Ethical and Professional Issues in Psychology Testing

CHAPTER 1 Applications and Consequences of Psychological Testing TOPIC 1A The Nature and Uses of Psychological Testing

If you ask average citizens “What do you know about psychological tests?” they might mention something about intelligence tests, inkblots, and true-false inventories such as the widely familiar MMPI. Most likely, their understanding of tests will focus on quantifying intelligence and detecting personality problems, as this is the common view of how tests are used in our society. Certainly, there is more than a grain of truth to this common view: Measures of personality and intelligence are still the essential mainstays of psychological testing. However, modern test developers have produced many other kinds of tests for diverse and imaginative purposes that even the early pioneers of testing could not have anticipated. The purpose of this chapter is to discuss the varied applications of psychological testing and also to review the ethical and social consequences of this enterprise. The chapter begins with a panoramic survey of psychological tests and their often surprising applications. In Topic 1A, The Nature and Uses of Psychological Testing, we summarize the different types and varied applications of modern tests. We also introduce the reader to a host of factors that can influence the soundness of testing such as adherence to standardized procedures, establishment of rapport, and the motivation of the examinee to deceive. In Topic 1B, Ethical and Social Implications of Testing, we further develop the theme that testing is a consequential endeavor. In this topic, we survey professional guidelines that impact testing and review the influence of cultural background on test results.

1.1 THE CONSEQUENCES OF TESTING From birth to old age, we encounter tests at almost every turning point in life. The baby’s first test conducted immediately after birth is the Apgar test, a quick, multivariate assessment of heart rate, respiration, muscle tone, reflex irritability, and color. The total Apgar score (0 to 10) helps determine the need for any immediate medical attention. Later, a toddler who previously received a low Apgar score might be a candidate for developmental disability assessment. The preschool child may take school-readiness tests. Once a school career begins, each student endures hundreds, perhaps thousands, of academic tests before graduation —not to mention possible tests for learning disability, giftedness, vocational interest, and college admission. After graduation, adults may face tests for job entry, driver’s license, security clearance, personality function, marital compatibility, developmental disability, brain dysfunction—the list is nearly endless. Some

persons even encounter one final indignity in the frailness of their later years: a test to determine their competency to manage financial affairs. Tests are used in almost every nation on earth for counseling, selection, and placement. Testing occurs in settings as diverse as schools, civil service, industry, medical clinics, and counseling centers. Most persons have taken dozens of tests and thought nothing of it. Yet, by the time the typical individual reaches retirement age, it is likely that psychological test results will have helped to shape his or her destiny. The deflection of the life course by psychological test results might be subtle, such as when a prospective mathematician qualifies for an accelerated calculus course based on tenth-grade achievement scores. More commonly, psychological test results alter individual destiny in profound ways. Whether a person is admitted to one college and not another, offered one job but refused a second, diagnosed as depressed or not—all such determinations rest, at least in part, on the

meaning of test results as interpreted by persons in authority. Put simply, psychological test results change lives. For this reason it is prudent —indeed, almost mandatory—that students of psychology learn about the contemporary uses and occasional abuses of testing. In Case Exhibit 1.1, the life-altering aftermath of psychological testing is illustrated by means of several true case history examples. CASE EXHIBIT 1.1 True-Life Vignettes of Testing The influence of psychological testing is best illustrated by example. Consider these brief vignettes: • A shy, withdrawn 7-year-old girl is

administered an IQ test by a school psychologist. Her score is phenomenally higher than the teacher expected. The student is admitted to a gifted and talented program where she blossoms into a self- confident and gregarious scholar.

• Three children in a family living near a lead smelter are exposed to the toxic effects of

lead dust and suffer neurological damage. Based in part on psychological test results that demonstrate impaired intelligence and shortened attention span in the children, the family receives an $8 million settlement from the company that owns the smelter.

• A candidate for a position as police officer is administered a personality inventory as part of the selection process. The test indicates that the candidate tends to act before thinking and resists supervision from authority figures. Even though he has excellent training and impresses the interviewers, the candidate does not receive a job offer.

• A student, unsure of what career to pursue, takes a vocational interest inventory. The test indicates that she would like the work of a pharmacist. She signs up for a prepharmacy curriculum but finds the classes to be both difficult and boring. After three years, she abandons pharmacy for a major in dance, frustrated that she still faces three more years of college to earn a degree.

These cases demonstrate that test results impact individual lives and the collective social fabric in powerful and far-reaching ways. In the first story about the hidden talent of a 7-year-old girl, cognitive test results changed her life trajectory for the better. In the second case involving the tragic saga of children exposed to lead poisoning, the test data helped redress a social injustice. In the third situation—the impulsive candidate for police officer—personality test results likely served the public interest by tipping the balance against a questionable applicant. But test results do not always provide a positive conclusion. In the last case mentioned above, a young student wasted time and money following the seemingly flawed guidance of a well-known vocational inventory. The idea of a test is thus a pervasive element of our culture, a feature we take for granted. However, the layperson’s notion of a test does not necessarily coincide with the more restrictive view held by psychometricians. A psychometrician is a specialist in psychology or education who develops and evaluates

psychological tests. Because of widespread misunderstandings about the nature of tests, it is fitting that we begin this topic with a fundamental question, one that defines the scope of the entire book: What is a test?

1.2 DEFINITION OF A TEST A test is a standardized procedure for sampling behavior and describing it with categories or scores. In addition, most tests have norms or standards by which the results can be used to predict other, more important behaviors. We elaborate these characteristics in the sections that follow, but first it is instructive to portray the scope of the definition. Included in this view are traditional tests such as personality questionnaires and intelligence tests, but the definition also subsumes diverse procedures that the reader might not recognize as tests. For example, all of the following could be tests according to the definition used in this book: a checklist for rating the social skills of a youth with mental retardation; a nontimed measure of

mastery in adding pairs of three-digit numbers; microcomputer appraisals of reaction time; and even situational tests such as observing an individual working on a group task with two “helpers” who are obstructive and uncooperative. In sum, tests are enormously varied in their formats and applications. Nonetheless, most tests possess these defining features: • Standardized procedure • Behavior sample • Scores or categories • Norms or standards • Prediction of nontest behavior

In the sections that follow, we examine each of these characteristics in more detail. The portrait that we draw pertains especially to norm- referenced tests—tests that use a well-defined population of persons for their interpretive framework. However, the defining characteristics of a test differ slightly for the special case of criterion-referenced tests—tests that measure what a person can do rather than comparing results to the performance levels of

others. For this reason, we provide a separate discussion of criterion-referenced tests. Standardized procedure is an essential feature of any psychological test. A test is considered to be standardized if the procedures for administering it are uniform from one examiner and setting to another. Of course, standardization depends to some extent on the competence of the examiner. Even the best test can be rendered useless by a careless, poorly trained, or ill-informed tester, as the reader will discover later in this topic. However, most examiners are competent. Standardization, therefore, rests largely on the directions for administration found in the instructional manual that typically accompanies a test. The formulation of directions is an essential step in the standardization of a test. In order to guarantee uniform administration procedures, the test developer must provide comparable stimulus materials to all testers, specify with considerable precision the oral instructions for each item or subtest, and advise the examiner

how to handle a wide range of queries from the examinee. To illustrate these points, consider the number of different ways a test developer might approach the assessment of digit span—the maximum number of orally presented digits a subject can recall from memory. An unstandardized test of digit span might merely suggest that the examiner orally present increasingly long series of numbers until the subject fails. The number of digits in the longest series recalled would then be the subject’s digit span. Most readers can discern that such a loosely defined test will lack uniformity from one examiner to another. If the tester is free to improvise any series of digits, what is to prevent him or her from presenting, with the familiar inflection of a television announcer, “1-800-325-3535”? Such a series would be far easier to recall than a more random set, such as, “7-2-8-1-9-4-6-3-7-4-2.” The speed of presentation would also crucially affect the uniformity of a digit span test. For purposes of standardization, it is essential that every

examiner present each series at a constant rate, for example, one digit per second. Finally, the examiner needs to know how to react to unexpected responses such as a subject asking, “Could you repeat that again?” For obvious reasons, the usual advice is “No.” A psychological test is also a limited sample of behavior. Neither the subject nor the examiner has sufficient time for truly comprehensive testing, even when the test is targeted to a well- defined and finite behavior domain. Thus, practical constraints dictate that a test is only a sample of behavior. Yet, the sample of behavior is of interest only insofar as it permits the examiner to make inferences about the total domain of relevant behaviors. For example, the purpose of a vocabulary test is to determine the examinee’s entire word stock by requesting definitions of a very small but carefully selected sample of words. Whether the subject can define the particular 35 words from a vocabulary subtest (e.g., on the Wechsler Adult Intelligence Scale-IV, or the WAIS-IV) is of little direct consequence. But the indirect meaning of such

results is of great import because it signals the examinee’s general knowledge of vocabulary. An interesting point—and one little understood by the lay public—is that the test items need not resemble the behaviors that the test is attempting to predict. The essential characteristic of a good test is that it permits the examiner to predict other behaviors—not that it mirrors the to-be-predicted behaviors. If answering “true” to the question “I drink a lot of water” happens to help predict depression, then this seemingly unrelated question is a useful index of depression. Thus, the reader will note that successful prediction is an empirical question answered by appropriate research. While most tests do sample directly from the domain of behaviors they hope to predict, this is not a psychometric requirement. A psychological test must also permit the derivation of scores or categories. Thorndike (1918) expressed the essential axiom of testing in his famous assertion, “Whatever exists at all exists in some amount.” McCall (1939) went a step further, declaring, “Anything that exists in

amount can be measured.” Testing strives to be a form of measurement akin to procedures in the physical sciences whereby numbers represent abstract dimensions such as weight or temperature. Every test furnishes one or more scores or provides evidence that a person belongs to one category and not another. In short, psychological testing sums up performance in numbers or classifications. The implicit assumption of the psychometric viewpoint is that tests measure individual differences in traits or characteristics that exist in some vague sense of the word. In most cases, all people are assumed to possess the trait or characteristic being measured, albeit in different amounts. The purpose of the testing is to estimate the amount of the trait or quality possessed by an individual. In this context, two cautions are worth mentioning. First, every test score will always reflect some degree of measurement error. The imprecision of testing is simply unavoidable: Tests must rely on an external sample of behavior to estimate an unobservable and,

therefore, inferred characteristic. Psychometricians often express this fundamental point with an equation: X = T + e where X is the observed score, T is the true score, and e is a positive or negative error component. The best that a test developer can do is make e very small. It can never be completely eliminated, nor can its exact impact be known in the individual case. We discuss the concept of measurement error in Topic 3B, Concepts of Reliability. The second caution is that test consumers must be wary of reifying the characteristic being measured. Test results do not represent a thing with physical reality. Typically, they portray an abstraction that has been shown to be useful in predicting nontest behaviors. For example, in discussing a person’s IQ, psychologists are referring to an abstraction that has no direct, material existence but that is, nonetheless, useful in predicting school achievement and other outcomes.

A psychological test must also possess norms or standards. An examinee’s test score is usually interpreted by comparing it with the scores obtained by others on the same test. For this purpose, test developers typically provide norms—a summary of test results for a large and representative group of subjects (Petersen, Kolen, & Hoover, 1989). The norm group is referred to as the standardization sample. The selection and testing of the standardization sample is crucial to the usefulness of a test. This group must be representative of the population for whom the test is intended or else it is not possible to determine an examinee’s relative standing. In the extreme case when norms are not provided, the examiner can make no use of the test results at all. An exception to this point occurs in the case of criterion-referenced tests, discussed later. Norms not only establish an average performance but also serve to indicate the frequency with which different high and low scores are obtained. Thus, norms allow the tester to determine the degree to which a score

deviates from expectations. Such information can be very important in predicting the nontest behavior of the examinee. Norms are of such overriding importance in test interpretation that we consider them at length in a separate section later in this text. Finally, tests are not ends in themselves. In general, the ultimate purpose of a test is to predict additional behaviors, other than those directly sampled by the test. Thus, the tester may have more interest in the nontest behaviors predicted by the test than in the test responses per se. Perhaps a concrete example will clarify this point. Suppose an examiner administers an inkblot test to a patient in a psychiatric hospital. Assume that the patient responds to one inkblot by describing it as “eyes peering out.” Based on established norms, the examiner might then predict that the subject will be highly suspicious and a poor risk for individual psychotherapy. The purpose of the testing is to arrive at this and similar predictions—not to determine whether the subject perceives eyes staring out from the blots.

The ability of a test to predict nontest behavior is determined by an extensive body of validational research, most of which is conducted after the test is released. But there are no guarantees in the world of psychometric research. It is not unusual for a test developer to publish a promising test, only to read years later that other researchers find it deficient. There is a lesson here for test consumers: The fact that a test exists and purports to measure a certain characteristic is no guarantee of truth in advertising. A test may have a fancy title, precise instructions, elaborate norms, attractive packaging, and preliminary findings—but if in the dispassionate study of independent researchers the test fails to predict appropriate nontest behaviors, then it is useless. 1.3 FURTHER DISTINCTIONS IN TESTING The chief features of a test previously outlined apply especially to norm-referenced tests, which constitute the vast majority of tests in use. In a norm-referenced test, the performance of each

examinee is interpreted in reference to a relevant standardization sample (Petersen, Kolen, & Hoover, 1989). However, these features are less relevant in the special case of criterion-referenced tests, since these instruments suspend the need for comparing the individual examinee with a reference group. In a criterion-referenced test, the objective is to determine where the examinee stands with respect to very tightly defined educational objectives (Berk, 1984). For example, one part of an arithmetic test for 10-year-olds might measure the accuracy level in adding pairs of two-digit numbers. In an untimed test of 20 such problems, accuracy should be nearly perfect. For this kind of test, it really does not matter how the individual examinee compares to others of the same age. What matters is whether the examinee meets an appropriate, specified criterion—for example, 95 percent accuracy. Because there is no comparison to the normative performance of others, this kind of measurement tool is aptly designated a criterion-referenced test. The important distinction here is that,

unlike norm-referenced tests, criterion- referenced tests can be meaningfully interpreted without reference to norms. We discuss criterion-referenced tests in more detail in Topic 3A, Norms and Test Standardization. Another important distinction is between testing and assessment, which are often considered equivalent. However, they do not mean exactly the same thing. Assessment is a more comprehensive term, referring to the entire process of compiling information about a person and using it to make inferences about characteristics and to predict behavior. Assessment can be defined as appraising or estimating the magnitude of one or more attributes in a person. The assessment of human characteristics involves observations, interviews, checklists, inventories, projectives, and other psychological tests. In sum, tests represent only one source of information used in the assessment process. In assessment, the examiner must compare and combine data from different sources. This is an inherently subjective process that requires the examiner to

sort out conflicting information and make predictions based on a complex gestalt of data. The term assessment was invented during World War II (WWII) to describe a program to select men for secret service assignment in the Office of Strategic Services (OSS Assessment Staff, 1948). The OSS staff of psychologists and psychiatrists amassed a colossal amount of information on candidates during four grueling days of written tests, interviews, and personality tests. In addition, the assessment process included a variety of real-life situational tests based on the realization that there was a difference between know-how and can-do: We made the candidates actually attempt the

tasks with their muscles or spoken words, rather than merely indicate on paper how the tasks could be done. We were prompted to introduce realistic tests of ability by such findings as this: that men who earn a high score in Mechanical Comprehension, a paper-and-pencil test, may be below average when it comes to solving

mechanical problems with their hands. (OSS Assessment Staff, 1948, pp. 41–42)

The situational tests included group tasks of transporting equipment across a raging brook and scaling a 10-foot-high wall, as well as individual scrutiny of the ability to survive a realistic interrogation and to command two uncooperative subordinates in a construction task. On the basis of the behavioral observations and test results, the OSS staff rated the candidates on dozens of specific traits in such broad categories as leadership, social relations, emotional stability, effective intelligence, and physical ability. These ratings served as the basis for selecting OSS personnel.

1.4 TYPES OF TESTS Tests can be broadly grouped into two camps: group tests versus individual tests. Group tests are largely pencil-and-paper measures suitable to the testing of large groups of persons at the same time. Individual tests are instruments that

by their design and purpose must be administered one on one. An important advantage of individual tests is that the examiner can gauge the level of motivation of the subject and assess the relevance of other factors (e.g., impulsiveness or anxiety) on the test results. For convenience, we will sort tests into the eight categories depicted in Table 1.1. Each of the categories contains norm-referenced, criterion- referenced, individual, and group tests. The reader will note that any typology of tests is a purely arbitrary determination. For example, we could argue for yet another dichotomy: tests that seek to measure maximum performance (e.g., an intelligence test) versus tests that seek to gauge a typical response (e.g., a personality inventory). In a narrow sense, there are hundreds—perhaps thousands—of different kinds of tests, each measuring a slightly different aspect of the individual. For example, even two tests of intelligence might be arguably different types of measures. One test might reveal the assumption that intelligence is a biological construct best

measured through brain waves, whereas another might be rooted in the traditional view that intelligence is exhibited in the capacity to learn acculturated skills such as vocabulary. Lumping both measures under the category of intelligence tests is certainly an oversimplification, but nonetheless a useful starting point. TABLE 1.1 The Main Types of Psychological Tests

Intelligence Tests: Measure an individual’s ability in relatively global areas such as verbal comprehension, perceptual organization, or reasoning and thereby help determine potential for scholastic work or certain occupations. Aptitude Tests: Measure the capability for a relatively specific task or type of skill; aptitude tests are, in effect, a narrow form of ability testing. Achievement Tests: Measure a person’s degree of learning, success, or accomplishment in a subject or task. Creativity Tests: Assess novel, original thinking and the capacity to find unusual or unexpected solutions, especially for vaguely defined problems. Personality Tests: Measure the traits, qualities, or behaviors that determine a person’s individuality; such tests include checklists, inventories, and projective techniques. Interest Inventories: Measure an individual’s preference for certain activities or topics and thereby help determine occupational choice. Behavioral Procedures: Objectively describe and count the frequency of a behavior,

Intelligence tests were originally designed to sample a broad assortment of skills in order to estimate the individual’s general intellectual level. The Binet-Simon scales were successful, in part, because they incorporated heterogeneous tasks, including word definitions, memory for designs, comprehension questions, and spatial visualization tasks. The group intelligence tests that blossomed with such profusion during and after WWII also tested diverse abilities—witness the Army Alpha with its eight different sections measuring practical judgment, information, arithmetic, and reasoning, among other skills. Modern intelligence tests also emulate this historically established pattern by sampling a wide variety of proficiencies deemed important in our culture. In general, the term intelligence test refers to a test that yields an overall summary score based on results from a heterogeneous sample of items. Of course, such a test might also provide a profile of subtest scores as well, but it is the overall score that generally attracts the most attention.

Aptitude tests measure one or more clearly defined and relatively homogeneous segments of ability. Such tests come in two varieties: single aptitude tests and multiple aptitude test batteries. A single aptitude test appraises, obviously, only one ability, whereas a multiple aptitude test battery provides a profile of scores for a number of aptitudes. Aptitude tests are often used to predict success in an occupation, training course, or educational endeavor. For example, the Seashore Measures of Musical Talents (Seashore, 1938), a series of tests covering pitch, loudness, rhythm, time, timbre, and tonal memory, can be used to identify children with potential talent in music. Specialized aptitude tests also exist for the assessment of clerical skills, mechanical abilities, manual dexterity, and artistic ability. The most common use of aptitude tests is to determine college admissions. Most every college student is familiar with the SAT (Scholastic Assessment Test, previously called the Scholastic Aptitude Test) of the College Entrance Examination Board. This test contains

a Verbal section stressing word knowledge and reading comprehension; a Mathematics section stressing algebra, geometry, and insightful reasoning; and a Writing section. In effect, colleges that require certain minimum scores on the SAT for admission are using the test to predict academic success. Achievement tests measure a person’s degree of learning, success, or accomplishment in a subject matter. The implicit assumption of most achievement tests is that the schools have taught the subject matter directly. The purpose of the test is then to determine how much of the material the subject has absorbed or mastered. Achievement tests commonly have several subtests, such as reading, mathematics, language, science, and social studies. The distinction between aptitude and achievement tests is more a matter of use than content (Gregory, 1994a). In fact, any test can be an aptitude test to the extent that it helps predict future performance. Likewise, any test can be an achievement test insofar as it reflects how much the subject has learned. In practice,

then, the distinction between these two kinds of instruments is determined by their respective uses. On occasion, one instrument may serve both purposes, acting as an aptitude test to forecast future performance and an achievement test to monitor past learning. Creativity tests assess a subject’s ability to produce new ideas, insights, or artistic creations that are accepted as being of social, aesthetic, or scientific value. Thus, measures of creativity emphasize novelty and originality in the solution of fuzzy problems or the production of artistic works. A creative response to one problem is illustrated in Figure 1.1. Tests of creativity have a checkered history. In the 1960s, they were touted as a useful alternative to intelligence tests and used widely in U.S. school systems. Educators were especially impressed that creativity tests required divergent thinking—putting forth a variety of answers to a complex or fuzzy problem—as opposed to convergent thinking— finding the single correct solution to a well- defined problem. For example, a creativity test

might ask the examinee to imagine all the things that would happen if clouds had strings trailing from them down to the ground. Students who could come up with a large number of consequences were assumed to be more creative than their less-imaginative colleagues. However, some psychometricians are skeptical, concluding that creativity is just another label for applied intelligence.

FIGURE 1.1 Solutions to the Nine-Dot Problem as Examples of Creativity Note: Without lifting the pencil, draw through all the dots with as few straight lines as possible. The usual

solution is shown in a. Creative solutions are depicted in b and c. Personality tests measure the traits, qualities, or behaviors that determine a person’s individuality; this information helps predict future behavior. These tests come in several different varieties, including checklists, inventories, and projective techniques such as sentence completions and inkblots (Table 1.2). Interest inventories measure an individual’s preference for certain activities or topics and thereby help determine occupational choice. These tests are based on the explicit assumption that interest patterns determine and, therefore, also predict job satisfaction. For example, if the examinee has the same interests as successful and satisfied accountants, it is thought likely that he or she would enjoy the work of an accountant. The assumption that interest patterns predict job satisfaction is largely borne out by empirical studies, as we will review in a later chapter. TABLE 1.2 Examples of Personality Test Items

(a) An Adjective Checklist Check those words which describe you: ( ) relaxed ( ) assertive ( ) thoughtful ( ) curious ( ) cheerful ( ) even-tempered ( ) impatient ( ) skeptical ( ) morose ( ) impulsive ( ) optimistic ( ) anxious (b) A True-False Inventory

Circle true or false as each statement applies to you: T   F   I like sports magazines. T   F   Most people would lie to get

a job. T   F   I like big parties where there

is lots of noisy fun. T   F   Strange thoughts possess me

for hours at a time. T   F   I often regret the missed

opportunities in my life. T   F   Sometimes I feel anxious for

no reason at all. T   F   I like everyone I have met. T   F   Falling asleep is seldom a

problem for me. (c) A Sentence Completion Projective Test Complete each sentence with the first thought that comes to you: I feel bored when What I need most is I like people who My mother was

Many kinds of behavioral procedures are available for assessing the antecedents and consequences of behavior, including checklists, rating scales, interviews, and structured observations. These methods share a common assumption that behavior is best understood in terms of clearly defined characteristics such as frequency, duration, antecedents, and consequences. Behavioral procedures tend to be highly pragmatic in that they are usually interwoven with treatment approaches. Neuropsychological tests are used in the assessment of persons with known or suspected brain dysfunction. Neuropsychology is the study of brain–behavior relationships. Over the years, neuropsychologists have discovered that certain tests and procedures are highly sensitive to the effects of brain damage. Neuropsychologists use these specialized tests and procedures to make inferences about the locus, extent, and consequences of brain damage. A full neuropsychological assessment typically requires three to eight hours of one-on-one testing with an extensive battery of measures.

Examiners must undergo comprehensive advanced training in order to make sense out of the resulting mass of test data.

1.5 USES OF TESTING By far the most common use of psychological tests is to make decisions about persons. For example, educational institutions frequently use tests to determine placement levels for students, and universities ascertain who should be admitted, in part, on the basis of test scores. State, federal, and local civil service systems also rely heavily on tests for purposes of personnel selection. Even the individual practitioner exploits tests, in the main, for decision making. Examples include the consulting psychologist who uses a personality test to determine that a police department hire one candidate and not another, and the neuropsycholo-gist who employs tests to conclude that a client has suffered brain damage.

But simple decision making is not the only function of psychological testing. It is convenient to distinguish five uses of tests: • Classification • Diagnosis and treatment planning • Self-knowledge • Program evaluation • Research

These applications frequently overlap and, on occasion, are difficult to distinguish one from another. For example, a test that helps determine a psychiatric diagnosis might also provide a form of self-knowledge. Let us examine these applications in more detail. The term classification encompasses a variety of procedures that share a common purpose: assigning a person to one category rather than another. Of course, the assignment to categories is not an end in itself but the basis for differential treatment of some kind. Thus, classification can have important effects such as granting or restricting access to a specific college or determining whether a person is hired for a particular job. There are many variant

forms of classification, each emphasizing a particular purpose in assigning persons to categories. We will distinguish placement, screening, certification, and selection. Placement is the sorting of persons into different programs appropriate to their needs or skills. For example, universities often use a mathematics placement exam to determine whether students should enroll in calculus, algebra, or remedial courses. Screening refers to quick and simple tests or procedures to identify persons who might have special characteristics or needs. Ordinarily, psychometricians acknowledge that screening tests will result in many misclassifications. Examiners are, therefore, advised to do follow- up testing with additional instruments before making important decisions on the basis of screening tests. For example, to identify children with highly exceptional talent in spatial thinking, a psychologist might administer a 10- minute paper-and-pencil test to every child in a school system. Students who scored in the top

10 percent might then be singled out for more comprehensive testing. Certification and selection both have a pass/fail quality. Passing a certification exam confers privileges. Examples include the right to practice psychology or to drive a car. Thus, certification typically implies that a person has at least a minimum proficiency in some discipline or activity. Selection is similar to certification in that it confers privileges such as the opportunity to attend a university or to gain employment. Another use of psychological tests is for diagnosis and treatment planning. Diagnosis consists of two intertwined tasks: determining the nature and source of a person’s abnormal behavior, and classifying the behavior pattern within an accepted diagnostic system. Diagnosis is usually a precursor to remediation or treatment of personal distress or impaired performance. Psychological tests often play an important role in diagnosis and treatment planning. For example, intelligence tests are absolutely

essential in the diagnosis of mental retardation. Personality tests are helpful in diagnosing the nature and extent of emotional disturbance. In fact, some tests such as the MMPI were devised for the explicit purpose of increasing the efficiency of psychiatric diagnosis. Diagnosis should be more than mere classification, more than the assignment of a label. A proper diagnosis conveys information— about strengths, weaknesses, etiology, and best choices for remediation/treatment. Knowing that a child has received a diagnosis of learning disability is largely useless. But knowing in addition that the same child is well below average in reading comprehension, is highly distractible, and needs help with basic phonics can provide an indispensable basis for treatment planning. Psychological tests also can supply a potent source of self-knowledge. In some cases, the feedback a person receives from psychological tests can change a career path or otherwise alter a person’s life course. Of course, not every instance of psychological testing provides self-

knowledge. Perhaps in the majority of cases the client already knows what the test results divulge. A high-functioning college student is seldom surprised to find that his IQ is in the superior range. An architect is not perplexed to hear that she has excellent spatial reasoning skills. A student with meager reading capacity is usually not startled to receive a diagnosis of “learning disability.” Another use for psychological tests is the systematic evaluation of educational and social programs. We have more to say about the evaluation of educational programs when we discuss achievement tests in a later chapter. We focus here on the use of tests in the evaluation of social programs. Social programs are designed to provide services that improve social conditions and community life. For example, Project Head Start is a federally funded program that supports nationwide pre-school teaching projects for underprivileged children (McKey and others, 1985). Launched in 1965 as a precedent-setting attempt to provide child development programs to low-income families,

Head Start has provided educational enrichment and health services to millions of at-risk preschool children. But exactly what impact does the multi-billion- dollar Head Start program have on early childhood development? Congress wanted to know if the program improved scholastic performance and reduced school failure among the enrollees. But the centers vary by sponsoring agencies, staff characteristics, coverage, content, and objectives, so the effects of Head Start are not easy to ascertain. Psychological tests provide an objective basis for answering these questions that is far superior to anecdotal or impressionistic reporting. In general, Head Start children show immediate gains in IQ, school readiness, and academic achievement, but these gains dissipate in the ensuing years (Figure 1.2). So far we have discussed the practical application of psychological tests to everyday problems such as job selection, diagnosis, or program evaluation. In each of these instances, testing serves an immediate, pragmatic purpose: helping the tester make decisions about persons

or programs. But tests also play a major role in both the applied and theoretical branches of behavioral research. As an example of testing in applied research, consider the problem faced by neuropsychologists who wish to investigate the hypothesis that low-level lead absorption causes behavioral deficits in children. The only feasible way to explore this supposition is by testing normal and lead-burdened children with a battery of psychological tests. Needleman and associates (1979) used an array of traditional and innovative tests to conclude that low-level lead absorption causes decrements in IQ, impairments in reaction time, and escalations of undesirable classroom behaviors. Their conclusions inspired a tumultuous and bitter exchange of opinions that we will not review here (Needleman et al., 1990). However, the passions inspired by this study epitomize an instructive point: Academicians and public policymakers respect psychological tests. Why else would they engage in lengthy, acrimonious debates about the validity of testing-based research findings?

FIGURE 1.2 Longitudinal Test Results from the Head Start Project Source: From McKey, R. H., and others. (1985). The impact of Head Start on children, families and communities. Washington, DC: U.S. Government Printing Office. In the public domain.

1.6 FACTORS INFLUENCING THE SOUNDNESS OF TESTING Psychological testing is a dynamic process influenced by many factors. Although examiners strive to ensure that test results accurately reflect the traits or capacities being assessed,

many extraneous factors can sway the outcome of psychological testing. In this section, we review the potentially crucial impact of several sources of influence: the manner of administration, the characteristics of the tester, the context of the testing, the motivation and experience of the examinee, and the method of scoring. The sensitivity of the testing process to extraneous influences is obvious in cases where the examiner is cold, hurried, or incompetent. However, invalid test results do not originate only from obvious sources such as blatantly nonstandard administration, hostile tester, noisy testing room, or fearful examinee. In addition, there are numerous, subtle ways in which method, examiner, context, or motivation can alter test results. We provide a comprehensive survey of these extraneous influences in the remainder of this topic.

1.7 STANDARDIZED PROCEDURES IN TEST ADMINISTRATION The interpretation of a psychological test is most reliable when the measurements are obtained under the standardized conditions outlined in the publisher’s test manual. Nonstandard testing procedures can alter the meaning of the test results, rendering them invalid and, therefore, misleading. Standardized procedures are so important that they are listed as an essential criterion for valid testing in the Standards for Educational and Psychological Testing (1999), a reference manual published jointly by the American Psychological Association and other groups: In typical applications, test administrators

should follow carefully the standardized procedures for administration and scoring specified by the test publisher. Specifications regarding instructions to test takers, time limits, the form of item

presentation or response, and test materials or equipment should be strictly observed. Exceptions should be made only on the basis of carefully considered professional judgment, primarily in clinical applications. (AERA, APA, NCME, 1999)

Suppose the instructions to the vocabulary section of a children’s intelligence test specify that the examiner should ask, “What does sofa mean, what is a sofa?” If a subject were to reply, “I’ve never heard that word,” an inexperienced tester might be tempted to respond, “You know, a couch—what is a couch?” This may strike the reader as a harmless form of fair play, a simple rephrasing of the original question. Yet, by straying from standardized procedures, the examiner has really given a different test. The point in asking for a definition of sofa (and not couch) is precisely that sofa is harder to define and, therefore, a better index of high-level vocabulary skills. Even though standardized testing procedures are normally essential, there are instances in which flexibility in procedures is desirable or even

necessary. As suggested in the APA Standards, such deviations should be reasoned and deliberate. An analogy to the spirit of the law versus the letter of the law is relevant here. An overly zealous examiner might capture the letter of the law, so to speak, by adhering literally and strictly to testing procedures outlined in the publisher’s manual. But is this really what most test publishers intend? Is it even how the test was actually administered to the normative sample? Most likely publishers would prefer that examiners capture the spirit of the law even if, on occasion, it is necessary to adjust testing procedures slightly. The need to adjust standardized procedures for testing is especially apparent when examining persons with certain kinds of disabilities. A subject with a speech impediment might be allowed to write down the answers to orally presented questions or to use gesture and pantomime in response to some items. For example, a test question might ask, “What shape is a ball?” The question is designed to probe the subject’s knowledge of common shapes, not to

examine whether the examinee can verbalize “round.” The written response round and the gestured response (a circular motion of the index finger) are equally correct, too. Minor adjustments in procedures that heed the spirit in which a test was developed occur on a regular basis and are no cause for alarm. These minor adjustments do not invalidate the established norms—on the contrary, the appropriate adaptation of procedures is necessary so that the norms remain valid. After all, the testers who collected data from the standardization sample did not act like heartless robots when posing questions to subjects. Examiners who wish to obtain valid results must likewise exercise a reasoned flexibility in testing procedures. However, considerable clinical experience is needed to determine whether an adjustment in procedure is minor or so substantial that existing norms no longer apply. This is why psychological examiners normally receive extensive supervised experience before they are

allowed to administer and interpret individual tests of ability or personality. In certain cases an examiner will knowingly depart from standard procedures to a substantial degree; this practice precludes the use of available test norms. In these instances, the test is used to help formulate clinical judgments rather than to determine a quantitative index. For example, when examining aphasic patients, it may be desirable to ignore time limits entirely and accept roundabout answers. The examiner might not even calculate a score. In these rare cases, the test becomes, in effect, an adjunct to the clinical interview. Of course, when the examiner does not adhere to standardized procedures, this should be stated explicitly in the written report.

1.8 DESIRABLE PROCEDURES OF TEST ADMINISTRATION

A small treatise could be written on desirable procedures of test administration, but we will have to settle for a brief listing of the most essential points. For more details, the interested reader can consult Sattler (2001) on the individual testing of children and Clemans (1971) on group testing. We discuss individual testing first, then briefly list some important points about desirable procedures in group testing. An essential component of individual testing is that examiners must be intimately familiar with the materials and directions before administration begins. Largely this involves extensive rehearsal and anticipation of unusual circumstances and the appropriate response. A well-prepared examiner has memorized key elements of verbal instructions and is ready to handle the unexpected. The uninitiated student of assessment often assumes that examination procedures are so simple and straightforward that a quick once- through reading of the manual will suffice as preparation for testing. Although some

individual tests are exceedingly rudimentary and uncomplicated, many of them have complexities of administration that, unheeded, can cause the examinee to fail items unnecessarily. For example, Choi and Proctor (1994) found that 25 of 27 graduate students made serious errors in the administration of the Stanford-Binet: Fourth Edition, even though the sessions were videotaped and the students knew their testing skills were being evaluated. Ramos, Alfonso, and Schermerhorn (2009) reviewed 108 protocols from the Woodcock Johnson III Tests of Cognitive Abilities administered by 36 first- year graduate students in a school psychology doctoral program. The researchers found an average of almost 5 errors per test, including the use of incorrect ceilings, failure to record errors, and failure to encircle the correct row for the total number correct. Loe, Kadlubek, and Williams (2007) reviewed 51 WISC-IV protocols administered by graduate students and found an average of almost 26 errors per protocol. The two most common errors were the failure to query incomplete or ambiguous verbal

responses, and granting too many points for substandard answers. In many cases, these errors materially affected the Full Scale IQ, shifting it upward or downward from the likely true score. What these studies confirm is that appropriate attention to the details of administration and scoring is essential for valid results. The necessity for intimate familiarity with testing procedures is well illustrated by the Block Design subtest of the WAIS-IV (Wechsler, 2008). The materials for the subtest include nine blocks (cubes) colored red on two sides, white on two sides, and red/white on two sides. The examinee’s task is to use the blocks to construct patterns depicted on cards. For the initial designs, four blocks are needed, while for more difficult designs, all nine blocks are provided (Figure 1.3).

FIGURE 1.3 Materials Similar to WAIS-IV Block Design Subtest Bright examinees have no difficulty comprehending this task and the exact instructions do not influence their performance appreciably. However, persons whose intelligence is average or below average need the elaborate demonstrations and corrections that are specified in the WAIS-IV manual

(Wechsler, 2008). In particular, the examiner demonstrates the first two designs and responds to the examinee’s success or failure on these according to a complex flow of reaction and counterreaction, as outlined in three pages of instructions. Woe to the tester who has not rehearsed this subtest and anticipated the proper response to examinees who falter on the first two designs. Sensitivity to Disabilities Another important ingredient of valid test administration is sensitivity to disabilities in the examinee. Impairments in hearing, vision, speech, or motor control may seriously distort test results. If the examiner does not recognize the physical disability responsible for the poor test performance, a subject may be branded as intellectually or emotionally impaired when, in fact, the essential problem is a sensory or motor disability. Vernon and Brown (1964) reported the tragic case of a young girl who was relegated to a hospital for the mentally retarded as a

consequence of the tester’s insensitivity to physical disability. The examiner failed to notice that the child was deaf and concluded that her Stanford-Binet IQ of 29 was valid. She remained in the hospital for five years, but was released after she scored an IQ of 113 on a performance-based intelligence test! After dismissal from the hospital, she entered a school for the deaf and made good progress. Persons with disabilities may require specialized tests for valid assessment. The reader will encounter a lengthy discussion of available tests for exceptional examinees in Chapter 7, Testing Special Populations. In this section, we concentrate on the vexing issues raised when standardized tests for normal populations are used with mildly or moderately disabled subjects. We include separate discussions of the testing process for examinees with a hearing, vision, speech, or motor control problem. However, the reader needs to know that many exceptional examinees have multiple disabilities.

Valid testing of a subject with a hearing impairment requires first of all that the examiner detect the existence of the disability! This is often more difficult than it seems. Many persons with mild hearing loss learn to compensate for this disability by pretending to understand what others say and waiting for further conversational cues to help clarify faintly perceived words or phrases. As a result, other persons—including psychologists—may not perceive that an individual with mild hearing loss has any disability at all. Failure to notice a hearing loss is particularly a problem with young examinees, who are usually poor informants about their disabilities. Young children are also prone to fluctuating hearing losses due to the periodic accumulation of fluid in the middle ear during intervals of mild illness (Vernon & Alles, 1986). A child with a fluctuating hearing loss may have normal hearing in the morning, but perceive conversational speech as a whisper just a few hours later.

Indications of possible hearing difficulty include lack of normal response to sound, inattentiveness, difficulty in following oral instructions, intent observation of the speaker’s lips, and poor articulation (Sattler, 1988). In all cases in which hearing impairment is suspected, referral for an audiological examination is crucial. If a serious hearing problem is confirmed, then the examiner should consider using one of the specialized tests discussed in Chapter 7, Testing Special Populations. In persons with a mild hearing loss, it is essential for the examiner to face the subject squarely, speak loudly, and repeat instructions slowly. It is also important to find a quiet room for testing. Ideally, a testing room will have curtains and textured wall surfaces to minimize the distracting effects of background noises. In contrast to those with hearing loss, subjects with visual disabilities generally attend well to verbally presented test materials. The examinee with visual impairment introduces a different kind of challenge to the examiner: detecting that

a visual impairment exists, and then ensuring that the subject can see the test materials well. Detecting visual impairment is a straightforward matter with adult subjects—in most cases, a mature examinee will freely volunteer information about visual impairment, especially if asked. However, children are poor informants about their visual capacities, so testers need to know the signs and symptoms of possible visual impairment in a young examinee. Common sense is a good starting point: Children who squint, blink excessively, or lose their place when reading may have a vision problem. Holding books or testing materials up close is another suspicious sign. Blurred or double vision may signify visual problems, as may headaches or nausea after reading. In general, it is so common for children to require corrective lenses that examiners should be on the lookout for a vision problem in any young subject who does not wear glasses and has not had a recent vision exam. Depending on the degree of visual impairment, examiners need to make corresponding

adjustments in testing. If the child’s vision is of no practical use, special instruments with appropriate norms must be used. For example, the Perkins-Binet is available for testing children who are blind. These tests are discussed in Topic 7B, Testing Persons with Disabilities. For obvious reasons, only the verbal portions of tests should be administered to sighted children with an uncorrected visual problem. Speech impairments present another problem for diagnosticians. The verbal responses of subjects with speech impairment are difficult to decipher. Owing to the failed comprehension of the examiner, subjects may receive less credit than is due. Sattler (1988) relates the lamentable case of Daniel Hoffman, a youngster with speech impairment who spent his entire youth in classes for those with mental retardation because his Stanford-Binet IQ was 74. In actuality, his intelligence was within the normal range, as revealed by other performance-based tests. In another tragic miscarriage of assessment, a patient in England was mistakenly confined to a ward for those with severe

retardation because cerebral palsy rendered his speech incomprehensible. The patient was wheelchair-bound and had almost no motor control, so his performance on nonverbal tests was also grossly impaired. The staff assumed he was severely retarded, so the patient remained on the back ward for decades. However, he befriended a fellow resident who could comprehend the patient’s gutteral rendition of the alphabet. The friend was severely retarded but could nonetheless recognize keys on a typewriter. With laborious letter-by-letter effort, the patient with incapacitating cerebral palsy wrote and published an autobiography, using his friend with mental disability as a conduit to the real world. Even if their disability is mild, persons with cerebral palsy or other motor impairments may be penalized by timed performance tests. When testing a person with a mild motor disability, examiners may wish to omit timed performance subtests or to discount these results if they are consistently lower than scores from untimed subtests. If a subject has an obvious motor

disability—such as a difficulty in manipulating the pieces of a puzzle—then standard instruments administered in the normal manner are largely inappropriate. A number of alternative instruments have been developed expressly for examinees with cerebral palsy and other motor impairments, and standard tests have been cleverly adapted and renormed (Topic 7B, Testing Persons with Disabilities). Desirable Procedures of Group Testing Psychologists and educators commonly assume that almost any adult can accurately administer group tests, so long as he or she has the requisite manual. Administering a group test would appear to be a simple and straightforward procedure of passing out forms and pencils, reading instructions, keeping time, and collecting the materials. In reality, conducting a group test requires as much finesse as administering an individual test, a point recognized years ago by Traxler (1951). There are numerous ways in which careless administration and scoring can impair group test

results, causing bias for the entire group or affecting only certain individuals. We outline only the more important inadequacies and errors in the following paragraphs, referring the reader to Traxler (1951) and Clemans (1971) for a more complete discussion. Undoubtedly the greatest single source of error in group test administration is incorrect timing of tests that require a time limit. Examiners must allot sufficient time for the entire testing process: setup, reading instructions out loud, and the actual test taking by examinees. Allotting sufficient time requires foresightful scheduling. For example, in many school settings, children must proceed to the next class at a designated time, regardless of ongoing activities. Inexperienced examiners might be tempted to cut short the designated time limit for a test so that the school schedule can be maintained. Of course, reduced time on a test renders the norms completely invalid and likely lowers the score for most subjects in the group. Allowing too much time for a test can be an equally egregious error. For example, consider

the impact of receiving extra time on the Miller Analogies Test (MAT), a high-level reasoning test once required by many universities for graduate school application. Since the MAT is a speeded test that requires quick analogical thinking, extra time would allow most examinees to solve several extra problems. This kind of testing error would likely lower the validity of the MAT results as a predictor of graduate school performance. A second source of error in group test administration is lack of clarity in the directions to the examinees. Examiners must read the instructions slowly in a clear, loud voice that commands the attention of the subjects. Instructions must not be paraphrased. Where allowed by the manual, examiners must stop and clarify points with individual examinees who are confused. Noise is another factor that must be controlled in group testing. It has been known for some time that noise causes a decrease in performance, especially for tasks of high complexity (e.g., Boggs & Simon, 1968).

Surprisingly, there is little research on the effects of noise on psychological tests. However, it seems almost certain that loud noise, especially if intermittent and unpredictable, will cause test scores to decline substantially. Elementary schoolchildren should not be expected to perform well while a construction worker jackhammers a cement wall in the next room. In fairness to the examinees, there are times when the test administrator should reschedule the test. Another source of error in the administration of a group test is failure to explain when and if examinees should guess. Perhaps more frequently than any other question, examiners are asked, “Is there a penalty if I guess wrong?” In most instances, test developers anticipate this issue and provide explicit guidance to subjects as to the advantages and/or pitfalls of guessing. Examiners should not give supplementary advice on guessing—this would constitute a serious deviation from standardized procedure. Most test developers incorporate a correction for guessing based on established principles of

probability. Consider a multiple-choice test that has four alternatives per item. On those items that the subject makes a wild, uneducated guess, the odds on being correct are 1 out of 4, while the odds on being wrong are 3 out of 4. Thus, for every three wrong guesses, there will be one correct guess that reflects luck rather than knowledge. Suppose a young girl answers correctly on 35 questions from a 50-item test but answers erroneously on 9 questions. In all, she has answered 44 questions, leaving 6 blank. The fact that she selected the wrong alternative on 9 questions suggests that she also gained 3 correct answers due to luck rather than knowledge. Remember, on wild guesses we expect there to be, on average, 3 wrong answers for every correct answer, so for 9 wrong guesses we would expect 3 correct guesses on other questions. The subject’s corrected score—the one actually reported and compared to existing norms—would then be 32; that is, 35 minus 3. In other words, she probably knew 32 answers but by guessing on 12 others she boosted her score another 3 points.

The scoring correction outlined in the preceding paragraph pertains only to wild, uneducated guesses. The effect of such a correction is to eliminate the advantage otherwise bestowed on unabashed risk takers. However, not all guesses are wild and uneducated. In some instances, an examinee can eliminate one or two of the alternatives, thereby increasing the odds of a correct guess among the remaining choices. In this situation, it may be wise for the examinee to guess. Whether an educated guess is really to the advantage of the examinee depends partly on the diabolical skill of the item writer. Traxler (1951) notes: In effect, the item writer attempts to make each

wrong response so plausible that every examinee who does not possess the desired skill or ability will select a wrong response. In other words, the item writer’s aim is to make all or nearly all considered guesses wrong guesses.

A skilled item writer can fashion questions so that the correct alternative is completely

counterintuitive and the wrong alternatives are persuasively appealing. For these items, an educated guess is almost always wrong. Nonetheless, many test developers now advise subjects to make educated guesses but warn against wild guesses. For example, a recent edition of the test preparation manual Taking the SAT advises: Because of the way the test is scored, haphazard

or random guessing for questions you know nothing about is unlikely to change your score. When you know that one or more choices can be eliminated, guessing from among the remaining choices should be to your advantage.

Whether or not a group test uses a scoring correction, the important point to emphasize in this context is that the administrator should follow standardized procedure and never offer supplementary advice about guessing. In group testing, deviations from the instructions manual are simply unacceptable.

1.9 INFLUENCE OF THE EXAMINER The Importance of Rapport Test publishers urge examiners to establish rapport—a comfortable, warm atmosphere that serves to motivate examinees and elicit cooperation. Initiating a cordial testing milieu is a crucial aspect of valid testing. A tester who fails to establish rapport may cause a subject to react with anxiety, passive-aggressive noncooperation, or open hostility. Failure to establish rapport distorts test findings: Ability is underestimated and personality is misjudged. Rapport is especially important in individual testing and particularly so when evaluating children. Wechsler (1974) has noted that establishing rapport places great demands on the clinical skills of the tester: To put the child at ease in his surroundings, the

examiner might engage him in some informal conversation before getting down to the more serious business of giving the

test. Talking to him about his hobbies or interests is often a good way of breaking the ice, although it may be better to encourage a shy child to talk about something concrete in the environment—a picture on the wall, an animal in his classroom, or a book or toy (not a test material) in the examining room. In general, this introductory period need not take more than 5 to 10 minutes, although the testing should not start until the child seems relaxed enough to give his maximum effort.

Testers may differ in their abilities to establish rapport. Cold testers will likely obtain less cooperation from their subjects, resulting in reduced performance on ability tests or distorted, defensive results on personality tests. Overly solicitous testers may err in the opposite direction, giving subtle (and occasionally blatant) cues to correct answers. Both extremes should be avoided. Examiner Sex, Experience, and Race

A wide body of research has sought to determine whether certain characteristics of the examiner cause examinee scores to be raised or lowered on ability tests. For example, does it matter whether the examiner is male or female? Experienced or novice? Same or different race from the examinee? We will contain the urge to review these studies—with a few exceptions— for one simple reason: The results are contradictory and, therefore, inconclusive. Most studies find that sex, experience, and race of the examiner make little, if any, difference. Furthermore, the few studies that report a large effect in one direction (e.g., female examiners elicit higher IQ scores) are contradicted by other studies showing the opposite trend. The interested reader can consult Sattler (1988) for a discussion and extensive listing of references. Yet, it would be unwise to conclude that sex, experience, or race of the examiner never affect test scores. In isolated instances, a particular examiner characteristic might very well have a large effect on examinee test scores. For example, Terrell, Terrell, and Taylor (1981)

ingeniously demonstrated that the race of the examiner interacts potently with the trust level of African American examinees in IQ testing. These researchers identified African American college students with high and low levels of mistrust of whites; half of each group was then administered the WAIS by a white examiner, the other half by an African American examiner. The high-mistrust group with an African American examiner scored significantly higher than the high-mistrust group with a white examiner (average IQs of 96 versus 86, respectively). In addition, the low-mistrust group with a white examiner scored slightly higher than the low-mistrust group with an African American examiner (average IQs of 97 versus 92, respectively). In sum, the authors concluded that mistrustful African Americans do poorly when tested by white examiners. Data bearing on this type of racial effect are meager, and there is certainly room for additional research.

1.10 BACKGROUND AND MOTIVATION OF THE EXAMINEE Examinees differ not only in the characteristics that examiners desire to assess but also in other extraneous ways that might confound the test results. For example, a bright subject might perform poorly on a speeded ability test because of test anxiety; a sane murderer might seek to appear mentally ill on a personality inventory to avoid prosecution; a student of average ability might undergo coaching to perform better on an aptitude test. Some subjects utterly lack motivation and don’t care if they do well on psychological tests. In all of these instances, the test results may be inaccurate because of the filtering and distorting effects of certain examinee characteristics such as anxiety, malingering, coaching, or cultural background. Test Anxiety Test anxiety refers to those phenomenological, physiological, and behavioral responses that

accompany concern about possible failure on a test. There is no doubt that subjects experience different levels of test anxiety ranging from a carefree outlook to incapacitating dread at the prospect of being tested. Several true-false questionnaires have been developed to assess individual differences in test anxiety (e.g., Lowe, Lee, Witteborg, & others, 2008; Spielberger, Gonzalez, Taylor, & others, 1980; Spielberger & Vagg, 1995). Following, we list characteristic items and their direction of keying (T for True, F for False): • (T) When taking an important examination,

I sweat a great deal. • (T) I freeze up when I take intelligence tests

or school exams. • (F) I really don’t understand why some

people get so upset about tests. • (T) I dread courses in which the instructor

likes to give “pop” quizzes. An extensive body of research has confirmed the commonsense notion that test anxiety is negatively correlated with school achievement, aptitude test scores, and measures of

intelligence (e.g., Chapell, Blanding, & Silverstein, 2005; Naveh-Benjamin, McKeachie, & Lin, 1987; Ortner & Caspers, 2011). However, the interpretation of these correlational findings is not straightforward. One possibility is that students develop test anxiety because of a history of performing poorly on tests. That is, the decrements in performance may precede and cause the test anxiety. In support of this viewpoint, Paulman and Kennelly (1984) found that—independent of their anxiety—many test-anxious students also display ineffective test taking in academic settings. Such students would do poorly on tests whether or not they were anxious. Moreover, Naveh-Benjamin et al. (1987) determined that a large proportion of test-anxious college students have poor study habits that predispose them to poor test performance. The test anxiety of these subjects is partly a by-product of lifelong frustration over mediocre test results. Other lines of research indicate that test anxiety has a directly detrimental effect on test performance. That is, test anxiety is likely both

cause and effect in the equation linking it with poor test performance. Consider the seminal study on this topic by Sarason (1961), who tested high- and low-anxious subjects under neutral or anxiety-inducing instructions. The subjects were college students required to memorize two-syllable words low in meaningfulness—a difficult task. Half of the subjects performed under neutral instructions— they were simply told to memorize the lists. The remaining subjects were told to memorize the lists and told that the task was an intelligence test. They were urged to perform as well as possible. The two groups did not differ significantly in performance when the instructions were neutral and non-threatening. However, when the instructions aroused anxiety, performance levels for the high-anxious subjects dropped markedly, leaving them at a huge disadvantage compared to low-anxious subjects. This indicates that test-anxious subjects show significant decrements in performance when they perceive the situation as a test. In contrast,

low-anxious subjects are relatively unaffected by such a simple redefinition of the context. Tests with narrow time limits pose a special problem to persons with high levels of test anxiety. Time pressure seems to exacerbate the degree of personal threat, causing significant reductions in the performance of test-anxious persons. Siegman (1956) demonstrated this point many years ago by comparing performance levels of high- and low-anxious medical/psychiatric patients on timed and untimed subtests from the WAIS. The WAIS consists of eleven subtests, including six subtests for which the examiner uses a stopwatch to enforce strict time limits, and five subtests for which the subject has unlimited time to respond. Interestingly, the high- and low-anxious subjects were of equal overall ability on the WAIS. However, each group excelled on different kinds of subtests in predictable directions. In particular, the low- anxious subjects surpassed the high-anxious subjects on timed subtests, whereas the reverse

pattern was observed on untimed subtests (Figure 1.4). Motivation to Deceive Test results also may be inaccurate if the examinee has reasons to perform in an inadequate or unrepresentative manner. Overt faking of test results is rare, but it does happen. A small fraction of persons seeking benefits from rehabilitation or social agencies will consciously fake bad on personality and ability tests. The topic of malingering (faking bad for personal gain) is discussed in a later chapter.

FIGURE 1.4 Influence of Timing and Anxiety Level on WAIS Subtest Results Source: Based on data from Siegman, A. W. (1956). The effect of manifest anxiety on a concept formation task, a nondirected learning task, and on timed and untimed intelligence tests. Journal of Consulting Psychology, 20, 176–178. TOPIC 1B Ethical and Social Implications of Testing 1.11 The Rationale for Professional Testing Standards Case Exhibit 1.2 Ethical and Professional Quandaries in Testing 1.12 Responsibilities of Test Publishers 1.13 Responsibilities of Test Users Case Exhibit 1.3 Overzealous Interpretation of the MMPI 1.14 Testing of Cultural and Linguistic Minorities 1.15 Unintended Effects of High-Stakes Testing 1.16 Reprise: Responsible Test Use

The general theme of this book is that psychological testing is a beneficial influence in modern society. When used ethically and responsibly, testing provides a basis for arriving at sensible inferences about individuals and groups. After all, the intention of the enterprise is to promote proper guidance, effective treatment, accurate evaluation, and fair decision making—whether in one-on-one clinic testing or institutional group testing. Who could possibly complain about these goals? Thankfully, tests generally are applied in an ethical and responsible manner by psychologists, educators, administrators, and others. But there are exceptions. Almost everyone has heard the horrific anecdotes: the minority grade schooler casually labeled as having mental retardation on the basis of a single IQ score; the college student implausibly diagnosed as schizophrenic from a projective test; the job applicant wrongfully screened from

employment based on an irrelevant measure; the aspiring teacher given unfair advantage when a competency test is mysteriously leaked beforehand; or the minority child penalized in testing because English is not her first language. Exceptions such as these illustrate the need for ethical and professional standards in testing. A major purpose of this topic is to introduce the reader to the ethical and professional standards that inform the practice of psychological testing. We also pursue the related theme of special considerations in the testing of cultural and linguistic minorities. The two topics share substantial overlap: When an examinee is not from the majority Anglo-American culture (predominantly Caucasian, English-speaking, individualistic, future-oriented), ethical and professional concerns in testing rise to the forefront. Finally, we examine a troubling and under- reported implication of widespread testing, namely, to the extent that society uses test results to make important decisions, the motivation for stakeholders to cheat is

intensified. As a result, cheating has emerged as a dark, unintended consequence of high-stakes testing, especially in the school systems of our nation.

1.11 THE RATIONALE FOR PROFESSIONAL TESTING STANDARDS Testing is generally applied in a responsible manner, but as previously noted, there are exceptions. On rare occasions, testing is irresponsible by design rather than by accident. Consider, with shuddering amazement, the advertisement for Mind Prober featured in a pop psychology magazine: Read Any Good Minds Lately? With the Mind

Prober you can. In just minutes you can have a scientifically accurate personality profile of anyone. This new expert systems software lets you discover the things most people are afraid to tell you. The strengths,

weaknesses, sexual interests and more. (Eyde & Primhoff, 1992)

In this case the irresponsibility is so blatant that discussion of ethical and professional guidelines is almost superfluous. However, testing practices do not always present in sharply contrasting shades, responsible or irresponsible. The real challenge of competent assessment is to determine the boundaries of ethical and professional practice. As usual, it is the borderline cases that provide pause for thought. The reader is encouraged to read the quandaries of testing described in Case Exhibit 1.2 and form an opinion about each. These examples are based on firsthand reports to the author. At the close of this chapter, we will return to these problematic vignettes. CASE EXHIBIT 1.2 Ethical and Professional Quandaries in Testing 1. A consulting psychologist agrees to perform

preemployment screening for psychopathology in police officer

candidates. At the beginning of each consultation, the psychologist asks the candidate to read and sign a detailed consent form that openly and honestly describes the evaluation process. However, the consent form explains that specific feedback about the test results will not be provided to job candidates. Question: Is it ethical for the psychologist to deny such feedback to the candidates?

2. A competent counselor who has received extensive training in the interpretation of the MMPI continues to use this instrument even though it has been superceded by the MMPI-2. His rationale is simply that there is a huge body of research on the MMPI and, he feels secure about the meaning of elevated MMPI test profiles, whereas he knows very little about the MMPI-2. He intends to switch over to the MMPI-2 at some undetermined future date, but finds no compelling reason to do so immediately. Question: Is the counselor’s refusal to use

the MMPI-2 a breach of professional standards?

3. A consulting psychologist is asked to evaluate a 9-year-old boy of Puerto Rican descent for possible learning disability. The child’s primary language is Spanish and his secondary language is English. The psychologist intends to use the Wechsler Intelligence Scale for Children-IV (WISC- IV) and other tests. Because he knows almost no Spanish, the psychologist asks the child’s after-school babysitter to act as translator when this is required to communicate test directions, specific questions, or the child’s responses. Question: Is it an appropriate practice to use a translator when administering an individual test such as the WISC-IV?

4. In the midst of taking a test battery for learning disability, a distraught 20-year-old female college student confides a terrifying secret to the psychologist. The client has just discovered that her 25-year-old brother, who died three months ago, was most likely a

pedophile. She shows the psychologist photographs of naked children posing in the brother’s bedroom. To complicate matters, the brother lived with his mother—who is still unaware of his well-concealed sexual deviancy. Question: Is the psychologist obligated to report this case to law enforcement?

The dilemmas of psychological testing do not always have simple, obvious answers. Even thoughtful and experienced psychologists may disagree as to what is ethical or professional in a given instance. Nonetheless, the scope of ethical and professional practice is not a matter of individual taste or personal judgment. Responsible test use is defined by written guidelines published by professional associations such as the American Psychological Association, the American Counseling Association, the National Association of School Psychologists, and other groups. Whether they know it or not, all practitioners owe allegiance to these guidelines, which we review in the following sections.

In general, the evolution of professional and ethical standards has been almost uniformly restrictive, providing an ever-narrowing demarcation of where, when, and how psychological tests may be used. Partly in response to the modern climate of litigation, organizations concerned with psychological testing have published guidelines that collectively define the ethical and professional standards relevant to the practice of assessment. These standards also pertain to corporations and individuals who publish tests. We begin with a survey of guidelines for test publishers before examining the responsibilities of test users. The chapter closes with a review of special concerns in the testing of cultural and linguistic minorities.

1.12 RESPONSIBILITIES OF TEST PUBLISHERS The responsibilities of publishers pertain to the publication, marketing, and distribution of their tests. In particular, it is expected that publishers

will release tests of high quality, market their product in a responsible manner, and restrict distribution of tests only to persons with proper qualifications. We consider each of these points in turn. Publication and Marketing Issues Regarding the publication of new or revised instruments, the most important guideline is to guard against premature release of a test. Testing is a noble enterprise but it is also big business driven by the profit motive, which provides an inherent pressure toward early release of new or revised materials. Perhaps this is why the American Psychological Association and other organizations have published standards that relate to test publication (AERA/ APA/NCME, 1999). These standards pertain especially to the technical manuals and user guides that typically accompany a test. These sources must be sufficiently complete so that a qualified user or reviewer can evaluate the appropriateness and technical adequacy of the test. This means that manuals and guides will

report detailed statistics on reliability analyses, validity studies, normative samples, and other technical aspects. Marketing tests in a responsible manner refers not only to advertising (which should be accurate and dignified) but also to the way in which information is portrayed in manuals and guides. In particular, test authors should strive for a balanced presentation of their instruments and refrain from a one-sided presentation of information. For example, if some preliminary studies reflect poorly on a test, these should be given fair weight in the manual alongside positive findings. Likewise, if a potential misuse or inappropriate use of a test can be anticipated, the test author needs to discuss this matter as well. Competence of Test Purchasers Test publishers recognize the broad responsibility that only qualified users should be able to purchase their products. By way of brief review, the reasons for restricted access include the potential for harm if tests fall into the wrong

hands (e.g., an undergraduate psychology major administers the MMPI-2 to his friends and then makes frightful pronouncements about the results) and the obvious fact that many tests are no longer valid if potential examinees have previewed them (e.g., a teacher memorizes the correct answers to a certification exam). These examples illustrate that access to psychological tests needs to be limited. But limited to whom? The answer, it turns out, depends on the complexity of the specific test under consideration. Guidelines proposed many years ago by the American Psychological Association (APA, 1953) are still relevant today, even though they are not enforced by all publishers. The APA proposed that tests fall into three levels of complexity (Levels A, B, and C) that require different degrees of expertise from the examiner. Level A comprised simple paper- and-pencil tests that require minimal training. These can be used by responsible nonpsychologists such as educational administrators. Examples include group educational tests and vocational proficiency

scales. Level B tests require training in statistics and knowledge of test construction. Some graduate training is needed. This group includes aptitude tests and personality inventories relevant to normal populations. Level C includes the most complex instruments. Minimum training required is a master’s degree in psychology or a related field. Instruments include projective personality tests, individual tests of intelligence, and neuropsychological test batteries. In general, test publishers try to screen out inappropriate requests by requiring that purchasers have the necessary credentials. For example, the Psychological Corporation, one of the major suppliers of test materials in the United States, requires prospective customers to fill out a registration form detailing their training and experience with tests. Buyers who do not hold an advanced degree in psychology must list details of courses in the administration and interpretation of tests and in statistics. References are required, too.

Most test publishers also specify that individuals or groups who provide testing and counseling by mail are not allowed to purchase materials. On a related note, ethical standards now discourage practitioners from giving “take- home” tests to clients. Until recent years, this has been an occasional practice with lengthy personality tests such as the MMPI. The ethics committee endorsed the following point: Nonmonitored administration of the MMPI

generally does not represent sound testing practice and may result in invalid assessment for a variety of reasons (e.g., influence from other people or completion of the test while intoxicated).

In general, users are advised to refrain from giving take-home tests and publishers are counseled to deny access to practitioners or groups who promote this practice. Even though publishers attempt to filter out unqualified purchasers, there may still be instances in which sensitive tests are sold to unscrupulous individuals. Oles and Davis (1977) discovered that graduate students in

psychology could purchase the WISC-R, MMPI, TAT, Stanford-Binet, and 16PF if they typed their orders on college stationery, placed the letters Ph.D. after their names, enclosed payment, and used a post office box return address. Although illicit test orders are few in number, they do occur.

1.13 RESPONSIBILITIES OF TEST USERS The psychological assessment of personality, interests, brain functioning, aptitude, or intelligence is a sensitive professional action that should be completed with utmost concern for the well-being of the examinee, his or her family, employers, and the wider network of social institutions that might be affected by the results of that particular clinical assessment (Matarazzo, 1990). Over the years, the profession of psychology has proposed, clarified, and sharpened a series of thorough and thoughtful standards to provide guidance for the individual practitioner. Professional

organizations publish formal ethical principles that bear upon test use, including the American Psychological Association (APA, 2002), the American Association for Counseling and Development (AACD, 1988), the American Speech-Language-Hearing Association (ASHA, 1991), and the National Association of School Psychologists (NASP, 2010). In addition to ethical principles, several testing organizations have published practice guidelines to help define the scope of responsible test use. Sources of test use guidelines include teaching groups (AFT, NCME, NEA, 1990), the American Psychological Association (APA, 1992b), the Educational Testing Service (ETS, 1989), the Joint Committee on Testing Practices (JCTP, 1988), the Society for Industrial and Organizational Psychology (SIOP, 1987), and professional alliances (AERA, APA, NCME, 1999). Finally, we should mention that the principles of responsible test use have been distilled in an illuminating casebook published jointly by several testing groups (Eyde, Robertson, & Krug, 2009).

The dozens of guidelines relevant to testing are quite specific, for example: Standard 5.9: When test score information is

released to students, parents, legal representatives, teachers, clients, or the media, those responsible for testing programs should provide appropriate interpretations. The interpretations should describe in simple language what the test covers, what scores mean, the precision of the scores, common misinterpretations of test scores, and how scores will be used.

Because of their specificity, a detailed analysis of relevant ethical and professional standards is beyond the scope of this text. What follows is a summary of the general provisions that pertain to the responsible practice of psychological testing and clinical psychological assessment. These principles apply to psychologists, students of psychology, and others who work under the supervision of a psychologist. We restrict our discussion to those principles that are directly pertinent to the practice of psychological testing. Proper adherence to these

principles would eliminate most—but not all— legal challenges to testing. Best Interests of the Client Several ethical principles recognize that all psychological services, including assessment, are provided within the context of a professional relationship. Psychologists are, therefore, enjoined to accept the responsibility implicit in this relationship. In general, the practitioner is guided by one overriding question: What is in the best interests of the client? The functional implication of this guideline is that assessment should serve a constructive purpose for the individual examinee. If it does not, the practitioner is probably violating one or more specific ethical principles. For example, Standard 11.15 in the Standards manual (AERA, APA, NCME, 1999) warns testers to avoid actions that have unintended negative consequences. Allowing a client to attach unsupported surplus meanings to test results would not be in the best interests of the client and would, therefore, constitute an unethical

testing practice. In fact, with certain worry- prone and self-doubting clients, a psychologist may choose not to use an appropriate test, since these clients are almost certain to engage in self- destructive misinterpretation of virtually any test findings. Confidentiality and the Duty to Warn Practitioners have a primary obligation to safeguard the confidentiality of information, including test results, that they obtain from clients in the course of consultations (Principle 5; APA, 1992a). Such information can be ethically released to others only after the client or a legal representative gives unambiguous consent, usually in written form. The only exceptions to confidentiality involve those unusual circumstances in which the withholding of information would present a clear danger to the client or other persons. For example, most states have passed laws that mandate that health care practitioners must report all cases of suspected abuse in children and vulnerable elderly persons. In most states, a psychologist

who learns in the course of testing that the client has physically or sexually abused a child is obligated to report that information to law enforcement. Psychologists also have a duty to warn that stems from the 1976 decision in the Tarasoff case (Wrightsman, Nietzel, Fortune, & Greene, 2002). Tanya Tarasoff was a young college student in California who was murdered by Prosenjit Poddar, a student from India. What makes the case relevant to the practice of psychology is that Poddar had made death threats regarding Tarasoff to his campus-based therapist. Although the therapist warned the police that Poddar had made death threats, he did not warn Tarasoff. Two months later, Poddar stabbed Tarasoff to death at her home. The parents of Tanya Tarasoff sued, and the California Supreme Court later agreed that therapists have a duty to use “reasonable care” to protect potential victims from their clients. Although the Tarasoff ruling has been modified by legislation in many states, the thrust of the case still stands: Clinicians must communicate

any serious threat to the potential victim, law enforcement agencies, or both. Finally, the clinician should consider the client’s welfare in deciding whether to release information, especially when the client is a minor who is unable to give voluntary, informed consent. When appropriate, practitioners are advised to inform their clients of the legal limits of confidentiality. Expertise of the Test User A number of principles acknowledge that the test user must accept ultimate responsibility for the proper application of tests. From a practical standpoint, this means that the test user must be well trained in assessment and measurement theory. The user must possess the expertise needed to evaluate psychological tests for proper standardization, reliability, validity, interpretive accuracy, and other psychometric characteristics. This guideline has special significance in areas such as job screening, special education, testing of persons with

disabilities, or other situations in which potential impact is strong. Psychologists who are poorly trained in their chosen instruments can make serious errors of test interpretation that harm examinees. Furthermore, inept test usage may expose the examiner to professional sanctions and civil lawsuits. A common error observed among inexperienced test users is the overzealous, pathologized interpretation of personality test results (Case Exhibit 1.3). CASE EXHIBIT 1.3 Overzealous Interpretation of the MMPI An inexperienced consulting psychologist routinely used the MMPI for preemployment screening of law enforcement candidates. One candidate subsequently filed a lawsuit, alleging that she had been harmed by the psychologist’s report. The plaintiff, a young woman with extensive training and background in law enforcement, was denied a position as police officer because of a supposedly “defensive” MMPI profile. Her profile was entirely within

normal limits, although she did obtain a T score of 72 on the K scale. The K scale is usually considered a good index of defensive test-taking attitudes, especially for mental health evaluations with clinic or hospital referrals. By way of quick review, MMPI T scores of approximately 50 are average, whereas elevations of 70 or higher are considered noteworthy. The consulting psychologist noticed the candidate’s elevated score on the K scale, surmised hastily that the candidate was unduly defensive, and cautioned the police chief not to hire her. What the psychologist did not know is that elevated K-scale scores are extremely common among law enforcement job applicants. For example, Hiatt and Hargrave (1988) found that about 25 percent of a sample of peace officers produced MMPI profiles with K scales at or above a T score of 70. In fact, successful police officers tend to have higher K-scale scores than “problem” peace officers! In this case the test user did not possess sufficient expertise to use the MMPI for job screening. His ignorance on

this point constituted a breach of professional ethics. Incidentally, the case was settled out of court for a substantial sum of money, showing that trespasses of responsible test use can have serious legal consequences. The expertise of the psychologist is particularly relevant when test scoring and interpretation services are used. The Ethical Principles of the American Psychological Association leave no room for doubt: Psychologists retain appropriate responsibility

for the appropriate application, interpretation, and use of assessment instruments, whether they score and interpret such tests themselves or use automated or other services. (APA, 1992a)

The reader is referred to Topic 12B, Computerized Assessment and the Future of Testing, for further discussion of this point. Informed Consent Before testing commences, the test user needs to obtain informed consent from test takers or their legal representatives. Exceptions to informed

consent can be made in certain instances, for example, legally mandated statewide testing programs, school-based group testing, and when consent is clearly implied (e.g., college admissions testing). The principle of informed consent is so important that the Standards manual devotes a separate standard to it: Informed consent implies that the test takers or

representatives are made aware, in language that they can understand, of the reasons for testing, the type of tests to be used, the intended use and the range of material consequences of the intended use. If written, video, or audio records are made of the testing session, or other records are kept, test takers are entitled to know what testing information will be released and to whom. (AERA et al., 1999)

Even young children or test takers with limited intelligence deserve an explanation of the reasons for assessment. For example, the examiner might explain, “I’m going to ask you some questions and have you work on some

puzzles so I can see what you can do and find out what things you need more help with.” From a legal standpoint, the three elements of informed consent include disclosure, competency, and voluntariness (Melton, Petrila, Poythress, & Slobogin, 1998). The heart of disclosure is that the client receive sufficient information (e.g., about risks, benefits, release of reports) to make a thoughtful decision about continued participation in the testing. Competency refers to the mental capacity of the examinee to provide consent. In general, there is a presumption of competency unless the examinee is a child, very elderly, or has mental disabilities (e.g., has mental retardation). In these cases, a guardian will need to provide legal consent. Finally, the standard of voluntariness implies that the choice to undergo an assessment battery is given freely and not based on subtle coercion (e.g., inmates are promised release time if they participate in research testing). In most cases, the examiner uses a written informed consent form such as that found in Figure 1.5.

FIGURE 1.5 Abbreviated Example of Informed Consent for Psychological Assessment Note: This form is illustrative only. Practitioners should consult legal counsel in regard to the details of an informed consent form. Obsolete Tests and the Standard of Care Standard of care is a loose concept that often arises in the professional or legal review of specific health practices, including psychological testing. The prevailing standard

of care is one that is “usual, customary or reasonable” (Rinas & Clyne-Jackson, 1988). To cite an extreme example, in medicine the standard of care for a fever might include the administration of aspirin—but would not include the antiquated practice of bleeding the patient. Practitioners of psychological testing must be wary of obsolete tests, because their use might violate the prevailing standard of care. A case in point is the MMPI versus the MMPI-2. Even though the MMPI-2 is a relatively conservative revision of the highly esteemed MMPI, the improvements in norming and scale construction are substantial. The MMPI-2 is now the standard of care in MMPI-based assessment of psychopathology. Practitioners who continue to rely on the original MMPI could be liable for malpractice suits, especially if the test interpretation resulted in misleading interpretive statements or an incorrect diagnosis. Another concern relevant to the standard of care is reliance on test results that are outdated for the current purpose. After all, individual

characteristics and traits show valid change over time. A student who meets the criteria for learning disability (LD) in the fourth grade might show large gains in academic achievement, such that the LD diagnosis is no longer accurate in the fifth grade. Personality test results are especially prone to quixotic change. A short-term personal crisis might cause an MMPI-2 profile to look like a range of mountains. A week later, the test profile could be completely normal. It is difficult to provide comprehensive guidelines as to the “shelf life” of psychological test results. For example, GRE test scores that are years old still might be validly predictive of performance in graduate school, whereas Beck Depression Inventory test results from yesterday could mislead a therapist as to the current level of depression. Practitioners must evaluate the need for retesting on an individual basis. Responsible Report Writing Except for group testing, the practice of psychological testing invariably culminates in a

written report that constitutes a semipermanent record of test findings and examiner recommendations. Effective report writing is an important skill because of the potential lasting impact of the written document. It is beyond the scope of this text to illuminate the qualities of effective report writing, although we can refer the reader to a few sources (Gregory, 1999; Tallent, 1993). Responsible reports typically use simple and direct writing that steers clear of jargon and technical terms. The proper goal of a report is to provide helpful perspectives on the client, not to impress the referral source that the examiner is a learned person! When Tallent (1993) surveyed more than one thousand health practitioners who made referrals for testing, one respondent declared his disdain toward psychologists who “reflect their needs to shine as a psychoanalytic beacon in revealing the dark, deep secrets they have observed.” On a related note, effective reports stay within the bounds of expertise of the examiner. For example:

It is never appropriate for a psychologist to recommend that a client undergo a specific medical procedure (such as a CT scan for an apparent brain tumor) or receive a particular drug (such as Prozac for depression). Even when the need for a special procedure seems obvious (e.g., the symptoms strongly attest to the rapid onset of a brain disease), the best way to meet the needs of the client is to recommend immediate consultation with the appropriate medical profession (e.g., neurology or psychiatry). (Gregory, 1999)

Additional advice on effective report writing can be found in Ownby (1991) and Sattler (2001). Communication of Test Results Individuals who take psychological tests anticipate that the results will be shared with them. Yet practitioners often do not include one- to-one feedback as part of the assessment. A major reason for reluctance is a lack of training in how to provide feedback, especially when the

test results appear to be negative. For example, how does a clinician tell a college student that her IQ is 93 when most students in that milieu score 115 or higher? Providing effective and constructive feedback to clients about their test results is a challenging skill to learn. Pope (1992) emphasizes the responsibility of the clinician to determine that the client has understood adequately and accurately the information that the clinician was attempting to convey. Furthermore, it is the responsibility of the clinician to check for adverse reactions: Is the client exceptionally depressed by the

findings? Is the client inferring from findings suggesting a learning disorder that the client—as the client has always suspected—is “stupid”? Using scrupulous care to conduct this assessment of the client’s understanding of and reactions to the feedback is no less important than using adequate care in administering standardized psychological tests; test administration and feedback are equally

important, fundamental aspects of the assessment process. (p. 271)

Proper and effective feedback involves give- and-take dialogue in which the clinician ascertains how the client has perceived the information and seeks to correct potentially harmful interpretations. Destructive feedback often arises when the clinician fails to challenge a client’s incorrect perceptions about the meaning of test results. Consider IQ tests in particular—a case in which many persons deify test scores and consider them an index of personal worth. Prior to providing test results, a clinician is advised to investigate the client’s understanding of what IQ scores mean. After all, IQ is a limited slice of intellectual functioning: It does not evaluate drive or character of any kind, it is accurate only to about ±5 points, it may change over time, and it does not assess many important attributes such as creativity, social intelligence, musical ability, or athletic skill. But a client may have an unrealistic perspective about IQ and, hence, might jump to erroneous conclusions when

hearing that her score is “only” 93. The careful practitioner will elicit the client’s views and challenge them when needed before proceeding. Further thoughts on feedback can be found in Pope (1992). Going beyond the general pronouncement to avoid harm when providing test feedback, Finn and Tonsager (1997) present the intriguing view that information about test results should be directly and immediately therapeutic to individuals experiencing psychological problems. In other words, they propose that psychological assessment is a form of short- term intervention, not just a basis for gathering information that is later used for therapeutic purposes. In one study (Finn & Tonsager, 1992), they examined the effects of a brief psychological assessment on clients at a university counseling center. Thirty-two students took part in an initial interview, completed the MMPI-2, and then received a one-hour feedback session conducted according to a method developed by Finn (1996). A comparison group of 29 students was

interviewed and received an equal amount of supportive, nondirective psychotherapy instead of the test feedback. The clients in the MMPI-2 assessment group showed a greater decline in symptomatic distress and a greater increase in self-esteem, immediately following their feedback session and also two weeks later, than the clients in the comparison group. The feedback group also felt more hopeful about their problems after the brief assessment. These findings illustrate the importance of providing thoughtful and constructive test feedback instead of rushing through a perfunctory review of the results. Consideration of Individual Differences Knowledge of and respect for individual differences is highlighted by all professional organizations that deal with psychological testing. The American Psychological Association lists this as one of six guiding principles: Principle D: Respect for People’s Rights and

Dignity . . . Psychologists are aware of

cultural, individual, and role differences, including those due to age, gender, race, ethnicity, national origin, religion, sexual orientation, disability, language, and socio- economic status. Psychologists try to eliminate the effect on their work of biases based on those factors, and they do not knowingly participate in or condone unfair discriminatory practices. (APA, 1992a)

The relevance of this principle to psychological testing is that practitioners are expected to know when a test or interpretation may not be applicable because of factors such as age, gender, race, ethnicity, national origin, religion, sexual orientation, disability, language, and socioeconomic status. We can illustrate this point with a case study reported in Eyde et al. (1993). A psychologist evaluated a 75-year-old man at the request of his wife, who had noticed memory problems. The psychologist administered a mental status examination and a prominent intelligence test. Performance on the mental status examination was normal, but standard scores on the intelligence test revealed

a large discrepancy between verbal subtests and subtests measuring spatial ability and processing speed. The psychologist interpreted this pattern as indicating a deterioration of intellectual functioning in the husband. Unfortunately, this interpretation was based on faulty use of non- age-corrected standard scores. Also, the psychologist did not assess for depression, which is known to cause visuospatial performance to drop sharply (Wolff & Gregory, 1992). In fact, a series of further evaluations revealed that the husband was a perfectly healthy 75-year-old man. The psychologist failed to consider the relevance of the gentleman’s age and emotional status when interpreting the intelligence test. This was a costly oversight that caused the client and his wife substantial unnecessary worry.

1.14 TESTING OF CULTURAL AND LINGUISTIC MINORITIES

Background and Historical Notes Persons of ethnic minority descent (non- European origin) currently constitute about a third of the U.S. population, and it is estimated that they will comprise more than 50 percent within several decades. Yet the enterprise of testing is based almost entirely on the efforts of white psychologists who bring an Anglo- American viewpoint to their work. The suitability of existing tests for the evaluation of diverse populations cannot be taken for granted. The assessment of ethnic minority individuals raises important questions, especially when test results translate to placement decisions or other sensitive outcomes, as is commonly the case within educational institutions. Unfortunately, the early pioneers in the testing movement largely ignored the impact of cultural background on test results. For example, in the 1920s Henry Goddard concluded that the intelligence of the average immigrant was alarmingly low, “perhaps of moron grade.” Yet he downplayed the likelihood that language and

cultural differences could explain the low test scores of immigrants. Goddard’s role in the history of testing is discussed in the next chapter. Perhaps as a rebound against these early methods, beginning in the 1930s psychologists displayed an increased sensitivity to cultural variables in the practice of testing. A shining example in this regard was Stanley Porteus, who undertook a wide-ranging investigation of the temperament and intelligence of Australian aboriginal peoples. Porteus (1931) used many traditional instruments (block designs, mazes, digit span), but to his credit he also devised an ecologically valid measure of intelligence for this group, namely, footprint recognition. Whereas the aboriginal examinees performed poorly on the Eurocentric tests, their ability to recognize photographed footprints was on a par with other racial groups studied. Even so, Porteus displayed an acute awareness that his procedures still might have handicapped the aboriginals:

The photograph of a footprint is not the same as the footprint itself, and quite probably a number of cues that are made use of by the aboriginal tracker are absent from a photograph. The varying depths of parts of the foot impression are not visible in the photograph, and the individual peculiarities other than general shape and size of the footprint may not be brought out clearly. Hence we must expect that the aboriginal subjects would be under some disadvantage in matching these photographs of footprints, as against recognition of the footprints themselves. (pp. 399–400)

In a similar vein, DuBois (1939) found that Pueblo Indian children displayed superior ability on his specially devised horse drawing test of mental ability, whereas they performed less well on the mainstream Goodenough (1926) Draw-A-Man test. From these early studies onward, psychologists have maintained a keen interest in the impact of language and culture on the meaning of test results.

The Impact of Cultural Background on Test Results Practitioners need to appreciate that the cultural background of examinees will impact the entire process of assessment. For this reason, Sattler (1988) advises assessment psychologists to approach their task from a pluralistic standpoint: Cultural groups may vary with respect to

cultural values (stemming in part from cultural shock, discontinuity, or conflict); language and nuances in language style; views of life and death; roles of family members; problem-solving strategies; attitudes toward education, mental health, and mental illness; and stage of acculturation (the group may follow traditional values, accept the dominant group’s values, or be at some point between the two). You should adopt a frame of reference that will enable you to understand how particular behaviors make sense within each culture. (p. 505)

For example, it is often noted that Native Americans display a distinctive conception of time, emphasizing present-time as opposed to the future-time orientation that is so powerfully formative in white, middle-class America (Panigua, 1994). A possible implication of this cultural difference is that time limits might not mean the same thing for a Native American child as for a child from the mainstream culture. Perhaps the minority child will disregard the subtest instructions and work at a careful, measured pace rather than seeking quick solutions. Of course, this child would then obtain a misleadingly low score on that measure. While acknowledging the impact of cultural differences on testing, it is also important to avoid stereotypical overgeneralization. Culture is not monolithic. Every person is unique. Some Native Americans will exhibit a distinctive orientation to time but perhaps most will not. The challenge for the practitioner is to observe the clinical details of performance and to

identify the culture-based nuances of behavior that help determine the test results. An ingenious study by Moore (1986) powerfully illustrates the relevance of cultural background for understanding the test performance of ethnic minority examinees. She compared not only the intelligence test scores but also the qualitative manner of responding to test demands in two groups of adopted African American children. One group of 23 children had been transracially adopted into middle-class white families. The other group of 23 children had been intraracially adopted into middle-class African American families. All children were adopted prior to age 2 and the backgrounds of the adoptive families were similar in terms of education and social class. Thus, group difference in test scores and test behaviors could be attributed mainly to differences in cultural background arising from the fact that one group was adopted into African American families, the other adopted into white families. Testing and observations were completed by two female African American examiners who were “blind” to the purposes of

the study. Tested at 7 to 10 years of age, the transracially adopted children scored an average IQ of 117 on the WISC compared to an average IQ of 104 for the traditionally adopted children. These IQ results were not remarkable, insofar as Scarr and Weinberg reported similar findings years before. The surprising and informative outcome of the study was that the two groups of children showed very different qualitative behaviors during testing. As a group, the children with lower IQ scores (those adopted by African American families) were less likely to spontaneously elaborate on their work responses and more likely simply to refuse to respond when presented with a test demand. Moore (1986) offers the following interpretations: Children’s tendency to spontaneously elaborate

on their work responses may be a very important index of their level of involvement in task performance, strategies for problem solving, level of motivation to generate a correct response, and level of adjustment to the standardized test

situation. . . . Although the terminal not- work response is treated as an incorrect response, it does not actually provide any empirical documentation of what the child does or does not know or of what the child can and cannot do. The only information available is that the child did not respond to the demand. (p. 322)

The essential lesson of this study is that culturally based differences in response style may function to conceal the underlying competence of some examinees. Cautious interpretation of test results is always advisable, but this is especially important for examinees from culturally or linguistically diverse backgrounds. The influence of cultural factors is not limited to the test performance of children but extends to adults as well. Terrell, Terrell, and Taylor (1981) investigated the effects of racial trust/mistrust on the intelligence test scores of African American college students. They identified African American students with high and low levels of mistrust of whites. Using a 2 × 2

design, half of each group was then administered an individual intelligence test by a white examiner, the other half by an African American examiner. As predicted, the analysis of variance revealed no differences for the main effects of race of examiner (white versus African American) or level of mistrust (high versus low) (Figure 1.6). But a substantial interaction was revealed; namely, the high- mistrust group with an African American examiner scored much better than the high- mistrust group with a white examiner (average IQs of 96 versus 86, respectively). Put simply, cultural mistrust among African Americans was associated with significantly lower IQ scores, but only when the examiner was white. Further illustrating cultural influences, Steele (1997) has proposed a theory that societal stereotypes about groups influence the immediate intellectual performance and also the long-term identity development of individual group members. He has applied this theory both to women—when stereotypes affect their achievement in math and sciences—and to

African Americans—when stereotypes apparently depress their performance on standardized tests. Here we discuss his research on stereotype threat with African American college students (Steele & Aronson, 1995).

FIGURE 1.6 Mean IQ Scores of African American Students as a Function of Race of Examiner and Cultural Mistrust Source: Based on data in Terrell, F., Terrell, S., & Taylor, J. (1981). Effects of race of examiner and

cultural mistrust on the WAIS performance of Black students. Journal of Consulting and Clinical Psychology, 49, 750–751. The idea of stereotype threat is essentially a sophisticated version of a self-fulfilling prophecy. The researchers define stereotype threat as the threat of confirming, as self- characteristic, a negative stereotype about one’s group. For example, based on published data and media coverage about race and IQ scores, African Americans are stereotyped as possessing less intellectual ability than others. As a consequence, whenever they encounter tests of intelligence or academic achievement, individuals from this group may perceive a risk that they will confirm the stereotype. In the short run, stereotype threat is hypothesized to depress test performance through heightened anxiety and other mechanisms. In the long run, it may have the further impact of pressuring African American students to “protectively disidentify” with achievement in school and related intellectual domains.

Steele and Aronson (1995) conducted a series of four studies to evaluate the hypothesis of stereotype threat. All the investigations supported the hypothesis. We focus here on the first study, in which African American and white college students were given a 30-minute test composed of challenging items from the verbal Graduate Record Examination. Students from both racial groups were randomly assigned to one of three test conditions: stereotype-threat, in which the test was described as diagnostic of individual verbal ability; control, in which the test was described as a research tool only; and control-challenge, in which the test was described as a research tool only but participants were exhorted to “take this challenge seriously.” Scores on the verbal test were adjusted (covariate analysis) on the basis of prior achievement scores so as to eliminate the effects of preexisting differences between groups. Race differences were small and nonsignificant in the control and control-challenge conditions, whereas African Americans scored much lower than whites in the stereotype-threat condition

(Figure 1.7). In other studies, Steele and Aronson (1995) investigated the mechanism of mediation by which stereotype threat caused African Americans to score lower on standardized tests. The details are beyond the scope of this text, but the overall conclusion is not: Our best assessment is that stereotype threat

caused an inefficiency of processing much like that caused by other evaluative pressures. Stereotype-threatened participants spent more time doing fewer items more inaccurately—probably as a result of alternating their attention between trying to answer the items and trying to assess the self-significance of their frustration. (Steele & Aronson, 1995, p. 809)

FIGURE 1.7 Average Verbal Items Correct for Whites and African Americans under Three Conditions Source: Based on data in Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of African Americans. Journal of Personality and Social Psychology, 69, 797–811. In sum, the authors propose a social- psychological perspective on the meaning of

lower test scores in African Americans and perhaps other stereotype-threatened groups as well. Their viewpoint emphasizes that test results do not reside within individuals. Test scores occur within a complex social- psychological field that is potentially influenced by national history, predicaments of race, and many other subtle factors.

1.15 UNINTENDED EFFECTS OF HIGH-STAKES TESTING The prevailing view in the general public is that cheating rarely or never occurs in nationally administered testing programs. We tend to think that the risks are too high and the opportunities too limited for cheaters to prevail. Therefore, we rest assured that test fraud must be a rare event. Unfortunately, this view is probably naive. After all, a growing number of people must pass a test to gain college entry, get a job, or obtain a promotion. Furthermore, school officials increasingly are evaluated on the basis of average test scores in their district. Precisely

because the stakes are so high, unscrupulous individuals will try to beat the system. Widespread cheating in public school systems is sporadically reported in many large cities across the United States. In most cases, the cheating is motivated by the desire of teachers and principals to further their own careers by creating the illusion of educational excellence. For example, in 1999, dozens of teachers and two principals in the New York City public school system were charged with helping students cheat on the standardized reading and math tests used to rank schools and determine whether students move on to the next grade (New York Times, December 12, 1999). The cheating scheme was described as “one of the largest in the recent history of American public schools.” In 2000, an entire eighth-grade class in a Chicago elementary school was required to retake the Iowa Tests of Basic Skills (ITBS) because a school administrator allegedly filled in incomplete tests and changed incorrect answers to correct ones (Chicago Tribune, June 2, 2000). Officials were tipped off to the fraud

because the test scores were simply too good to be true—the average score for the class was two years above their standing. In 2005, the Dallas Morning News reported strong evidence of “organized, educator-led cheating” in dozens of schools on the statewide achievement test and found suspicious scores in hundreds more (www.dallasnews.com, March 21, 2005). Disturbingly, one assessment expert noted, “You’re catching the dumb cheaters. The smart cheaters you’re not going to be able to detect.” We only read about the cases of cheating that are detected. The number of undetected cases is simply unknown, although probably larger than the public would like to believe. Cheating in public school systems is not a thing of the past. It continues unabated, year after year. In 2011, a decade long cheating scandal was revealed in the Atlanta, Georgia, public school system (Atlanta Journal-Constitution, July 6, 2011). Teachers and principals routinely changed students’ answer sheets to produce higher scores. The school system scores soared dramatically, bringing national acclaim to the

district and the superintendent. But it was all based on fraud perpetrated by 178 educators, including 38 principals. Cheating was confirmed in 44 of 56 schools examined. In 2011, six charter schools in Los Angeles were threatened with closure when it was discovered that the founding director had ordered principals to open the state standardized tests and train students on actual test questions (Los Angeles Times, June 22, 2011). Suspiciously, scores for the schools had vaulted upward in recent years. The director and the six principals were terminated. An especially flagrant instance of cheating on national tests was uncovered in Louisiana in 1997. This case involved wholesale circulation of the Educational Testing Service (ETS) exam administered to teachers who want to be school principals. As reported in the New York Times (September 28, 1997), copies of the 145-item test, along with correct answers, had circulated among teachers throughout southern Louisiana, most likely for several years. In a state ranked at or near the bottom on nearly every educational index, it appears that many potentially

unqualified persons cheated their way into running the schools. ETS handled this case quietly by asking more than 200 teachers to retake the test so as to “confirm” their initial scores. Unfortunately, the Louisiana case was not an isolated instance. In another case, ETS allegedly failed to monitor its handling of the federal government’s test for immigrants who want to become citizens, with the likely result that test supervisors accepted bribes. English- proficiency tests for foreign students also were vulnerable to cheating. In 1994, ETS canceled the scores of 30,000 students from China after discovering a ring that was selling the examinations abroad. Cizek (1999) catalogues literally dozens of ingenious ways that students have developed for cheating on tests: writing information on the floor, in tissues, on the back of a bottled water label; using an ultraviolet pen to write information on “blank” paper; and using a video transmitter (e.g., hidden in an eyeglass case) to send pictures of the test to an outside accomplice who then coaches the

student by means of an audio receiver (e.g., hidden in the ear). Stories about miniature transmitters are not fanciful. Consider the following story reported from a monolithic culture where test results literally make or break a child’s future. In China, 10 million 18-year-olds take a two day exam each year that determines whether they will be allowed to attend public universities. Success or failure drastically impacts their lives and those of their families who might depend on their future income. In 2009, eight parents were jailed for up to three years after it was determined that they were transmitting stolen test answers to their children through miniature earpieces. The subterfuge was discovered when police detected unusual radio signals near the school (www.guardian.co.uk, April 3, 2009). In 2012, cheating was brought to light on the board certification test for radiology (CNN, Prescription for Cheating, January 13, 2012). For years, doctors around the country have helped one another cheat by each memorizing one or two test questions verbatim, writing

down the questions after taking the test, and circulating the ever-expanding list of questions (dubbed “recalls”) to cooperating programs. The practice is so widespread and considered so egregious that the American Board of Radiology released a sternly worded video condemning the use of recalls as unethical. CNN found at least 15 years’ worth of test questions (with answers) on a website for residents in radiology. Recently, efforts to circumvent exam security have become even more brazen, with some test preparation companies encouraging students to steal copies of college entrance exams such as the Scholastic Assessment Tests (SAT) (Los Angeles Times, October 12, 2005). Fortunately, the publisher of the SAT was granted a restraining order in federal court, prohibiting individuals or companies from soliciting stolen copies of the test. Even so, this episode illustrates once again that high-stakes testing has had a corrupting influence on the testing process. Dishonest and inappropriate practices by school officials are implicated in the recent inflation of

scores on nationally normed group tests of achievement. By definition, for a norm- referenced test, 50 percent of the examinees should score above the 50th percentile, 50 percent below. If the same test is used in a large sample of typical and representative school systems, average scores for the school systems should be split evenly—about half above the nationally normed 50th percentile, half below. According to a survey reported in the news media (Foster, 1990), virtually all states of the union claim that average achievement scores for their school systems exceed the 50th percentile. The resulting overly optimistic picture of student achievement is labeled the Lake Wobegon Effect, in reference to humorist Garrison Keillor’s mythical Minnesota town where “all the children are above average.” How does inflation of achievement test scores arise? According to Cannell (1988), the major cause is educational administrators who are desperate to demonstrate the excellence of their school systems. Precisely because our society attaches so much importance to achievement

test results, some educators apparently help students cheat on standardized tests. The alleged cheating includes the following: • Teachers and principals coach students on

test answers. • Examiners give more than the allotted time

to take tests. • Administrators alter answer sheets. • Teachers teach directly to the specific test

items. • Teachers make copies of the tests to give to

their students. In sum, the importance that our society attaches to achievement test scores has caused a number of unappealing side effects that undermine the very foundations of nationally normed group- testing programs. Moore (1994) reports on a special case in educational testing, namely, the districtwide consequences of court-ordered achievement testing. He surveyed 79 teachers from third- through fifth-grade level in a midwestern town in which the court required the use of a standardized test to determine the effectiveness

of a desegregation effort. The test in question, the Iowa Tests of Basic Skills (ITBS), is a well- respected group achievement test that requires strict adherence to instructions and time limits for obtaining valid results. Yet the teachers found little value in the testing program, complaining that its benefits did not offset the time and costs involved. As a consequence of their devaluing the effort, nonstandard testing was practically the rule rather than the exception. The teachers engaged in several nonstandard practices, most of which tended to inflate the test scores. Inappropriate testing practices included praising students who answered a question correctly during the test (67 percent), using last year’s test questions for practice (44 percent), recoding a student’s answer sheet because he or she just “miscoded” the answer (26 percent), giving students as much time as they needed (24 percent), giving students items that were directly off the test (24 percent), and giving hints or clues during the test (23 percent). In general, Moore (1994) notes that teachers modified their instructional efforts

and curriculum in anticipation of having their students take the test. More than 90 percent of the teachers added test-related lessons to the curriculum, and more than 70 percent eliminated topics so that they could spend more time on test-related skills. What this study demonstrates is that mandated educational testing can have the unanticipated consequence of polluting the validity of a worthy test—especially when crucial stakeholders have no voice in the process. Further, in teaching to the tests, educators may emphasize bits and pieces of factual knowledge rather than imparting a general ability to think clearly and solve problems. In conclusion, it appears that an excessive emphasis on nationally normed achievement tests for selection and evaluation promotes inappropriate behavior, including outright fraud and cheating on the part of students and school officials. Just how widespread is the problem? Although we live with the optimistic assumption that fraud in nationally normed testing programs is rare, the

disturbing truth is that we really don’t know how often this occurs.

1.16 REPRISE: RESPONSIBLE TEST USE We return now to the real-life quandaries of testing mentioned at the beginning of the topic. The reader will recall that the first quandary had to do with whether a consulting psychologist responsibly could refuse to provide feedback to police officer candidates referred for preemployment screening. Surprisingly, the answer to this query is “Yes.” Under normal circumstances, a practitioner must explain assessment results to the client. But there are exceptions, as explained by Principle 9.10 of the APA Ethical Code: Psychologists take reasonable steps to ensure

that explanations of results are given to the individual or designated representative unless the nature of the relationship precludes provision of an explanation of results (such as in some organizational

consulting, preemployment or security screenings, and forensic evaluations), and this fact has been clearly explained to the person being assessed in advance.

The second quandary concerned a counselor who continued to use the MMPI even though the MMPI-2 has been available for several years. Is the counselor’s refusal to use the MMPI-2 a breach of professional standards? The answer to this query is probably “Yes.” The MMPI-2 is well validated and constitutes a significant improvement upon the MMPI. As mentioned previously, the MMPI-2 is now the standard of care in MMPI-based assessment of psychopathology. The counselor who continued to rely on the original MMPI could be liable for malpractice suits, especially if his test interpretations resulted in misleading interpretive statements or a false diagnosis. The third predicament involved the use of a neighborhood friend as translator in the administration of the WISC-IV to a 9-year-old boy whose first language was Spanish. This is usually a mistake as it sacrifices strict control of

the testing material. The examiner was not bilingual and, therefore, he would have no way of knowing whether the translator was remaining faithful to the original text or was possibly supplying additional cues. In an ideal world, the proper procedure would be to enlist a Spanish-speaking examiner who would use a test formally translated and also standardized with Hispanic examinees. For example, the Escala de Inteligencia Wechsler Para Ninos- Revisada de Puerto Rico (EIWN-R PR) would be a good choice. The final quandary concerned the client who informed a psychologist that her recently deceased brother was most likely a pedophile. Is the psychologist obligated to report this case to law enforcement? The answer to this query is probably “Yes,” but it may depend on the jurisdiction of the psychologist and the wording of the relevant statutes. In fact, the psychologist did report the case to authorities with unexpected consequences. Police obtained a search warrant, went to the home of the client’s mother (where the brother had lived), and

ransacked the brother’s bedroom. The mother was traumatized by the unexpected visit from the police and blamed the fiasco on her daughter. A bitter estrangement followed, and the client then sued the psychologist for violation of confidentiality!