Psychological testing and assessment aiken pdf

265

DOI: 10.1037/14047-015 APA Handbook of Testing and Assessment in Psychology: Vol. 1. Test Theory and Testing and Assessment in Industrial and Organizational Psychology, K. F. Geisinger (Editor-in-Chief) Copyright © 2013 by the American Psychological Association. All rights reserved.

C h a P t e r 1 5

EThICS In PSyCholoGICAl TESTInG And ASSESSmEnT

Frederick T. L. Leong, Yong Sue Park, and Mark M. Leach

Since their early origins in the use of intelligence tests for placement of schoolchildren through the recent attention to high-stakes educational testing, psychological testing and assessment have remained controversial and complex topics. This controversy underscores the importance of addressing the ethical challenges in the use and application of tests and assessment in psychology. In this chapter, we begin with an overview of the various professional ethical standards that guide our work in this area. This sec- tion is followed by a more detailed review and dis- cussion of the relevant sections of the American Psychological Association (APA) Ethical Principles of Psychologists and Code of Conduct (APA, 2010). In this review, we also provide some guidance on the application of these ethical principles to the testing and assessment enterprise. Given the increasing cul- tural diversity of the U.S. population and the rise of globalization, we end with a discussion of some unique challenges in conducting testing and assess- ment cross-culturally.

There are also legal issues associated with testing and assessment in psychology, but these issues are not covered in this chapter because they are addressed elsewhere in this handbook (see Chapter 28, this volume, and Volume 2, Chapters 6 and 34). It is interesting to note that the U.S. Office for Human Research Protections highlights the differ- ences between ethical principles and regulatory guidelines. Ethical principles refers to ethical values and principles aimed at the protection of human participants in research, whereas regulatory guide- lines refers to a list of procedural dos and don’ts

(“Distinguishing Statements of Ethical Principles and Regulatory Guidelines,” 2011). The purpose of this chapter is to discuss the ethical values and prin- ciples in professional psychology as they pertain to testing and assessment.

PROFESSIONAL ETHICS

Ethics is a broad term that encompasses the com- monly endorsed values of professional psychology (Groth-Marnat, 2006) and is the basis for ethics codes—rules and guidelines on appropriate behav- iors for the purpose of protecting the public and the profession (Meara, Schmidt, & Day, 1996). In the United States, three major sources of ethics codes related to psychological testing and assessments are available: (a) the Standards for Education and Psycho- logical Testing (American Educational Research Association [AERA], APA, & National Council on Measurement in Education [NCME], 1999), (b) the Guidelines for Computer-Based Tests and Interpreta- tions (APA Committee on Professional Standards & Committee on Psychological Tests and Assessment, 1986), and (c) the Ethical Principles of Psychologists and Code of Conduct (APA, 2010).

Standards for Education and Psychological Testing In 1985, AERA, APA, and NCME collaborated to develop the Standards for Education and Psychological Testing—a set of standards pertaining to professional and technical issues of test development and use in education, psychology, and employment. The

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Leong, Park, and Leach

266

Standards is organized in three sections: (a) Test Construction, Evaluation, and Documentation; (b) Fairness in Testing; and (c) Testing Applications. The Standards document was significantly revised in 1999 to contain a greater number of standards and updated to reflect changes in law and measurement trends, increased attention to diversity issues, and information on new tests and new uses of existing tests (AERA et al., 1999). An in-depth review of the Standards can be found in Chapter 13 of this volume.

Guidelines for Computer-Based Tests and Interpretations With the increased use of, and concern for the lack of regulation of, psychological computer-based test- ing (CBT), APA’s Committee on Professional Stan- dards and Committee on Psychological Tests and Assessment (1986) published the Guidelines for Computer-Based Tests and Interpretations, a set of 31 guidelines aimed at both test developers, to ensure the development of quality CBT products, and end users of these products, to ensure proper adminis- tration and interpretation of computer-based psy- chological tests (Schoenfeldt, 1989). More recently, the International Test Commission gave increased attention to CBT in its own set of CBT guidelines, adopted in 2005, titled the International Guidelines on Computer-Based and Internet-Delivered Testing. Similar to the objectives of the Guidelines for Computer-Based Tests and Interpretations, the general aim of the International Test Commission guidelines is to recommend standards for good practices for development and use of CBTs. The International Test Commission guidelines are organized along the following recommendations: (a) Give due regard to the technological issues in computer-based and Internet testing, (b) attend to quality issues in CBT and Internet testing, (c) provide appropriate levels of control over CBT and Internet testing, and (d) make appropriate provision for security and safe- guarding privacy in CBT and Internet testing.

American Psychological Association Ethical Principles of Psychologists and Code of Conduct APA adopted its first official code of ethics in 1952 in response to the field’s increased professionalism

and visibility after World War II (Fisher, 2009). Since then, the APA Ethics Code has been revised 10 times, with an amended version being adopted in 2010 by the APA Council of Representatives. The APA Ethics Code contains four major sec- tions. The first section, Introduction and Applica- bility, delineates the rationale, scope and limitations, and applicability of the Ethics Code and describes the possible consequences and sanc- tions imposed on APA members and student affili- ates who are found to have violated the standards of the Ethics Code. The second section, the Pream- ble, contains a statement of APA’s purpose as a profession and delineates the various roles and responsibilities held by psychologists. The third section, General Principles, contains the five aspi- rational general principles of APA meant “to guide and inspire psychologists toward the very highest ethical ideals of the profession” (APA, 2010, p. 3): Beneficence and Nonmaleficence, Fidelity and Responsibility, Integrity, Justice, and Respect for People’s Rights and Dignity. Finally, the fourth section, Ethical Standards, contains a set of 10 enforceable ethical standards by which psycholo- gists are obligated to abide. Sanctions may be imposed on psychologists who violate these ethical standards. The ninth section of the Ethical Stan- dards provides guidelines pertaining to the use of psychological tests and assessments (APA, 2010). In the next section, we discuss the APA ethical standards on assessments in greater detail as they apply to a variety of purposes and contexts in which psychological testing is conducted.

AMERICAN PSYCHOLOGICAL ASSOCIATION ETHICAL STANDARDS ON ASSESSMENTS

In the sections that follow, we highlight the 11 assessment standards associated with the APA Eth- ics Code. These standards have been found in the ethics codes of other countries, although the degree to which there is consistency differs based on a country’s use of testing. In addition, other countries did include an additional standard not found in the APA Ethics Code (Leach & Oakland, 2007). The consistency found indicates that these standards

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Ethics in Psychological Testing and Assessment

267

have international appeal and form the ethical foun- dation of test use and development.

Bases of Assessments APA Ethical Standard 9.01, Bases for Assessments, stipulates that all oral and written opinions and con- clusions made by psychologists be based on infor- mation and techniques grounded in the scientific and professional knowledge bases of professional psychology (Fisher, 2009). Adherence to the scien- tific and professional standards of the field builds public trust in the profession consistent with Princi- ple B, Fidelity and Responsibility, of the APA Ethics Code. When psychologists’ opinions and conclu- sions are not grounded in the scientific and profes- sional standards, the probability that their opinions may mislead and potentially harm the clients and patients whom they serve is greater. Professional discernment applies to all phases of the testing and assessment process, even in the preassessment phase of planning and information gathering (Jacob & Hartshorne, 2006).

Scientific and professional bases. According to APA Ethical Standard 9.01a, psychologists are obligated to base their recommendations, reports, and diagnostic or evaluative statements on tech- niques supported by the scientific and professional standards of the field. Moreover, Ethical Standard 9.01b stipulates that opinions on individuals’ psychological characteristics be drawn after an adequate examination is conducted on the basis of assessment procedures and tools that are consistent with the objective of the testing (e.g., that address the referral question), are sensitive to the cultural and linguistic characteristics of the examinee, are congruent with the examinee’s level of competency to be administered the assessment, and have been shown to be valid and reliable. Psychologists are responsible for personally ensuring that the reli- ability and validity of the assessment tools and techniques they use are adequate. Furthermore, psychologists should base their conclusions and recommendations on assessments that have been demonstrated to be reliable and valid. Reliability and validity issues are discussed in greater depth in Chapters 2 and 4 of this volume.

Limitations of assessment results. When limita- tions to the reliability and validity of the assess- ment procedures and tools are found, psychologists should appropriately limit the nature and extent of their conclusions and recommendations and refrain from drawing conclusions that are not adequately supported. Another scenario to limit conclusions may arise when psychologists are unable to person- ally evaluate an individual for various reasons, such as an examinee’s refusal to continue with assess- ment or an examinee’s relocation during the course of assessment. In these situations, psychologists should make reasonable efforts, when appropriate and practical, to reach examinees for assessment and thoroughly document the outcome of these efforts (Ethical Standard 9.01b). When a personal evalua- tion is not practical, psychologists are obligated to limit the scope of their decisions and recommenda- tions, in addition to delineating how the limited information influences the reliability and validity of their findings.

Cases may exist in which personal evaluation of an examinee is not warranted, such as when reviewing preexisting records in academic, legal, organizational, and administrative contexts or when examining secondary records provided by a third-party assessor, such as trainees or profession- als with whom psychologists supervise or consult, respectively (Fisher, 2009; Knapp & VandeCreek, 2003). In these cases, psychologists should clearly explain that their conclusions and recommenda- tions are based on a secondary analysis of informa- tion derived from alternate sources (Ethical Standard 9.01c).

Use of Assessments Psychological testing applies to a wide range of pur- poses and contexts, which include but are not lim- ited to screening applicants for job placement, diagnosing psychological disorders for mental health treatment, verifying health insurance cover- age, conducting focus groups for market research, informing legal decisions and governmental policies, and developing measures to reliably measure per- sonality characteristics (Aiken & Groth-Marnat, 2006; Fisher, 2009). According to the Eighteenth Mental Measurements Yearbook (Spies, Carlson, &

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Leong, Park, and Leach

268

Geisinger, 2010), there are no less than 19 major categories of psychological tests and assessments.

APA Ethical Standard 9.02 pertains to the proper selection and use of psychological tests and assess- ments. The first component of this ethical standard stipulates that psychologists administer, adapt, score, interpret, and use psychological testing in the manner and purpose for which the selected tests and assess- ments were designed to be used as indicated by research (Ethical Standard 9.02a). Furthermore, psy- chologists should select and use tests or assessments with members of populations for whom adequate reli- ability and validity of the test scores has been estab- lished. If the reliability and validity of the test scores has not been examined or verified for a particular population, psychologists are obligated to describe the strengths and limitations of the interpretations and recommendations derived from the test or assess- ment results (Ethical Standard 9.02b). The third aspect of this ethical standard obligates psychologists to select tests and assessments that are appropriate to the language preference and competence of the indi- viduals being assessed (Ethical Standard 9.02c).

Test selection and usage. Psychologists are responsible for selecting appropriate assessments for the intended purpose of the testing (Ethical Standard 9.02a). To guide the selection of appro- priate tests and assessments, psychologists should have adequate knowledge of the theoretical bases and empirical evidence that support the validity and reliability of the tests or assessments; stan- dardized administration and scoring procedures; approaches to interpreting the results; and the popu- lations for which the assessment was normed and designed (Fisher, 2009; see Ethical Standard 9.07, Assessment by Unqualified Persons). Psychologists should also keep themselves updated on the most recent versions of the tests and assessments that they commonly use because testing and assessment procedures and parameters may change in light of theoretical advances and new research (see Ethical Standard 9.08, Obsolete Tests and Outdated Test Results). Finally, psychologists should select tests and assessments that have been empirically vali- dated to be used in the specific contexts and settings in which the testing occurs.

Testing across diverse populations. According to Principles D (Justice) and E (Respect for People’s Rights and Dignity) of the APA Ethics Code, psy- chologists strive to establish fair and equal access to and benefit of psychological contributions for all individuals and populations, which include but are not limited to diversity in age, gender, gender identity, race, ethnicity, culture, national origin, religion, disability, language, and socioeconomic status. Although psychological testing represents a unique contribution of professional psychology to benefiting larger society, ensuring the fair and equal access to and benefit of psychological testing has historically been challenging for the field. According to Reynolds (1982), the reliability and validity of test and assessment scores have predominately been established with White, middle-class samples and may not generalize well to other populations, espe- cially those that represent a minority in the United States. This historical precedence conflicts with Ethical Standard 9.02b, which stipulates the selec- tion and use of assessments that have been found to be adequately valid and reliable for drawing particu- lar inferences for specific populations being assessed. When tests are administered across diverse popula- tions, psychologists are obligated to select and use tests and assessments that have measurement equiv- alence in that the psychometric properties (i.e., mea- surement and structural models) have been shown to be equivalent or invariant between members of culturally different populations and those from the reference population for which the test and assess- ment scores were validated, normed, and found to be reliable (Schmitt, Golubovich, & Leong, 2010).

Testing and language. APA Ethical Standard 9.02c stipulates that psychologists select tests that are appropriate to be used with the language prefer- ences and levels of competence of the individuals or groups being assessed. Thus, before selecting assess- ments, it is helpful for psychologists to gather infor- mation on examinees’ cultural background (e.g., acculturation) and native and English language ability with regard to written, reading, and spoken language proficiencies (Jacob & Hartshorne, 2006; Takushi & Uomoto, 2001). According to Groth- Marnat (2009), literal translation of testing

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Ethics in Psychological Testing and Assessment

269

and assessment materials and tools using the commonly implemented method of translation– back-translation may not be adequate because of cross-cultural differences in the conceptual inter- pretation of items, noncomparable idioms, and within-group differences in dialect and word usage. Furthermore, from an item response theory frame- work, literal translation of testing and assessment items from one language to another may change the properties of the items’ difficulty, which may in turn diminish the measurement equivalence of tests or assessments. For these reasons, the psychometric properties of the original-language version of tests or assessments cannot be assumed to generalize to the alternate-language versions that were developed from a translation–back-translation method. More information on testing and language can be found in Volume 3, Chapter 26, of this handbook.

With regard to testing conducted in person (e.g., interviews) with linguistically different clients, psy- chologists may consider enlisting the services of a translator for interpretation purposes or consider referring clients to colleagues who have professional proficiency in the clients’ language. Professional organizations may be useful resources for identify- ing and referring clients to professional colleagues with the appropriate linguistic background; for example, the National Association of School Psy- chologists maintains a directory of bilingual school psychologists that can be found on its website (http://www.nasponline.org/about_nasp/ bilingualdirectory.aspx).

Informed Consent in Assessments Before administering an assessment, psychologists are obligated to obtain from examinees, or their par- ents, guardians, or legal representatives, informed consent that includes an explanation of the nature and purpose of the assessment, fees, involvement of third parties (e.g., referral source), and limits of confidentiality (see Ethical Standard 3.10, Informed Consent). The informed consent stage of testing may also be the opportune time to provide examin- ees with an explanation of their rights as test takers. The Joint Committee on Testing Practices (1998) developed the Rights and Responsibilities of Test Tak- ers: Guidelines and Expectations to inform test takers

about and clarify expectations for the testing pro- cess. Because consent refers to examinees’ legal sta- tus to autonomously decide whether to be assessed, informed consent must be communicated in a clear and comprehensible manner that is appropriate to the age of examinees and their mental abilities (Fisher, 2009).

As stipulated by Ethical Standard 9.03a, informed consent can be dispensed with in the fol- lowing situations: when “(1) testing is mandated by law or governmental regulations; (2) informed con- sent is implied because testing is conducted as a routine educational, institutional or organizational activity; or (3) one purpose of the testing is to evalu- ate decisional capacity” (APA, 2010, p. 12). Even though informed consent is not required in these cases, psychologists are recommended to, when appropriate, continue to provide examinees with an explanation of the nature and purpose of the testing.

When assessing individuals younger than age 18 (i.e., minors), informed consent from parents or legal guardians is required because minors are viewed, from a legal standpoint, as being unable to make autonomous and well-informed decisions per- taining to psychological services. Thus, minors do not have the legal right to assent, consent, or object to a proposed psychoeducational assessment; how- ever, it is recommended that minors be fully informed about the nature and purpose of the test- ing and assessment in a clear and understandable manner (Jacob & Hartshorne, 2006).

Nature and purpose of assessment. Informed consent in the assessment context includes an explanation of the nature and purpose of the test or assessment. Thus, psychologists are obligated to clearly explain how results will be used, the admin- istration procedure, and possible benefits and risks or consequences of being assessed. With regard to informing examinees about the administration procedure, psychologists are advised to provide a general description of the procedure because fore- knowledge of the testing may influence examinees’ responses and thus alter the validity of the test or assessment results. Psychologists should also be sensitive to the possible risks and consequences of the testing, especially with regard to the negative

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Leong, Park, and Leach

270

feelings that may be generated by the testing pro- cess. Some assessment topics or questions may elicit uncomfortable feelings in examinees, such as those that involve private or taboo topics (Groth-Marnat, 2009). Thus, psychologists, in most cases, should not pressure or force examinees to answer all ques- tions, especially those that create undue discomfort or emotionally painful feelings.

Confidentiality and release of information. A core component of informed consent is explaining the limits of confidentiality. Confidentiality refers to a professional standard that requires psychologists to maintain the privacy of any assessment informa- tion unless disclosure is permitted or requested by examinees through a release of information. According to Ethical Standard 4.05, Disclosures, psychologists may breach confidentiality without examinees’ permission when disclosure is mandated by law or when permitted by law for a valid pur- pose, such as to

(1) provide needed professional services; (2) obtain appropriate professional con- sultations; (3) protect the client/patient, psychologist, or others from harm [e.g., danger to self and others, elder and child abuse]; and (4) obtain payment for services from a client/patient, in which instance disclosure is limited to only information that is necessary to obtaining the payment. (APA, 2010, p. 7)

In situations in which breach of confidentiality is necessary or legally mandated, psychologists should share only information that is necessary to accom- plish the purpose of the disclosure in an effort to respect examinees’ right to privacy.

Health Insurance Portability and Accountability Act and Family Educational Rights and Privacy Act. Because of the increased reliance on electronic databases to store client–patient information, psy- chologists are responsible for effectively protecting the confidentiality and security of the information contained in these databases (Aiken & Groth- Marnat, 2006). The Health Insurance Portability and Accountability Act (HIPAA) was established in 1996 to regulate the protection of protected health

information. Protected health information refers to any information that

(a) is created or received by a health care provider, health plan, public health authority, employer, life insurer, school or university, or health care clearing- house; and (b) relates to the past, pres- ent, or future physical or mental health or condition of any individual, the provi- sion of health care to an individual, or the past, present, or future payment for the provision of health care to an indi- vidual. (Title 42, U.S.C. § 1320d)

Any health care provider who electronically trans- mits health information is considered a covered entity by HIPAA and must comply with HIPAA reg- ulations. Within the informed consent, covered enti- ties should provide examinees with a written document titled Notice of Practice Practices; this doc- ument contains a description of the examinee’s rights, the legal duty to protect protected health information, and the routine uses and disclosures of protected health information. The Family Educa- tional Rights and Privacy Act of 1974 pertains to issues of confidentiality and release of information in the educational setting. The act stipulates that assessment information and school records of stu- dents maintained by educational institutions that receive federal funding may be disclosed to others only with the written consent of the student exam- inees or their parents or legal guardians.

Language and use of interpretation services. Ethical Standards 9.03b and 9.03c refer to the psychologists’ responsibility to provide informed consent in the language of the examinee or at a lan- guage proficiency level the examinee can reasonably understand. Psychologists may enlist the services of an interpreter when working with examinees who have limited English proficiency. When using inter- preters, psychologists are responsible for ensuring that interpreters are not only competent in commu- nicating the informed consent in a reasonable and understandable manner but also comply with the ethical standard on maintaining the confidential- ity of examinees’ identity, assessment results, and

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Ethics in Psychological Testing and Assessment

271

test security (Fisher, 2009; Knapp & VandeCreek, 2003).

Release of Test Data According to Fisher (2009), a growing trend in the legal system is toward affirming the autonomy of patients’ access to their health care records, a trend that is consistent with Principle E, Respect for Peo- ple’s Rights and Dignity, of the APA Ethics Code, emphasizing self-determination. HIPAA stipulates that patients have the right to access, inspect, and receive copies of their medical and billing records on their request for the release of this information. Related to the assessment context, examinees or oth- ers identified in the release have the right, in most cases, to have access to their test data (Ethical Stan- dard 9.04a). Test data refers to raw and scaled scores on the assessment items, any responses to test ques- tions or stimuli, and psychologists’ written notes or recordings of the testing.

Test data versus test materials. It is important to note the difference between test data and test materials. Test materials refers to test manuals, administration and scoring protocols, and test items. According to Ethical Standard 9.11, test materials do not need to be released pursuant to a client or patient request for test data because test materials are protected by copyright laws, and inappropriate release of such test materials is legally considered a breach of trade secrets (Groth-Marnat, 2009; Knapp & VandeCreek, 2003). However, when examinees’ identifying information or responses are written on test materials, the test material is considered test data and may need to be released on examinees’ request (Ethical Standard 9.04a). Thus, examiners are recommended, whenever possible, to record any identifying information and responses on a separate document from the actual test materials.

Potential misuse of test data. When examinees provide a release to request test data for themselves or identified others, it is important that psycholo- gists explain the potential for test data to be mis- used if the people interpreting the test data do not have the proper qualifications to do so (see Ethical Standard 9.07, Assessment by Unqualified Persons). According to Ethical Standard 9.04a, psychologists

may refrain from releasing test data to the examin- ees or others if the release may result in substantial harm resulting from misuse or misinterpretation of the test data. In these cases, psychologists are obli- gated to document the specific rationale for why they believe that the test data would result in sub- stantial harm (Fisher, 2009).

Court order for test data. According to Ethical Standard 9.04b, psychologists are obligated to release test data when the disclosure is required by the law or court order. When release of test data is court mandated, Fisher (2009) recommended that psychologists seek legal counsel to determine the legitimacy of the request and ascertain their legal responsibility to release the test data. Another rec- ommendation is that psychologists request the court for a protective order to prevent the inappropriate disclosure of the confidential test data and recom- mend that test data be reviewed by another health care professional who is qualified to provide appro- priate and competent interpretations. Furthermore, psychologists are recommended to make reason- able efforts to notify examinees when test data are released to the court and to document these efforts (Fisher, 2009).

Test Construction Ethical Standard 9.05, Test Construction, refers to test developers’ responsibility to ensure that the development of tests and assessments incorporates appropriate psychometric procedures that are guided by the current scientific and professional knowledge of test design, standardization, valida- tion, reduction or elimination of bias, and recom- mendations for use.

Standardization. Test developers are responsible for providing specific and clear guidelines to quali- fied test users with regard to the proper and stan- dardized procedure for administering and scoring tests and assessments. Furthermore, test developers are responsible for specifying the scoring cutoffs and norms for the populations for which the tests and assessments were developed and intended to be used. Scoring norms are commonly found in norm- referenced tests, which allows for comparison of individual scores to the distribution of scores from

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Leong, Park, and Leach

272

the reference group. It is important that the char- acteristics of the reference group sample are clearly described in the test or assessment manual and are representative of the population to which the test is targeted.

Validity. According to the Standards for Educational and Psychological Tests (AERA et al., 1999), valid- ity is defined as the degree to which the theoretical basis for the assessment and accumulated empiri- cal evidence support the intended interpretation of the scores for which the assessment was designed. In general, validity refers to the degree to which an assessment measures what it purports to measure. Several types of evidence are used to justify claims of validity, such as content-related evidence and criterion-related evidence. For an in-depth review, readers are referred to Chapter 4 in this volume.

Reliability. The Standards for Educational and Psychological Tests (AERA et al., 1999) stipu- late that test developers are obligated to provide reliability estimates—the degree to which the assessment results are consistent over repeated administrations—of their tests and assessments. Jacob and Hartshorne (2006) recommended that reliability estimates be provided for each demo- graphic subpopulation of the population for which the assessment was intended, such as for age groups and class levels. Several methods can estab- lish the reliability of an assessment: internal consis- tency, test–retest, split-half test, and alternative-form comparisons. For an in-depth review, readers are referred to Chapter 2 in this volume.

Interpreting Assessment Results Interpretations of test and assessment results influ- ence the decisions and recommendations that are made in reference to the purpose of the testing (see Ethical Standard 9.02, Use of Assessments), such as diagnosing and informing treatment plans in clinical settings and educational placements in academic set- tings and determining employment selections and promotions. Interpretations should be based on proper administration of tests and assessments as outlined by the testing manual to ensure the inter- pretations are in line with the evidence to support the validity and reliability of the test or assessment

scores (Fisher, 2009). It is the psychologist’s respon- sibility to ensure that his or her interpretations of test or assessment results are useful and relevant to the purpose of the assessment and take into account various test factors, test-taking abilities, and other characteristics of individuals being assessed (Ethical Standard 9.06).

Interpretation of multiple sources. Interpretations of test and assessment results should not be derived from a simple, mechanical process that is based solely on the test or assessment scores, score cut- offs, or reliance on automated interpretations (Fisher, 2009; Groth-Marnat, 2009) but that takes into consideration a host of factors, including but not limited to examinees’ characteristics, test-taking abilities, styles, issues of fatigue, perceptual and motor impairments, illnesses, language proficien- cies, and cultural orientations (Fisher, 2009). Furthermore, Groth-Marnat (2009) recommended that psychologists base their interpretations on mul- tiple sources of data, including behavioral observa- tions, examinee background information, and other assessments. Often, testing is administered using an integrated battery of assessments, and inconsistent findings across the various assessments may result. In these situations, it is the psychologist’s respon- sibility to analyze the contradictions and use his or her clinical and professional judgment to offer the most accurate and relevant interpretation in relation to the purpose of testing (Groth-Marnat, 2009).

Automated interpretations. There are many well- established, standardized assessments, such as the Minnesota Multiphasic Personality Inventory—2, for which one can receive a computer-generated automated interpretative report. Although these automated interpretations are based on a body of past empirical evidence and theoretical models, it is important to highlight that interpretations are not sophisticated enough to take into account examin- ees’ unique characteristics and test-taking contexts. Thus, psychologists should not base their interpreta- tions solely on automated interpretations but rather use automated interpretations as supplemental resources for integrated interpretations that take into consideration a host of other factors that may influence the testing.

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Ethics in Psychological Testing and Assessment

273

Limitations of interpretations. According to Ethical Standard 9.06, Interpreting Assessment Results, psychologists are obligated to indicate any significant limitations of their interpretations, espe- cially when the interpretations are not supported by the established validity and reliability of the test or assessment scores in making particular inferences. When interpretation of test or assessment scores is made outside their established validity and reliability, Fisher (2009) recommended that such interpreta- tions be posed as hypotheses, rather than conclu- sions, to elucidate the limitations of such findings. Another limitation that needs to be indicated is when testing procedures and materials, evidence for validity and reliability, and score cutoffs and norms have become obsolete in the face of new research or changes in the populations for which tests and assessments were designed (see Ethical Standard 9.08, Obsolete Tests and Outdated Test Results).

Assessment by Unqualified People APA Ethical Standard 9.07, Assessment by Unquali- fied Persons, warns against the promotion of psy- chological assessment techniques being used by unqualified people. Psychologists are obligated to ensure that testing is carried out by qualified indi- viduals within the scope of their competence as indicated by their education and training back- ground and past experiences (Fisher, 2009). Fur- thermore, qualified psychologists have knowledge of the nature and purpose of the assessments, their psychometric properties, standardized procedure for administration and scoring, proper interpretation of results, and assessment limitations (Groth-Marnat, 2009). Unqualified users may also include psycholo- gists who are working with populations or problem areas that are outside the scope of their competen- cies (see Ethical Standard 2.01, Boundaries of Com- petence), such as working with culturally and linguistically different clients whom they are not multiculturally competent to serve.

Assessment by unqualified people may result in misdiagnosis of the examinees’ presenting con- cerns and potentially result in psychological harm (Jacob & Hartshorne, 2006). Aiken and Groth- Marnat (2006) suggested that the unqualified use of assessments has greater consequences when

assessing individuals (e.g., intelligence and person- ality assessments) as opposed to groups because misuse of assessment results can have direct nega- tive consequences on people’s livelihoods, such as being prescribed a treatment plan for an incorrect diagnosis or being placed at the wrong educational level or in the wrong job placement. In relation to Principle A, Beneficence and Nonmaleficence, of the APA Ethics Code, psychologists should be aware of the boundaries or limitations of their competence to prevent unqualified use of assessments and make appropriate referrals or seek supervision or consul- tation from specialists in these situations (Aiken & Groth-Marnat, 2006). Furthermore, psychologists are recommended to obtain access to or create a directory of local assessment specialists for referral purposes (Jacob & Hartsthorne, 2006).

Qualifications. According to Turner, DeMers, Fox, and Reed (2001), qualified use of assessments often includes graduate course work and supervised train- ing experiences pertaining to the use of specific assess- ments. In 2002, the Psychological Assessment Work Group convened at the Competencies Conference: Future Directions in Education and Credentialing in Professional Psychology and identified a set of eight core competencies in psychological testing:

1. A background in the basics of psychometric theory.

2. Knowledge of the scientific, theoretical, empirical, and contextual bases of psycho- logical assessment.

3. Knowledge, skill, and techniques to assess the cognitive, affective, behavioral, and per- sonality dimensions of human experience with reference to individuals and systems.

4. The ability to assess outcomes of treatment/ intervention.

5. The ability to evaluate critically the multiple roles, contexts, and relationships within which clients and psychologists function, and the reciprocal impact of these roles, contexts, and relationships on assessment activity.

6. The ability to establish, maintain, and under- stand the collaborative professional relationship that provides a context for all psychological activity including psychological assessment.

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Leong, Park, and Leach

274

7. An understanding of the relationship between assessment and intervention, assessment as an intervention, and inter- vention planning.

8. Technical assessment skills that include: (a) problem and/or goal identification and case conceptualization, (b) understanding and selection of appropriate assessment meth- ods including both test and non-test data (e.g., suitable strategies, tools, measures, time lines, and targets), (c) effective appli- cation of the assessment procedures with clients and the various systems in which they function, (d) systematic data gathering, (e) integration of information, inference, and analysis, (f) communication of findings and development of recommendations to address problems and goals, (g) provision of feedback that is understandable, useful, and responsive to the client, regardless of whether the client is an individual, group, organization or referral source. (Krish- namurthy et al., 2004, pp. 732–733)

The Psychological Assessment Workgroup also delineated core competencies of training programs in providing quality educational and training experi- ences for psychological testing.

Ethical responsibility for qualified use applies not only to individual psychologists but also to test developers with regard to the distribution of their test materials. Standards for qualified use have been established by test developers to prohibit unqualified users’ access to test materials. Thus, test developers should include information on the required qualifi- cations for use in the test’s promotional materials and require end users to meet the minimum require- ments to purchase and use their tests and assess- ments. Aiken and Groth-Marnat (2006) provided a sample qualification form for test developers that includes questions for the potential end user with regard to the purpose for using the test, area of pro- fessional expertise, level of training, specific courses taken, and quality control over test use (e.g., test security, appropriate tailoring of interpretations).

Assessment by trainees. Although APA Ethical Standard 9.07 stipulates that psychologists should

not promote unqualified use of assessments, an exception is made for training purposes as long as trainees have adequate supervision while the assess- ments are provided. More specifically, for trainees to be qualified in administering tests or assessments, they must have been or concurrently be enrolled in a graduate-level course, practicum externship, or pre- or postdoctoral training program that provides training in the specific assessment that is being administered. In addition to the formal training, trainees must receive adequate supervision from a qualified user of the test or assessment. In cases in which unqualified trainees have not received suf- ficient training and supervision to administer the assessment, they must clearly inform examinees that the test or assessment is being administered for training purposes only and adequately describe the limitations of their assessment interpretations, con- clusions, and recommendations (Fisher, 2009). It is important to note that when supervising psycholo- gists sign their trainees’ assessment reports, they are ultimately held responsible for the contents of the report (Jacob & Hartshorne, 2006).

Obsolete Tests and Outdated Test Results Psychologists are prohibited from basing their deci- sions and recommendations on test data that are outdated for the test’s current use (Ethical Standard 9.08a) and from tests and assessments that are obso- lete and not useful for the current use (Ethical Stan- dard 9.08b). Use of outdated test data is prohibited because examinees may have changed since the time of the prior assessment owing to such factors as maturational and developmental effects, develop- ment of new presenting problems, and changes in the environment (Fisher, 2009). When outdated test results are used, psychologists are obligated to pro- vide an explanation for why outdated test data are used and to clearly communicate the limitations of such outdated information.

Old test data are often kept stored in outdated files or databases even after examiners no longer work at the testing location. In this situation, psy- chologists are recommended to prevent the misuse of outdated test results by taking reasonable steps to remove or destroy obsolete data and files. In cases in which clients or patients request that outdated test

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Ethics in Psychological Testing and Assessment

275

data be sent to a new clinician who is currently pro- viding services to them, psychologists are recom- mended to include a cover page detailing the limitations of outdated test results.

APA Ethical Standard 9.08 also stipulates that psychologists should not base their decisions and recommendations on use of obsolete assessments. According to Fisher (2009), tests developers often revise their assessments to reflect significant advances and changes in the theoretical constructs underlying the psychological characteristics being assessed; changes in the assessment’s test item valid- ity owing to various cultural, educational, linguistic, or societal influences; and shifts in the demograph- ics of the target population, which in turn affect the standardized norms and score cutoffs. Use of obso- lete tests may be applicable when long-term compar- isons of test performance are needed, but psychologists are obligated to adequately describe the differences between test versions and explain the limitations of their comparisons when obsolete tests are used. According to Fisher (2009), the expense associated with updating to new versions is not an adequate ethical justification for using obsolete tests and assessments.

Test Scoring and Interpretation Services APA Ethical Standard 9.09 applies to psychologists who provide test scoring and interpretation services. Within their promotional and other administrative materials (e.g., manuals), these psychologists are obligated to accurately describe the nature and pur- pose of the assessments, the basis for the standard- ized norms, and validity and reliability information for their assessment results and interpretations and to specify the qualifications for using the services. When interpretations and recommendations from assessment results are made, psychologists are obli- gated to provide the theoretical rationale and psy- chometric evidence for justifying their conclusions and to adequately explain the limitations of their interpretations and recommendations.

Ethical responsibility for the appropriate use of test scoring and interpretation services also applies to psychologists who are consumers of these ser- vices. These psychologists are obligated to select ser- vices that adequately provide evidence for the

validity and reliability of their procedures for administering, scoring, and interpreting test and assessment results. Furthermore, psychologists using these services are obligated to have the qualifi- cations and competence to ensure that the scoring and interpretations made by these services are con- sistent with APA Ethical Standard 9.06, Interpreting Assessment Results. When these services are used, the HIPAA Notice of Privacy Practices obligates psy- chologists to inform and obtain authorization from their clients or patients to permit the release of test or assessment information to these services.

Explaining Assessment Results According to Ethical Standard 9.10, Explaining Assessment Results, psychologists are obligated to provide competent feedback to examinees, or to par- ents or legal guardians of minors, explaining any interpretations, decisions, and recommendations in relation to the purpose of testing. Groth-Marnat (2009) recommended that the feedback begin with a clear explanation of the rationale for testing, fol- lowed by the nature and purpose of the assessment, general conclusions drawn from assessment results, limitations, and common misconceptions or misin- terpretations of assessment results. When examinees are minors, psychologists are obligated to provide the feedback to both examinees and their parents or legal guardians.

Sensitivity in the communication of assessment results. The Standards for Educational and Psychological Tests (AERA et al., 1999) stipulates that simple, clear, everyday language should be used when providing feedback so that the feedback is readily understood by its recipients. Psychologists should tailor their level of communication to recipients’ personal characteristics, such as their educational and linguistic backgrounds, level of knowledge of psychological testing, and pos- sible emotional reactions to the assessment results (Groth-Marnat, 2009). With regard to the pos- sible emotional reactions generated by feedback, it may be helpful for psychologists to make available options for follow-up counseling to facilitate ser- vices for examinees who may need support in pro- cessing the feedback information. When providing

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Leong, Park, and Leach

276

feedback on mental health status, Aiken and Groth- Marnat (2006) recommended that the least stig- matizing label be used to describe the examinees’ psychological conditions or diagnoses.

Written reports. In addition to the oral feedback session, psychologists commonly provide written reports to examinees, or their referral source, regard- ing the assessment results, interpretations, and rec- ommendations. Written reports should be centered on referral questions and the purpose of the testing and adequately describe the characteristics of the examinees and how they relate to the assessments used and the test situations (Aiken & Groth-Marnat, 2006). According to Jacob and Hartshorne (2006), written reports should be comprehensible to both professionals and nonprofessionals and should be written in a succinct, clear, and comprehensible manner while avoiding overgeneralizations (Aiken & Groth-Marnat, 2006). Psychologists are respon- sible for signing off on assessment reports only after ensuring the accuracy of the contents contained in the reports.

Maintaining Test Security According to Ethical Standard 9.11, Maintaining Test Security, psychologists are obligated to main- tain the security of test materials, which are defined as manuals, instruments, protocols, and test questions or stimuli. As noted in Ethical Stan- dard 9.04, although examinees have the right to request and access test data, they do not have the right to access test materials for reasons related to threats to validity and copyright protection. For these reasons, test materials should be stored in a secure location, and only authorized and qualified individuals should have access to them. Further- more, test materials, even sample items, should not be reprinted in any form, such as in newspapers and magazines, without the written consent of the test developers.

Threat to validity. A primary reason for the ethi- cal obligation to maintain test security is the threat to test validity that is posed when individuals have access to test materials before administration of the test. Having foreknowledge of the test questions and answers may alter the psychometric properties

of the test, including its standardized score cutoffs and norms and validity (Fisher, 2009). Furthermore, access to test materials before administration may increase the likelihood of some individuals manipu- lating their responses for purposes of malingering or obtaining an unfair advantage on a given assessment relative to others (Knapp & VandeCreek, 2003).

Copyright law. Pursuant to copyright protection laws, it is illegal and an ethical violation to repro- duce test materials without obtaining permission from test developers or publishers. Maintaining test security allows for the protection of trade secrets and honors the terms of agreement made with the test publisher on obtaining access to the test materi- als (Groth-Marnat, 2009). With regard to HIPAA, which stipulates that examinees have the right to access their protected health information (e.g., test data), psychologists should separate, when appro- priate, test materials from test data to protect the copyrighted test materials from being disclosed when releases of information are requested by cli- ents or patients.

CROSS-CULTURAL ISSUES

Testing and assessment become inherently more complex when considering cross-cultural issues. Our position is that to be ethically and multicultur- ally competent when conducting testing and assess- ments, the psychologist should consider the client’s cultural context. Approximately one third of the U.S. population consists of ethnic minorities, and when one includes the potential influence of other diverse groups (e.g., language, disability, socioeco- nomic status) on testing, it becomes evident that to be competent in testing and assessment requires much more than basic knowledge of test use.

All of the principles described in APA’s (2010) Ethical Principles of Psychologists and Code of Con- duct apply to cross-cultural testing, yet two are briefly highlighted that seem particularly salient. These are Principle D (Justice) and Principle E (Respect for People’s Rights and Dignity). First, Principle D refers not only to equal access and fair- ness but to psychologists’ ensuring that their biases, boundaries of competence, and level of expertise do

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Ethics in Psychological Testing and Assessment

277

not influence their work and lead to unjust practices. Second, Principle E refers to respecting differences among individuals and cultural groups and the belief in autonomous self-determination. Unfortu- nately, sound ethical testing practices have not always been the norm when considering the history of the testing movement in psychology. Although progress in ethical testing practices has been made over the years and the field has improved signifi- cantly in the development, measurement, and implementation of testing with regard to culture, further developments are needed.

Psychological testing has made great strides in the understanding of psychological constructs, and it continues to do so. It also has a well-referenced history of bias against those who are not White, middle class, and male. The acceptance of the belief in universality, that the mainstream American expe- rience is applicable to everyone, has long been at odds with a multicultural framework. This frame- work states that testing and assessment cannot be uniformly applied to all groups (Leong, Qin, & Huang, 2008). Using a simple example, readers would probably agree that assessing women if a test was normed on men or adults if a test was normed on elementary school-aged children would not be ethically appropriate. Similarly, there may be con- cerns about the application of tests primarily normed on the dominant group when considering use with nondominant group members. Consistent with many psychologists today, Burlew (2003) cau- tioned against taking a universal philosophical approach in that theories may not be transferable across cultures, that researchers are limited from developing alternative theories, that protective mea- sures unique to a particular cultural group are neglected, and that any deviation from the universal perspective leads to a pathological or deviational view of nondominant outgroups. Only during the past few decades has research attention been given to the inclusion of diverse individuals and groups as they relate to the richness in understanding human behavior.

Etic Versus Emic Validity from a cross-cultural perspective begins with knowledge of differences between etic and

emic approaches to testing. Simply defined, etic approaches assess constructs across cultures, whereas emic approaches examine a construct within a particular culture. Understanding these validity issues is crucial when developing or using tests because tests are generally developed within a particular cultural context. Both etic and emic approaches are discussed in greater detail next, and examples from history are included to highlight eth- ical issues that have emerged.

Etic Psychological testing has been at the forefront of controversy since the early part of the 20th century because of differences found among ethnic groups on a variety of tests, most notably intelligence tests. Imposed etics surrounding psychological assess- ment probably began with Galton’s (1883/2003) treatise, “Inquiries Into Human Faculty and Its Development.” This document led to the “mental test,” which then helped launch psychology’s ver- sion of the eugenics movement (Schultz & Schultz, 2011). Other psychologists such as Cattell, God- dard, and Terman were influential in launching intelligence and ability testing into conventional psychology. These famous psychologists, along with other equally as recognizable names such as Yerkes, were influential in putting forth testing practices that were unfavorable toward ethnic minorities, those of lower socioeconomic status, and others. More recently, Herrnstein and Murray’s (1994) con- troversial book The Bell Curve revived the debate over the relationship among (primarily ethnic) groups and intelligence. Their thesis that ethnic minorities do not score well on tests of intelligence and achievement because of genetic and biological limitations harkens back to earlier testing history in psychology (for a review of the issues surrounding The Bell Curve and a rebuttal, see Jacoby & Glau- berman, 1995).

Culturally appropriate and ethical test develop- ment has recently gained significant attention in the professional literature (e.g., Dana, 2005; Groth- Marnat, 2009). In this vein, to work toward compe- tent, ethical, and culturally valid testing practices, psychologists and others have begun discussing test equivalence (or invariance). Equivalence refers to the

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Leong, Park, and Leach

278

degree to which the parameters of a test’s measure- ment model are comparable across groups (Cheung, van de Vijver, & Leong, 2011). Measurement equiv- alence is a prerequisite before one can make reason- able and ethical interpretations of the results across cultural groups. Historically, equivalence in psycho- logical testing was omitted or significantly flawed given that many psychological tests were either normed on or developed in a framework of the dom- inant culture. Quite simply, using a psychological test that has not included a broader multicultural framework may introduce bias and is ethically dubi- ous. It may be unethical because, among a myriad reasons, the psychologist is not acting competently and the foundation on which the tests were devel- oped is flawed. More specifically, the APA Ethics Code acknowledges that ethical test use requires that the test be appropriate for the individual or group under investigation. Determination of whether a psychological instrument is valid for use with a particular cultural group is based on multiple factors, such as an individual’s level of accultura- tion, translation of the instrument, language abili- ties, whether the construct measured with the instrument is consistent across cultures, and norm availability, among others. These can be accom- plished through the assessment of four types of equivalence: linguistic, conceptual, metric, and functional (Leong, Leung, & Cheung, 2010).

Linguistic Equivalence Linguistic equivalence, or translation equivalence, is primarily concerned with the translation of a psy- chological instrument and its application in another culture (Groth-Marnat, 2009). Brislin (1970) was one of the first to discuss the back-translation method, which involves translating an instrument into another language and then back-translating it into the primary language. The two versions are compared, and differences are resolved. Linguistic equivalence merely permits comprehensibility and does not, however, postulate about the instrument’s validity. It is still a common translation method, although more recent procedures regarding the area of linguistic equivalence are expounded on in Ham- bleton, Merenda, and Spielberger (2005) and Vol- ume 3, Chapter 26, of this handbook.

Conceptual Equivalence Unfortunately, linguistic equivalence may be suffi- cient with some tests, but conceptual equivalence is also needed to behave in the highest ethical manner. Conceptual equivalence determines the degree to which a concept is consistent cross-culturally. This concept is more difficult to attain because what may be considered a similar concept between cultures may actually be a close proximity to it or interpreted differently altogether, resulting in conceptual vari- ability. To decrease this variability, Usunier (1998) suggested that the translation process include multi- ple sources and target languages. Briefly, multiple native speakers independently develop words con- sistent with a concept, and a cross-cultural research team identifies the most commonly cited terms and back-translates them. Etic and emic conceptual dimensions are then determined (see also Leong et al., 2010).

Metric Equivalence Metric equivalence is concerned with whether the psychometric properties of an instrument are con- sistent across cultural groups (Groth-Marnat, 2009). This type of equivalence is delineated into two cate- gories, measurement invariance and structural invariance. Measurement invariance is related to variables’ relationships to latent constructs, whereas structural invariance involves the actual latent vari- ables themselves. Another way of considering the two is that measurement invariance is concerned with consistent matrices and scalar equivalence, for example, whereas structural invariance is concerned with whether the structural models, for example, are consistent across cultural groups. The more metric variability introduced, the greater the likelihood is that using the test across cultures is invalid and unethical.

Functional Equivalence Functional equivalence addresses the idea that pat- terns of relationships between various constructs and a target measure are equivalent. If one construct in one culture does not function in the same manner in another culture, then variability is increased. For example, cognitive distortions may be associated with depression in one culture but not in another.

Co py

ri gh

t Am

er ic

an P sy

ch ol og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

Ethics in Psychological Testing and Assessment

279

To test for cognitive distortions in one culture because of its cultural consideration as a common feature of depression in another culture could be inaccurate. To derive meaning and make interpreta- tions from test results based on functional invari- ance could be considered unethical behavior (for a brief overview of strategies to offset measurement inequivalence, see Leong et al., 2008, 2010).

At least five ethical standards should be consid- ered when evaluating tests without equivalence. We first consider a translated test developed in the Eng- lish language and administered, for example, to an individual whose native language is Spanish. As indicated earlier, Ethical Standards 9.01, 9.02, and 9.06 are directly related to test use, and these three standards are central to linguistic equivalence. Stan- dard 9.01, Bases for Assessments, states, “Psycholo- gists base the opinions contained in their recommendations, reports, and diagnostic or evalua- tive statements, including forensic testimony, on information and techniques sufficient to substanti- ate their findings” (APA, 2010, p. 12). Without linguistic equivalence, for example, a simple transla- tion without the back-translation, the psychologist is acting unethically because whether the translation is accurate is not clear. Whether the results can be used to substantiate the findings cannot be known.