings on negative effects led to a splurge of research. Since the 1970s, dozens of experiments, using a common set of proce- dures, have been conducted to investigate the undermin- ing effects of reward. 556 Intrinsic Versus Extrinsic Motivation In a typical experiment, people are presented with an interesting task (e.g., solving puzzles, drawing pic- tures,orplayingwordgames)forwhichtheyreceive praise, money, candy, gold stars, and so forth. A con- trol group performs the act ivity without receiving a reward. Both groups are then observed during a non- reward period in which they are free to continue per- forming the task or to engage in some alternative activity (free-choice period) . The time that participants spend on the target activity during this period (free time), their performance on the task during the free- choice session, and/or their expressed interest in the activity are used as measures of intrinsic motivation. If participants in the reward condition spend less free time on the activity, perform at a lower level, or report less task interest than those in the nonreward group, reward is said to undermin e intrinsic motivation. Over the past few years, there have been several sta- tistical reviews of experiments on rewards and intrinsic motivation. The most recent meta-analytic review by Canadian researcher Judy Ca meron and her associates examined the results from 150 experiments. They found that extrinsic rewards do not always have nega- tive effects on people’s intrinsic interest; in fact, rewards were found to produce positive effects under several conditions. In general, on high-interest tasks, rewards were found to undermine intrinsic motivation when they were tangible, expected (promised before- hand), and loosely tied to performance. In contrast, tan- gible rewards increase intri nsic motivation when they are contingent upon achieving a speciﬁc level or stan- dard of performance. Verbal praise and positive feed- back also lead to positive effects on measures of intrinsic motivation. As well, all types of rewards have been found to increase intrinsic motivation on tasks of low initial interest. Recen t research has shown that intrinsic motivation increases when rewards are offered and given for successfully achieving a graded level of performance leading to mastery. As well, when individ- uals are rewarded for achievement during learning and on tests, people show even higher levels of intrinsic motivation. Theories of Extrinsic and Intrinsic Motivation One theoretical account of the relationship between extrinsic and intrinsic motivation comes from Edward Deci and Richard Ryan’s cognitive evaluation the- ory (CET). CET proposes that intrinsic motivation springs from two innate sources: the need for self- determination and the need for competence. Intrinsic motivation arises from activities that lead people to fulﬁll these basic needs. When people engage in activities such as puzzle solving, they are said to feel free and highly competent; these perceptions in turn activate the innate energy of intrinsic motivation. The energy of intrinsic motivation heightens interest and sustains involvement in an activity. Extrinsically motivated behaviors, on the other hand, refer to actions that are directly linked to an external cause such as an explicit reward, incentive, or threat. Extrin- sically motivated actions are said to be characterized by pressure and tension and to result in a loss of per- ceived competence and personal freedom. These assumptions provide the theoretical basis for CET’s account of the effects of extrinsic rewards on people’s intrinsic motivation. CET has typically focused on negative effects of rewards and their controlling versus informational aspects. From the CET perspective, all rewards are experienced as controlling, but rewards that are closely tied to achievement of performance standards are said to be the most controlling. Offering rewards for meet- ing performance standards leads individuals to feel pressured; such rewards undermine perceptions of self- determination leading to a reduction in intrinsic moti- vation. On the other hand, rewards linked to achieve- ment can also provide information about competence. When people succeed at atta ining a performance stan- dard, the rewards convey information about perfor- mance that is positively evaluated by the individual; this evaluation leads to greater perceived competence and may offset some of the controlling aspects of rewards. For CET, the critical process for intrinsic motivation involves people’ s evaluation of rewards in terms of perceived self-determination, and the second- ary process concerns perceived competence. Another explanation of the effects of rewards on intrinsic motivation comes from attribution theory and Mark Lepper’s overjustiﬁcation hypothesis. As with CET, the focus has been on the negative effects of rewards; rewards are said to decrease intrinsic motiva- tion by altering people’s attributions of causation for their behavior. When rewards are offered for perform- ing an activity, people are said to discount the internal causes of their actions (e.g., interest, effort, skill, abil- ity) and to view their behavior as externally moti- vated. Heightened attributions of control by external factors lead to low levels of personal autonomy, Intrinsic Versus Extrinsic Motivation 557 resulting in a loss of intrinsic motivation. At the same time, however, rewards for achievement may convey information about one’s ability, mastery, or compe- tence. Perceptions of high competence enhance intrin- sic motivation. This is most likely when rewards are given for successful performance on an activity. Thus, rewards given for achievement may increase percep- tions of competence as well as perceptions of external control; the effects of such rewards on intrinsic motivation will depend on the relative magnitude of these two competing processes. The overjustiﬁcation hypothesis emphasizes the shift in causal attribution from internal to external sources as the basis for negative effects of rewards on intrinsic motivation. Perceived competence plays a role only if people attribute causation for performance to themselves (internal). Achievement cannot signify competence if people infer that their performance was caused by external factors (rewards or external pressure). In contrast to social psychological theories that depict rewards as harmful to self-determination, com- petence, and internal attribution, Albert Bandura’s social cognitive theory (also called social learning the- ory ) holds that aspects of per sonal agency (e.g., self- motivation and self-directedness) may be enhanced by external inﬂuences such as rewards. That is, social cog- nitive theory rejects CET’s appeal to innate sources of motivation and attribution theory’s shift from internal to external sources as the bases for the effects of rewards on intrinsic motivat ion. Social cognitive theory contends that personal competencies (and other aspects of self-regulation) often are developed with the aid of extrinsic rewards. Social cognitive theory deals with situations in which individuals extrac t personal standards from reward contingencies. The theory states that rewards given for achievement contribute to high personal stan- dards. Attainment of high personal standards leads to task involvement (interest), positive evaluations of per- formance, and increased self-efﬁcacy or competence. In contrast to CET and attribution theory, a high perception of competence or efﬁcacy is the primary basis for intrinsic motivation—not self-determination or shifts in attribution. Another important distinction between social cognitive theory and other approaches is that achievement-based rewards can lead directly to greater involvement and inte rest. Social cognitive the- ory proposes that rewards for achieving challenging performance standards are likely to act as positive feedback leading to high task interest and involvement, high personal evaluatio ns of performance, and increased competence; these processes in turn increase intrinsic motivation. Social cognitive theory distinguishes between non- competency-contingent and competency-contingent rewards. Non-competency-contingent rewards include rewards given without regard to mastery of perfor- mance (e.g., rewards offered for doing an activity with no criterion for performance). From a social cognitive perspective, non-competency-contingent rewards impart little indications of competency in that the rewards are loosely tied to behavior. This means that intrinsic motivation decreases when people are offered rewards simply for doing an activity without regard to level of performance, and meta-analytic ﬁnd- ings support this view. Bandura describes rewards given for mastery (i.e., achieving relatively challeng- ing behavioral standards) as competency-contingent rewards. Competency-contingent rewards are said to enhance intrinsic motivation by increasing people’s perceptions of self-efﬁcacy and competence. Cameron and colleagues found that intrinsic motivation increased when the rewards were tied to achieving mas- tery or a speciﬁc level of performance, a ﬁnding that supports social cognitive theory. Practical Implications Reviews of the research indicate that extrinsic rewards can have negative, neutral, or positive effects on people’s intrinsic motivation. On tasks of low initial interest, extrinsic rewards can be used to increase motivation and performance. This ﬁnding indicates that rewards can be arranged to enhance time and performance on activities that initially hold little enjoyment. In education, a major goal is to instill motivation and enjoyment of academic activities. Many academic activities are not of high initial inter- est to students. An implication of this ﬁnding for edu- cation is that extrinsic rewards may be used by teachers to increase students’ motivation and perfor- mance on low-interest academic activities. On high-interest tasks, verbal praise and tangible rewards linked to success or to obtaining or exceeding a speciﬁc performance standard can enhance people’s interest. These reward contingencies can be viewed as a subset of the many possible arrangements of the use of reward in everyday life. Extrinsic rewards can be arranged to progressively shape performance, to 558 Intrinsic Versus Extrinsic Motivation cultivate initial interest in an activity, and to maintain or enhance effort and persistence at a task. A negative effect occurs when an activity is of high initial interest, when the rewards are tangible and offered beforehand, and when the rewards are delivered without regard to success on the task or to any speciﬁed level of performance. How relevant to everyday life is the detrimental effect of reward, found in the experimental studies? In work settings, there are some circumstances in which individuals are rewarded irrespective of success or performance level. For example, because of compensation and promotion systems that are insensitive to performance (e.g., wage by job classiﬁcation), some employees can vary their performance substantially with little effect on tangible reward. In such a situation, employees may have considerable latitude in how well or poorly they perform t

bundet

heir jobs without any change in pay or fringe beneﬁts. If an employee has high intrinsic interest in the job and receives a reward independent of perfor- mance, interest in the job may decrease, thereby pro- ducing poorer performance than if the reward were contingent on performance. In educational settings, if a student has high inter- est in a particular subject matter, say, a history course, and receives the same grade regardless of perfor- mance, interest in history could deteriorate. The stu- dent may be less likely to spend time reading history during the course and following the conclusion of the course. Negative effects on intrinsic motivation also occur when rewards are promised but not delivered, offered in an authoritarian manner, or arranged for the beneﬁt of the authority rather than the recipient. As well, rewards must be tailored to the individual; what is rewarding for one person may not be for another. When rewards are used inappropriately, people can actually perceive them as punishers rather than incen- tives, a point documented by the writer Alﬁe Kohn. In order to avoid poor performance and reduced task interest in business and educational settings, employers, teachers, and administrators need to con- sider the basis upon which they allocate rewards, rec- ognition, and advancement. Careful arrangement of rewards in educational settings and in the work envi- ronment can enhance students’ and employees’ interest and performance. This occurs when extrinsic rewards are closely tied to the attainment of performance stan- dards. In addition, when tasks are challenging and graded in steps toward mastery of skills, extrinsic rewards given at each step of accomplishment can instill interest and high personal standards for perfor- mance. As well, according to Robert Eisenberger’s learned industriousness theory , rewards that are made contingent on effort lead people to work harder at an assigned task and show increased performance on other activities. Extrinsic rewards tied to high effort also result in higher intrinsic motivation than rewards given for low effort. Overall, rewards based on chal- lenge, mastery, and high effort build skills, personal standards, persistence, and intrinsic motivation. Current Directions Research on extrinsic and intrinsic motivation con- tinues to be important in psychology and education. One line of inquiry concerns the relationship between sources of motivation and cr eativity. Eisenberger and his colleagues have been working on the effects of giving rewards for novel performances. Creativity involves the generation of novel behavior that meets a standard or quality. Rewards for working hard at gen- erating novel responses ha ve been found to increase creative behavior. When rewards are given for creative thinking and performance on speciﬁc tasks, people show generalized creativi ty on other tasks. Rewards can also enhance creativity through increased intrinsic motivation. That is, rewards for high levels of creative performance have been found to increase perceived self-determination and competence, leading people to greater task enjoyment. In contrast, tying rewards to conventional-level perfor mance (noncreative) causes people to be less creative and unmotivated. Another line of research has focused on building intrinsic motivation without using extrinsic rewards. Deci and Ryan have developed self-determination theory as a more general approach to intrinsic moti- vation than CET. Self-determination theory focuses on the social conditions that activate or interfere with innate self-motivation. The theory postulates three psychological needs—competence, autonomy, and relatednes—that, when satisﬁed, lead to enhanced self-motivation. CET pointed to the importance of perceptions of autonomy and competence for intrin- sic motivation; a third factor, relatedness, is also seen to enhance self-motivation. Relatedness concerns the social interaction dynamics between students and teachers, employers and employees, and parents and children. Secure relationships lead people to feel accepted, whereas negative interactions lead individ- uals to feel that others are cold and uncaring. Based Intrinsic Versus Extrinsic Motivation 559 on relatedness, intrinsic motivation is enhanced or thwarted. Self-determination theory is being used in applied settings such as education, health care, par- enting, work organizations, sports, and mental health. Self-determination theory emphasizes the impor- tance of social contexts for intrinsic motivation. Cur- rent research is showing how social contexts can modify the effects of extrinsic rewards on intrinsic motivation. Speciﬁcally, giving rewards for activities within social contexts that emphasize choice, self- monitoring of performance, and self-pacing have been found to increase people’s intrinsic motivation. As well, rewards in these contexts lead people to perceive themselves as self-determined and competent. This research points to new approaches for arranging autonomy-supportive social contexts in classrooms and the workplace that promote enjoyment of activi- ties using competency-contingent rewards. Another direction of resea rch concerns testing the major theories that attempt to account for how extrinsic rewards affect intrinsic motivation. As noted, CET, attribution theory, and s ocial cognitive theory each emphasize different cogniti ve processes that mediate the effects of rewards on intrinsic motivation. CET emphasizes people’s percepti ons of self-determination and feelings of autonomy, att ribution theory focuses on the shift in causal attributio n from internal to external sources, and social cognitive theory postulates that per- ceptions of competence or self-efﬁcacy are the major underlying processes. Re cent research by Cameron and associates using path analysis has indicated that perceived competence is the central cognitive factor. When rewards for performance make people feel com- petent, intrinsic motivation is increased. The study also showed that rewards can activate processes that involve both internal and external sources of motivation. Exter- nal and internal motivations add together to affect overall motivation for an activity, as suggested by early investigators in organizational and industrial psychol- ogy. Further research on cognitive processes that medi- ate reward effects on these two sources of motivation is warranted. Judy Cameron and W. David Pierce See also Cognitive View of Learning; Motivation; Self-Determination; Self-Efﬁcacy; Social Learning Theory Further Readings Bandura, A. (1986). Social foundations of thought & action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall. Cameron, J., & Pierce, W. D. (2002). Rewards and intrinsic motivation: Resolving the controversy. Westport, CT: Greenwood Press. Eisenberger, R., & Cameron, J. (1996). The detrimental effects of reward: Myth or reality? American Psychologist , 51 , 1153–1166. Kohn, A. (1999). Punished by rewards: The trouble with gold stars, incentive plans, A’s, praise, and other bribes. Boston: Houghton Mifﬂin. Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist , 55 , 68–78. Sansone, C., & Harackiewicz, J. M. (Eds.). (2000). Intrinsic and extrinsic motivation: The search for optimal motivation and performance. San Diego, CA: Academic Press. 560 Intrinsic Versus Extrinsic Motivation K The most important human endeavor is the striving for morality in our actions. Our inner balance and even our very existence depend on it. Only morality in our actions can give beauty and dignity to life. —Charles Dickens, Dombey and Son , 1847 K OHLBERG ’ S S TAGES OF M ORAL D EVELOPMENT The word moral as it applies to human conduct and thinking has several commonly understood dictionary meanings. These include to act with virtue, to act hon- estly and with right character, and to act justly. The word moral is also deﬁned as a generally accepted standard or custom for right living in a society. All of these meanings are embedded within Lawrence Kohlberg’s stage theory of moral development. Historically, the study of m orality has fallen within the purview of theology and philosophy. Because of the inﬂuence of the positivist tradition in science, the study of moral development as a plausible subject of scientiﬁc inquiry did not fully emerge until the latter half of the 20th century. This breakthrough was signiﬁ- cantly inﬂuenced by the challenges that Thomas Kuhn made concerning the assumptions of the positivist tra- dition. During this same period, cognitive theory gained increasing inﬂuence over behavioral theory toward its efforts in explaining reasoning development in children and adolescents . Cognitive developmental theory, through the pioneering works of Jean Piaget and Lawrence Kohlberg, ha s become pivotal in creat- ing the study of moral reasoning development as a legit- imate research area in educational psychology. Cognitive Developmental Theory: Piaget’s Work Although Piaget is most noted for his theory of cogni- tive stage development, in 1932, he published a highly inﬂuential book titled The Moral Judgment of the Child. Piaget asserted that a child’s moral reasoning develops not merely as a function of internalizing the norms and values of an individual culture, but rather as a natural process of constructing ideas of justice and fairness largely through peer interaction. Through his interviews with children, Piaget identiﬁed three phases of moral reasoning development. He consid- ered these phases, which sequentially included the nonmoral, the heteronomous, and the autonomous, to describe only developmental trends. He noted that children move from an external morality in which jus- tice is judged by concrete events toward a morality judged by pragmatic acts of reciprocity and ﬁnally toward a more internalized morality in which just acts are judged by context, intentions, and an idealized sense of reciprocity. Kohlberg’s Theory of Stages Except for some notable differences in methodology, which included the additi on of adolescents to his sub- ject sample and the use of dilemma vignettes in place of story pairs in his interviews, Kohlberg’s 1955 561 dissertation study was ini tially intended to be an empirical extension and validation of Piaget’s work. Instead, his dissertation results became the foundation for Kohlberg’s stage theory of moral development, which, in later studies, d eparted somewhat from Piaget. First, Kohlberg argued that his subject responses should be categorized not merely in terms of overlap- ping phases but rather in terms of six distinct stages. These are further classiﬁed by three levels: (1) the preconventional, which includes the ﬁrst two stages; (2) the conventional, consisting of stages 3 and 4; and (3) the postconventional, which includes stages 5 and 6. Second, in contrast to Piaget’s less assertive idea that moral development proceeds merely in a deﬁnite direction, Kohlberg claimed that his stages are hierar- chical and progress in an invariant sequence, without stage reversal or stage skipping. Finally, whereas Pia- get placed great emphasis on the importance of peer interaction, Kohlberg argued that social interactions between adults and children involving role taking of many kinds are important as well for the moral devel- opment of children and adolescents. Dilemmas and Questions Kohlberg’s method of data collection consisted of presenting his subjects with hypothetical dilemma vignettes and then asking them a series of questions. The subjects’ responses were his data. The following is an abbreviated version of one of his well-known dilemmas. In Europe, a woman was near death from a cancer. There was one drug that doctors believed might save her, but the druggist was charging 10 times what the drug cost him to make. The sick woman’s husband, Heinz, went to everyone he knew to bor- row the required money, but he could only obtain half of what the drug cost. He told the druggist that his wife was dying and asked him to sell it cheaper or let him pay later, but the druggist said no. So Heinz got desperate and broke into the man’s store to steal the drug for his wife. In his interviews, Kohlberg would then ask his sub- jects questions such as the following: Should Heinz have stolen the drug? Which is worse, stealing or letting another person die a preventable death? What would you have done if you were Heinz? Levels and Stages: Kohlberg’s Criteria for Classifying Subject Responses Kohlberg used three criteria for judging and classify- ing subjects’ interview responses by level and stage. A response was judged for what a subject indicated was right or fair behavior, by the reasons given for right conduct, and by the indicated social perspective taken by the subject. At the preconventional level, for stages 1 and 2, subjects’ responses indicate a social perspective involv- ing only the welfare of the self or someone identiﬁed with the self, such as Heinz. Also, right behavior is judged by its concrete pleasure and pain consequences to the self, and the reasons are relatively few in number and simple in content. Stage 1 responses indicate that right behavior involves the avoidance of punishment, so it would be all right for Heinz to steal the drug as long as he did not get caught and punished. Stage 2 responses are more advanced in that right behavior is viewed as not only serving one’s own interests but also recognizingthatothersseektoservetheirowninterests as well. It would be all right for Heinz to steal the drug if he did not get caught and if he cared enough for his wife to take that risk. If Heinz did not care enough for his wife, then he should let her die. At the conventional level, for both stages 3 and 4, the social perspective involves the acceptance of social and societal conventions for judging right and wrong. At stage 3, what is important is the welfare of relationships with important people, such as family members, friends, and schoolmates. Right behavior involves living up to the expectations of, and gaining approval from, those close to the self. Being good means maintaining mutuality in such relationships through trust, loyalty, respect, and care. Reasons for right conduct include that one should care for others and relationships in this sense and that rules that sup- port this notion of good behavior should be followed. Despite the recognition that stealing and dishonesty are wrong, Heinz should still steal the drug because that is the level of caring that a husband should have for his wife. At stage 4, the social perspective becomes more abstract in that the laws of society or the social order structure morality. Good citizenship to the state 562 Kohlberg’s Stages of Moral Development becomes primary as a criterion for right conduct, which should be interpreted in terms of both the letter and the spirit of the laws of state. If Heinz had indeed exhausted all means of legally obtaining the needed money, then Heinz should steal the drug, but the courts should show clemency to him for the special circumstances surrounding his act of stealing. Kohlberg argued that responses in his ﬁrst three stages somewhat paralleled those of the three phases in Piaget’s work but that at stage 4 and beyond, sub- jects used more abstract and sophisticated reasoning for justifying right conduct, which also reﬂects the highest Piagetian level of thinking and reasoning, that is, formal operational thought. At the postconventional level, for both stages 5 and 6, the social perspective becomes more abstract and universal in that it is the wel fare of all humankind that becomes critical for moral judgments. At stage 5, one is aware that universal human rights have priority over social conventions and laws but that the latter are nec- essary to structure morality within a social order. Thus, as in a social contract, one should always abide by conventional laws except in cases in which they conﬂict with universal human rights. However, stage 5 individuals sometimes have d ifﬁculty integrating the conﬂicts between legal an d moral points of view. In such cases, the utilitarian principle of the greatest good for the greatest number should be followed. Stage 6 responses, which are very rare, are based upon self-chosen ethical principles of justice. At this stage, individuals show the highest level of integra- tion of thought, priority of values, and behavior with respect to moral issues. Right conduct is in accor- dance with universal principles, such as the respect for the dignity of all human beings as individuals and the quality of human rights. While recognizing the importance of the property rights of the druggist, Heinz should steal the drug, not merely because he is the husband of a suffering wife but because the pres- ervation of life should have priority over property rights, and breaking the law would have the potential to improve the conventional legal system. Validations, Controversies, and Criticisms Several cross-cultural as well as longitudinal studies support Kohlberg’s claim of developmental progres- sion through the ﬁrst four stages, but development beyond stage 4 is weakly supported at best. From his 1984 longitudinal results, Kohlberg reported that only 13 % reached stage 5 and none of his subjects had clearly reached stage 6. This led to some revisions. Although he defended stage 5 as an empirically sup- ported concept, Kohlberg suspended his empirical claims for stage 6 as an end point for moral maturity. Others have questioned if stage 5 is empirically valid for subjects in non-Western cultures because it is con- ceptualized in the philosophy of Immanuel Kant and other Western philosophers. One of the most inﬂuential challenges to Kohl- berg’s assertion that his stages are hierarchical with respect to gender differences, stems from the work of Carol Gilligan. This is based on the fact that subjects in Kohlberg’s early studies were exclusively male. In later studies, when he began to interview both adult male and female subjects, he found that females pre- dominantly reasoned at stage 3, whereas males more frequently reasoned at stage 4. Noting that females are more likely to be socialized for adult leadership roles that require maintaining relationships through caring, Gilligan has argued that stage 3 should not be considered as less mature than stage 4 but rather only different and representing the impact of the typical differences in gender role socialization. Educational Implications Like John Dewey, Kohlberg believed that moral development and education were inextricably related. Based upon his research in the 1970s and 1980s, he developed an approach to moral education emphasiz- ing the idea of building a just community in the class- room. He believed that moral dilemma discussion was the key process but that it required the following qualiﬁcations. Students and teachers engage in discus- sions that are challenging, and the dilemmas discussed are an integral part of the subject matter. Also, stu- dents and teachers discuss dilemmas from a mixture of different perspectives and stages of reasoning. Thus, despite controversy and criticism, Kohlberg, more than anyone, has had the greatest inﬂuence on the ﬁeld of moral development study in relation to education, educational psychology, and developmen- tal psychology over the past ﬁve decades. Richard E. Hult See also Moral Development; Social Development Kohlberg’s Stages of Moral Development 563 Further Readings Gilligan, C. (1982). In a different voice: Psychological theory and women’s development. Cambridge, MA: Harvard University Press. Kohlberg, L. (1981). Essays on moral development: Vol. 1. The philosophy of moral development. San Francisco: Harper & Row. Kohlberg, L. (1984). Essays on moral development: Vol. 2. The psychology of moral development. San Francisco: Harper & Row. Kohlberg, L., & Kramer, R. (1968). Continuities and discontinuities in childhood and adult moral development. Human Development , 12 , 93–120. Kohlberg, L., & Mayer, R. (1972). Development as the aim of education. Harvard Educational Review , 42 , 449–496. Piaget, J. (1965). The moral judgment of the child (M. Gabain, Trans.). New York: Free Press. (Original work published 1932) 564 Kohlberg’s Stages of Moral Development L I am learning all the time. The tombstone will be my diploma. —Eartha Kitt L ANGUAGE D ISORDERS Language plays a central role in the conduct of human transactions. It is the vehicle by which we form inter- personal relationships. It i s also a vehicle by which we gain access to knowledge and store the information that we’ve learned in memory. Finally, language is a means by which we create new knowledge, including great lit- erary works. Indeed, some h ave argued that language is one of the deﬁning characteristics of the human spe- cies. It comes about as an interaction of our genetic endowment and an environment that nurtures its emer- gence. Typically, language emerges effortlessly, but this is not always the case. When a child struggles to acquire his or her ﬁrst language, with its onset delayed and development protracted in the absence of any other sensory or cognitive deﬁcits and an intact environment, that child is considered to have a language disorder or language impairment. Language is a multifaceted phenomena, with partic- ular form through which unlimited meaning may be expressed and understood in the context of the situation of its use. A disruption may occur in any of these three facets of language or in their critical interactions, resulting in language disorders. Although the cause of such disruptions is typically unknown, the result is staggering because of the high societal value placed on verbal skill. Social interaction, knowledge acquisition, and one’s very quality of life are all jeopardized by a failure in typical languag e development. Understand- ing language disorders is particularly important in the school setting, given its ma ndate for the academic devel- opment of children and the key role schooling plays in their socialization. Typically, the delayed onset of lan- guage is identiﬁed in the preschool years, yet the impact of a language disorder is felt well into the school years and beyond, making it important for early educators, classroom teachers, and other education professionals to be alert to and to understand the debilitating impact that a language disorder has on a child’s life. Basics of the Linguistic System To understand language disorders, the multifaceted nature of the language sys tem itself must be under- stood, because any or all aspects of it may be impaired. The language system is often thought of as arising out of the intersecting components of form, content, and use. The form of language comprises small units that combine to create larger ones, and it is governed by tacitly understood rules for which combinations are permissible and which are not. The smallest unit of an oral language is a sound, or phoneme. Language disor- ders may, but do not always, include an impairment of the sound system. Even if children do not have difﬁ- culty in producing the sounds of their language, they may have difﬁculty segmenting and recognizing those 565 sounds as individual units. Awareness of these sound units forms the foundation of phonological awareness, which in turn is the most reliable predictor of early reading success. Phonemes are combined according to language-speciﬁc rules to c reate a slightly larger unit, the morpheme, the smallest unit that carries meaning in a language. A morpheme may be either what is com- monly considered a word, for example, jump,orwhat is typically thought of as the preﬁxes or sufﬁxes that shade the meaning of a word. For example, when the past tense sufﬁx -ed is added to jump (i.e., jumped), the meaning of this word is shaded to reﬂect not only theactionbutthetimeframeinwhichitoccurred. Children with a language disorder have particular difﬁ- culty acquiring morphemes. Morphemes are then com- bined in larger units to form the grammar or syntax of a language. These units are commonly thought of as simple sentences, or when embedded into one other, a complex sentence. Thus, the form or structure of the language within which we express our ideas to others is completed. The earliest emerging sentences are two- word combinations (e.g., ‘‘want cookie’’), and late onset of this near-universal stage of language acquisition is often one of the ﬁrst signs of a language disorder. The potential of a language’s form as a powerful means of expression is realized once it i ntersects with its content or meaning, for without ideas about the world or internal desires to express, the form of language is empty. The linguist Noam Chomsky illus trated this point in his now famous sentence, ‘‘Colorless green ideas sleep furi- ously,’’ exemplifying the notion that phonemes and mor- phemes may be combined into a sentence that has followed all the rules of a language but expresses noth- ing. The content component of language, or the ideas we hold about ourselves and other people and things in the world, both draw upon and contribute to the child’s con- ceptual development. Reduced or impoverished vocabu- lary development is often part of a language disorder. The third component of the system is use, or prag- matics. The central role that language plays in human affairs comes from its communicative function. Argu- ably, the main purpose of language is for communicating with other people. It is through social interaction with caregivers that the child’s genetic potential for language is guided to emerge. It has been suggested that the use of language in social contexts must also be governed by tacit rules, or successful conversational exchanges could not take place. The cooperative principle suggests that speakers and listeners have tacitly agreed on a shared goal to successfully exchange information in ways that are maximally relevant to one another. Additionally, contextual variations, such as the social status of those communicating, the circumstances, and the purpose of the communication, are all elements of language use and indeed inﬂuence the choice of content and form. For example, a child may use polite forms such as ‘‘Please, give me a piece of cake’’ or indirect requests such as ‘‘Gee, that cake looks good’’ when attempting to get something from a grandparent, but the child may choose a more direct approach when talking to a peer: ‘‘Give me a piece of cake too!’’ A listener may understand a state- ment to be sarcastic when sp oken by a peer but as a repri- mand if uttered by someone in a position of authority (e.g., ‘‘Do you always dress that way when you come to school?!’’). A disrupti on in the use of language in context is most clearly seen in the language disorders experienced by, but not limited to, children with autism spectrum disorders. Defining Language Disorders Based on the previously deﬁned components of the lin- guistic system, language disorders may then be thought of as a disruption in any one or all of them, making it a heterogeneous category of impairments. It may occur solely in the modality of expression, or it may include difﬁculty in understanding or comprehending form, content, or the context-sensitive variations described in the previous section. Language disorders are most often developmental in nature and of unknown origins, but they can also be the result of an acquired brain injury. Language disorders may occur in the absence of any other known condition and coexist with otherwise nor- mal cognitive function, in which case it is identiﬁed as a speciﬁc language impairment (SLI). A language dis- order also may accompany other developmental dis- abilities, such as hearing loss, cognitive impairments, cerebral palsy, autism spect rum disorder, or attention deﬁcit hyperactivity disord er (ADHD). Furthermore, it may accompany the impact o f certain negative envi- ronmental circumstances such as neglect and abuse. Regardless of whether or not there are associated disor- ders, a language disorder places the child at risk for negative social consequenc es because of the primary role that language plays in b oth social interaction and as a vehicle for the acquisition of knowledge. Both spoken and written forms of language may be affected by a language disorder. Indeed, the very nature of the disorder may change over time with changing contextual demands. Speciﬁcally, whereas children 566 Language Disorders may initially be identiﬁed with a spoken language dis- order during the preschool years, with treatment, the difﬁculties in oral communi cation may resolve. Yet, once these same children start school and are faced withthechallengesoflearningtoreadandwrite,they may once again experience di fﬁculty, the characteris- tics of the disorder changing with new demands being placed on the language system. Language disorders are chronic disabilities that show remarkable stability throughout the life span. Prevalence, Stability, and Prognosis Large-scale studies have been conducted in the United States, Great Britain, and Canada of primarily kinder- garten children to establis h the prevalence rate of lan- guage disorders. In studies of kindergarten children from the general population, the prevalence of language impairment without accom panying speech problems ranges between 6 % and 10 %. When language problems are combined with speech sound impairments, the rate rises to approximately 12 %. In high-risk populations, such as children living in poverty, recent studies have shown a much higher prevalence rate than expected in the general population. Reported stability rates are quite variable for children identiﬁed as having a language dis- order before they start sch ool (age 4 or 5). It has been reported that approximately 40 % of these children will go on to have a good outcome, particularly if only expressive language abilities are affected. These moder- ate rates of improvement are, however, somewhat illu- sory when regression to the mean or the impact that false positive identiﬁcati on has on subsequent assess- ments is taken into account. Furthermore, the prognosis is poorer for children with additional comprehension deﬁcits and below-normal nonverbal cognitive abilities. The bottom line is that for children who are correctly identiﬁed as having a language disorder by age 4 or 5, the prognosis for spontaneous recovery is poor. Lan- guage disorders are chronic, and longitudinal studies clearly demonstrate that they persist into adolescence and adulthood. Encouragingly, studies investigating the efﬁcacy of language intervention indicate those who have had treatment generall y fare better than those who have not, yet much more research is needed to clarify the impact of the amount, nature, and timing of treat- ment on long-term prognoses. Given that late onset of single spoken words and pro- tracted periods of vocabulary and grammatical develop- ment are hallmarks of language disorders, it would seem that identiﬁcation of a language disorder during tod- dlerhood would be ideal so that early intervention could be instituted. Two-year-olds with expressive vocabularies of fewer than 10 words are often consid- ered ‘‘late talkers’’ (about 10 % –15 % of toddlers); in recent years these children have been widely studied with the hope of determining which of them would ‘‘catch up’’ and which would not. Overall, the current consensus is that the majority of toddlers with early expressive language delay a re truly ‘‘late bloomers’’ andtheyare not at risk for further language disorders. In contrast, approximately 40 % of late talkers continue to have language difﬁculties. Although they seem to catch up in their vocabulary development, their con- tinuing difﬁculties are reﬂected in difﬁculties with developing sentence structure. It has been suggested thatalongwithhavingsmall er expressive vocabularies, these children also understand fewer words and use fewer communicative gestures, such as pointing and reaching, as a means of expression. Parent report ques- tionnaires are often the tool used to measure a toddler’s emerging language, but unfor tunately, they are not sen- sitive enough to accurately predict persistent language disorder in individual toddlers. A great deal of attention has been directed to those children who apparently have a language disorder in the absence of any other condition, that is, those children who have SLI. Although there indeed may be a sub- group of children who have a language disorder without any other concomitant condition, the speciﬁcity of lan- guage disorders in children is debated widely. Many of these children have nonspeciﬁc language impairment, that is, concomitant nonverbal cognitive limitations or other associated developmental disorders. Studies have indicated that nearly 50 % or more of children identiﬁed as having SLI also have associated developmental motor coordination disorder, ADHD, or other emotional or behavioral problems. Thus, in the educational setting, it would not be uncommon for a child with a language disorder to receive intervention by not only a speech language pathologist but also an occupational therapist, an educational resource specialist, or all three. Causation Any behavior as multifaceted and complex as language development most certainly must be affected by a variety of biological and environmental factors. Thus, it follows that language disorders can rarely be fully accounted for by one direct and isolated causal Language Disorders 567 condition. Researchers are just at the frontier of under- standing the neurobiology of developmental language disorders, but studies of family history patterns suggest familial inheritance as a factor in at least SLI, particu- larly expressive-only forms of the disorder. In the past 15 years, the genetics of language disorders, in particu- lar for SLI, has been of great interest, with some researchers proclaiming that a speciﬁc gene for gram- mar (syntax) has been located. However, there is no conclusive evidence that genes can inﬂuence the com- ponents of language differentially or indeed inﬂuence language differently from o ther cognitive processes. Even though studies with twi ns demonstrate the impor- tance of genetic inﬂuences upon spoken language disor- ders, researchers are quick to note that just as a shared environment makes the large st contribution to individ- ual differences in typical language development, it has substantial inﬂuence in la nguage disorders as well. More research is needed to identify those important contributing environmental v ariables. Certain sensory and cognitive deﬁcits have long been associated with language disorders in childhood, but association does not mean causation. Hearing loss is a case in point. It is commonly assumed that a hearing loss causes a spoken language disorder. Yet it has long been known that two children with exactly the same pattern of loss and residual hearing may have quite different spoken language abilities and disabi lities. It is not possible to predict language ability based upon an audiogram. In what sense then does the hearing loss cause the lan- guage disorder? Similarly, even though language and cognition are intimately interwoven, in cases of cogni- tive impairment, language abilities may or may not be commensurate with the child’s mental age. Indeed, lan- guage abilities in advance of cognition characterize children with William’s syndrome. So again, one must ask in what sense does cognitive impairment cause lan- guage disorder? Although a great deal more research is needed before a satisfactory answer to the question of ‘‘What causes childhood language disorders?’’ is found, it is most likely to reside in the complex interactions of biology and multiple environmental factors. Language Disorders and Academic Performance Language is the medium in which classroom instruc- tion takes place. Teacher-directed lessons, small student work groups, worksheets, and textbooks are all language dependent. Moreover, the nature of the language of the classroom is different from that which the preschool child experiences at home. For example, the allocation of speaking turns and topics are teacher controlled as compared with that within the home where conversation is most often child initiated and directed. The language itself is not as closely tied to the immediate context in the classroom, making it necessary to depend more heavily on the meaning inherent in words and sentences themselves. While in the home, the preschool child usually talks about the here and now, but once he or she enters school, much more of the talk is distanced in time and place. Finally, language itself not only becomes the medium by which curriculum is accessed through reading and writing but also becomes a subject of study. While the typically developing child navigates these changes inhowlanguageisusedwhenenteringschool,they present signiﬁcant challenges for the child with a lan- guage disorder. Difﬁculty following multiple instruc- tions, following classroom ‘‘rules’’ like raising a hand before speaking, or staying on topic can be misinter- preted as behavior problems if educators are not aware of the impact of a language disorder on the functional use of language in the classroom. Language Disorders and Reading The contributions of oral language abilities to devel- oping literacy skills have been increasingly recognized in recent years. Indeed one of the most consistent ﬁnd- ings in the reading literature is that a child’s level of phonological awareness is the best predictor of reading success in the early grades. As described earlier, pho- nological awareness involves the recognition and manipulation of the smalle st units of language. With renewed interest in children’s reading comprehension, it also has been recognized tha t important foundational skills relevant to readin g comprehension emerge in their comprehension of oral narratives such as stories well before a child enters school. With the well- established links between o ral language abilities and reading, spelling and written composition, it comes as no surprise that longitudinal studies of children with language disorders have rev ealed long-term difﬁculty inthelanguageartsareas andeveninmathematics. Hugh Catts and colleagues have found that although not all children identiﬁed with a language disorder at kindergarten will have reading difﬁculties, there is a signiﬁcantly high risk for most of them by the time they reach second grade. Two areas of language 568 Language Disorders impairment contribute somewhat independently to their reading difﬁculties. Impaired phonological awareness abilities contribute strongly to difﬁculties in learning to decode written language, whereas impairment in the semantic and syntactic components of oral language contribute more to difﬁcul ties in reading comprehen- sion. Indeed, second-grade c hildren who demonstrate poor reading comprehension but intact phonological awareness have been shown to have an oral language disorder that was not identiﬁed in their preschool years. Better reading outcomes can be expected for children whose oral language improved between kindergarten and second grade, but the fact remains that the more severe the language disorder is, the poorer the reading outcome will be. Thus, although phonological aware- ness instruction has been advocated for children with language disorders, it is with the caveat that it be part of comprehensive oral language remediation. Persis- tence of academic difﬁculties has been revealed in lon- gitudinal follow-up studies. Adolescents with histories of a language disorder in kindergarten continue to dem- onstrate spelling, word recognition, and reading com- prehension deﬁcits. When interviewed about their level of literacy, they report that while they read regularly, they do not always understand what they have read and that their written language skills do not meet the expectations of classroom curriculum. Even those whose language impairment appeared to have resolved by kindergarten continue t o demonstrate a high rate of literacy problems. Finally, whereas poor reading outcomes are easily understood given the nature of language disorders, the reported difﬁculties that children with language disor- ders have with mathematics are not necessarily as intuitive. Children with language disorder are at risk not only for difﬁculties with the more obvious areas of ‘‘word problems’’ and when asked to verbally explain their solutions to math problems but also for the rote memory aspects of math needed for number recall and calculation. The relationships between lan- guage disorders and success in the academic setting underscore the serious long-term impact of a language disorder, its consequences to both the individual and society. Language Disorders and Socialization An account of language disorders in children would be incomplete without a consideration of its impact on social competence. The school setting is highly inﬂuential in a child’s social development and recip- rocally depends upon it. Kindergarten teachers iden- tify social maturity as more important to school readiness than skills like counting and letter recogni- tion. School becomes a primary environment for peer interaction and the development of friendships. Thus, once again, language, as the medium for interpersonal interactions, plays a critical role in the development of social competence, and children with language dis- orders are disadvantaged. The social communication difﬁculties of children with language disorders, while variable, are well documented. These children may demonstrate difﬁ- culty initiating interactions with other children and may not be responsive to verbal overtures from peers; they may not be able to gain access to a playgroup or to be integrated into classroom work groups. The general social behavior pattern appears to be one of reticence and withdrawal, which make friendship for- mation challenging and peer acceptance problematic. As early as the preschool years, other children recog- nize children with diminished language abilities, and their social interactions are negatively affected. Although not well studied, there are reports that chil- dren with language disorders may experience peer victimization and bullying. The precise relationship of language disorders to social competence is a matter of some debate, but the elevated risk for negative social outcomes is not. The limited longitudinal studies that have examined psychosocial and quality of life issues in 20- and 30-year-old adults who have a history of language disorder suggest evolving consequences. The limited data available to date suggest that at least a cohort of mostly males in their 20s, who not only have a history of language disorders but continue to demonstrate diminished language abilities, are satis- ﬁed with their lives. It was suggested, however, that when these young men experienced the full weight of adult responsibility, satisfaction ratings might change. Indeed, longitudinal research out of Great Britain revealed that adults in their 30s with a history of severe developmental language disorders indeed had poorer social adaptation and psychosocial adjustment. These adults experienced diminished personal relationships, difﬁculty maintaining employment, and even difﬁculty with bullying from coworkers. Although any individ- ual’s personal outcome will surely vary, the data are clear on the long-range prognosis for the social compe- tence of individuals with developmental language disorders. Language Disorders 569 Childhood language disorders are a chronic condi- tion with signiﬁcant impact upon social, academic, and vocational success. Yet there is evidence, and more is accumulating, that intervention early and tailored to the changing facets of the disorder throughout the school years can improve outcomes for these individuals. Elizabeth Skarakis-Doyle See also Communication Disorders; Disabilities; Dyslexia; Learning Disabilities Further Readings Bishop, D. V. M. (2004). Speciﬁc language impairment: Diagnostic dilemmas. In L. Verhoeven & H. van Balkom (Eds.), Classiﬁcation of developmental language disorders: Theoretical issues and clinical implications (pp. 309–326). Mahwah, NJ: Lawrence Erlbaum. Brinton, B., & Fujiki, M. (2005). Social competence in children with language impairment: Making connections. Seminars in Speech and Language , 26 , 151–159. Catts, H. W. (1993). The relationship between speech-language impairments and reading disabilities. Journal of Speech and Hearing Research , 36 , 948–958. Catts, H. W., Fey, M. E., Tomblin, J. B., & Zhang, X. (2002). A longitudinal investigation of reading outcomes in children with language impairment. Journal of Speech, Language and Hearing Research , 45 , 1142–1157. Catts, H. W., Fey, M. E., Zhang, X., & Tomblin, J. B. (1999). Language basis of reading disabilities: Evidence form a longitudinal investigation. Scientiﬁc Studies of Reading , 3 , 331–361. Fazio, B. B. (1999). Arithmetic calculation, short-term memory, and language performance in children with speciﬁc language impairment: A 5-year follow-up. Journal of Speech, Language and Hearing Research , 42 , 420–431. Johnson, C. J., Beitchman, J. H., Young, A., Escobar, M., Atkinson, L., Wilson, B., et al. (1999). Fourteen-year follow-up of children with and without speech/language impairments: Speech/language stability and outcomes. Journal of Speech, Language and Hearing Research , 42 , 744–760. Nation, K., Clarke, P., Marshall, C. M., & Durand, M. (2004). Hidden language impairments in children: Parallels between poor reading comprehension and speciﬁc language impairment? Journal of Speech, Language and Hearing Research , 47 , 199–211. Naucler, K., & Magnusson, E. (1998). Reading and writing development: Report from an ongoing longitudinal study of language-disordered and normal groups from pre-school to adolescence. Folia Phoniatrica et Logopaedica , 50 , 272–282. Records, N. L., Tomblin, J. B., & Freese, P. R. (1992). The quality of life of young adults with histories of speciﬁc language impairment. American Journal of Speech Language Pathology , 1 , 44–53. Stothard, S. E., Snowling, M. J., Bishop, D. V. M., Chipchase, B. B., & Kaplan, C. A. (1998). Language- impaired preschoolers: A follow-up into adolescence. Journal of Speech, Language and Hearing Research , 41 , 407–418. Tomblin, J. B. (2006). A normativist account of language-based learning disability. Learning Disabilities Research & Practice , 2 , 8–18. L ATINOS See H ISPANIC A MERICANS L EARNED H ELPLESSNESS Learned helplessness is a term coined by Martin Seligman to characterize the overgeneralized learning of helpless responses that occur when animals are repeatedly exposed to noxious, uncontrollable, and inescapable situations. What is learned is that their actions will not effect the outcome they desire. In humans, learned helplessness is particularly problem- atic in education; students may falsely believe that effortful actions, such as studying, will have no effect on performance or learning. Seligman and colleagues’ explanation of learned helplessness focuses on three components: outcome contingency, mediating cognitions, and behavior out- comes. Contingency concerns individuals’ perception that outcomes are contingent on their behavior. Individ- uals ‘‘learn’’ helplessness when experiences create the belief that outcomes are not contingent upon their actions. Learned helplessness theory also postulates that certain types of cognitions, called causal attribu- tions , mediate this relation between experience and learned helplessness. Finally, learned helplessness theory postulates that the behavioral outcomes of perceived helplessness include passivity, quitting, and depression. The idea of learned helplessness has helped research- ers and educators understand why some students repeat- edly experience more failures and give up—often before even trying. The serio us implications of learned 570 Learned Helplessness