Ethical Concerns and Practice Recommendations
Ethical Concerns and Practice Recommendations
Psychological Testing in Child Custody Evaluations With Ethnically Diverse
Families: Ethical Concerns and Practice Recommendations
EDDIE Y. CHIU Alliant International University, San Francisco, California
Given increasing ethnic diversity of divorcing families, it is more important than ever for family court evaluators to understand the ethical concerns and principles underlying the use of psycho- logical testing with ethnic family members. Major professional organizations have promulgated the guidelines of child custody evaluation with an imperative to consider cultural factors in the evaluation. However, there is a paucity of literature that addresses ethnicity and psychological testing in child custody evaluations. This article addresses three major ethical domains: (a) compe- tence, (b) test fairness, and (c) test interpretation when psychologi- cal tests are used. Practice recommendations provide guidelines to ensure ethical standards and competence in using psychological testing with ethnically diverse families in child custody and other family court evaluations.
KEYWORDS child custody evaluations, ethnicity, culture, forensic assessment, psychological testing
Culture and ethnicity play a significant role in child custody evaluations. To provide a culturally competent evaluation with a divorcing family from a nondominant culture, the evaluator should consider the family’s ethnicity and other cultural factors in addition to the standard methods of data collection. An evaluation missing the cultural component could affect the
Address correspondence to Eddie Y. Chiu, California School of Professional Psychology, Alliant International University, One Beach Street, Suite 200, San Francisco, CA 94133. E-mail: firstname.lastname@example.org
Journal of Child Custody, 11:107–127, 2014 Copyright # Taylor & Francis Group, LLC ISSN: 1537-9418 print=1537-940X online DOI: 10.1080/15379418.2014.919245
admissibility of the evidence in court as the validity of the collected infor- mation might be questionable.
Although the importance of striving for cultural competence in child custody evaluations has been recognized by major professional organizations (e.g., American Psychological Association [APA], 2009; Association of Family and Conciliation Courts [AFCC], 2006), there is an absence of literature that addresses culture and ethnicity in child custody issues. A literature search in September 2013 using the term child custody on the PsycInfo database showed 2,838 results; however, an advanced search connecting child custody with the terms culture, cultural diversity, race, ethnicity, or ethnic showed only 5 to 31 results, approximately 1% or less of the total number of articles. It is problematic that the advancement of our knowledge about child custody issues relating to ethnically diverse groups is so sluggish.
Given the paucity of literature on the interplay of ethno-cultural factors in child custody evaluations, child custody experts are finding it difficult to keep updated on scientific knowledge about such dimensions so as to ensure the competency of their evaluation process. Evaluators using psychological testing in the evaluation process might easily commit errors in their selection and interpretation of the test data for multicultural families. A failure to elim- inate test bias can be a serious ethical concern in the practice of psychology and law (American Educational Research Association [AERA], APA, & National Council on Measurement in Education [NCME], 1999; AFCC, 2006; APA, 2001, 2002, 2013b). Following ethical standards, evaluators make every effort to eliminate and reduce test bias that might lead to over- or under-diagnosis, inaccurate impressions, and inappropriate child custody recommendations. This article addresses major ethical concerns in using psychological testing in child custody evaluations with ethnically diverse families. The discussion of ethical issues in assessment is based on the APA’s Ethical Principles of Psychologists and Code of Conduct (APA Ethics Code; APA, 2002). Although the APA Ethics Code applies only to psychologists, its ethical standards and guidelines are useful for other forensic evaluators in the practice of child custody evaluations. Other evaluators including psy- chiatrists, social workers, and marriage and family therapists are encouraged to comply with the ethical standards in using psychological testing promul- gated by their professional agency. Finally, practice recommendations for psychological testing with ethnically diverse families in the context of custody disputes are made.
ETHNICITY, IMMIGRANTS, AND DIVORCE RATES
Many child custody disputes today involve ethnically diverse families. According to the 2010 U.S. Census, over one-third of the U.S. population identified their race and ethnicity in categories other than non-Hispanic
108 E. Y. Chiu
white, representing a growth of 29% in the past decade (U.S. Census Bureau, 2011a). Moreover, many of these people are immigrants, who now exceed 40 million or 13% of the total U.S. population. The largest two immigrant groups are from Latin America (53%) and Asia (28%). These two immigrant groups have mostly populated the east and west coasts. For example, more than 1 in 4 residents of California and more than 1 in 5 residents of New York and New Jersey are immigrants. About 85% of them speak a language other than English at home (U.S Census Bureau, 2011a).
In 2011, just over 14 million children under age 18 lived in single-parent ethnic families, according to the Kids Count Data Center (2013). This number does not include children of two or more races. The report indicates that Hispanic=Latino and Black=African children are the two largest groups, accounting for 48% and 46% of the total ethnic children in single-parent families, respectively. One-fourth of Hispanic=Latino and one-half of African American children live in single mother families, compared with one-sixth of white children. Over half of the children in single parent families are under 9 years of age (Mather, 2010).
The divorce rate among U.S. ethnic families is very high. In 2009, the largest divorce rate at any year of marriage occurred among Black=African Americans, followed by white, Hispanic, and Asian populations (U.S. Census Bureau, 2011b). Only 37% of Blacks stay in their first marriage for 20 years, compared to 50% of non-Hispanic white couples. The percentage of U.S. immigrant families experiencing separation and divorce has been increasing. The U.S. Census (2011b) indicates that immigrant families currently in separation and divorce status are 11%, compared to 14% of native-born populations.
PSYCHOLOGICAL TESTING AND CHILD CUSTODY EVALUATION
When there is an unresolved dispute over the child custody arrangements between parents, the court will recommend that the parents and the child undergo a child custody evaluation, an assessment of the needs of the child and each parent’s ability to meet those needs. The evaluation is assumed to serve the ‘‘best interest of the child’’ but the standards vary from state to state. Relevant information is gathered through standard data collection methods including interviews, observations, home visits, collateral contacts, and psychological testing (AFCC, 2003). Increasingly, psychological testing is being adopted and recommended by evaluators and the courts, particularly for comprehensive or full child custody evaluations. It is assumed that psychological testing will generate useful data informing child custody eva- luators, attorneys, and judges about the parenting capacity of each parent and the child’s needs. It aims to assist in formulating opinions, conclusion, and recommendations in child custody cases. Given the high utilization of
Ethnicity and Psychological Testing 109
psychological testing in family court cases, test results are often accepted as supporting evidence for legal custody arrangements. However, whether information based on psychological testing has met the criteria for admissibility in court is rarely questioned. Recent literature has addressed the controversy about using psychological testing in family court and other forensic settings (Emery, Otto, & O’Donohue, 2005: Gould, Martindale, & Flens, 2009; Hagan & Hagan, 2008; Medoff, 2003, 2010; Sanders & Katz, 2013). Emery et al. found that there may be a role for psychological tests in child custody evaluations, but evaluators should not assume that the test instruments will address all questions in the custody context. For example, there are no valid psychological tests that directly measure the important custody factor of parenting capacity. Furthermore, it is debatable whether the constructs assessed by the measure are broadly conceived ‘‘states’’ or ‘‘traits.’’
Professionals have advocated for a high standard for evidence admissi- bility of psychological testing in the forensic setting (e.g., Hagan & Hagan, 2008; Medoff, 2003, 2010). In the context of child custody evaluations, the selection, administration, and interpretation of the tests must be in com- pliance with ethical and scientific standards as delineated in professional and legal communities. The APA Ethics Code (2002) states that ‘‘psycholo- gists [should] base the opinions contained in their recommendations, reports, and diagnostic or evaluative statements, including forensic testimony, on information and techniques sufficient to substantiate their findings’’ (Standard 9.01a) and ‘‘psychologists administer, adapt, score, interpret or use assessment techniques, interviews, tests, or instruments in a manner and for purposes that are appropriate in light of the research on or evidence of the usefulness and proper application of the techniques’’ (Standard 9.02a). The APA Ethics Code (2002) also emphasizes the scientific value of validity and reliability of the instruments (Standard 9.02b).
In forensic settings, Daubert standards are often used to advocate for a higher threshold for admissibility of psychological testing in child custody evaluations. The Daubert standard provides guidelines for the judge as to whether expert witness testimony can be admitted to or excluded from court proceedings. By and large, the judge looks for evidence that ‘‘an expert’s testimony both rests on a reliable foundation and is relevant to the task at hand’’ (Daubert v. Merrell Dow Pharmaceuticals, Inc., 1993, p. 580). In addition, evidence or testimony must be derived from scientifically valid methods. Today, the courts in many states hold to the Daubert standard, which has superseded the earlier well-known Frye criteria for admissibility of scientific evidence (Frye v. U.S., 1923). The Frye standard determines if expert witness testimony is admissible as evidence based on whether its method or technique is generally accepted as reliable in the relevant scientific community. In some states, the superior courts still follow the Frye standard.
110 E. Y. Chiu
Family court experts should adhere to the ethical and practice standards promulgated by their own professional community. Data collection methods that fail to meet the prevailing scientific and professional standards are generally poor in scientific value and are probably inadmissible in court. In other words, the court should not admit scientifically invalid or unreliable data as evidence in child custody disputes. For example, the use of projective tests such as the Rorschach (Exner, 2003) and the Thematic Apprehension Test (TAT; Murray, 1943) is controversial in child custody evaluations (e.g., Calloway, 2005; Erard, 2005, 2007; Erickson, Lilienfeld, & Vitacco, 2007a, 2007b). The scientific value of projective tests has been questioned as they are not well validated empirically in research (Erickson et al., 2007a; Medoff, 2010). On the other hand, some objec- tive tests such as the Wechsler Adult Intelligence Test-IV (WAIS-IV; Wechsler, 2008) and the Minnesota Multiphase Personality Inventory-2 (MMPI-2; Butcher, Dahlstorm, Graham, Tellegen, & Kaemmer, 1989) have been assumed to have much higher scientific value because their validity and reliability have been empirically demonstrated. Nevertheless, no tests are perfect. The stability of cognitive abilities and personality functioning on these instruments vary across cultural contexts and, as a result, can often be questioned. It is argued that personality should not be conceived as a fixed set of traits regardless of the indi- vidual’s ecological and cultural contexts (e.g., Piekkola, 2011; Suzuki, Onoue, & Hill, 2013). Moreover, the applicability of these Western standardized tests to individuals from ethnically diverse families has been challenged.
PSYCHOLOGICAL TESTING WITH ETHNIC FAMILIES IN CHILD CUSTODY EVALUATIONS
Debate over the appropriate use of psychological tests across racial, ethnic, and cultural backgrounds started several decades ago. Given that the norms and standardization of widely utilized instruments were mostly based on white European populations, the applicability of the tests may not transfer to diverse ethnic populations. For instance, the MMPI-2 and the MCMI-III are the most frequently used personality tests in child custody evaluations. These tests have demonstrated reliability and validity in research settings; however, there are concerns about the lack of representation of ethnic groups in the normative sample. The authors of the MMPI-2 recruited only 73 Hispanics=Latinos and 19 Asians in the standardization of the norms, 2.9% and 0.7% of the total sample, respectively (Butcher et al., 1989). Such small Hispanic and Asian sample sizes failed to represent adequately the Hispanic and Asian ethnics group in the United States at the time of standardization.
The first edition of MCMI was developed with separate norms for Black patients (Millon, 1977). However, the MCMI-II (Millon, 1987) and the MCMI-III (Millon, 1994; Millon, Millon, Davis, & Grossman, 2006) eliminated the race-specific norms. Moreover, the overall representation of ethnic
Ethnicity and Psychological Testing 111
groups in the standardization of the MCMI-III was inadequate. The normative sample included 84% White, 8.5% Black, 4.4% Hispanic, 1.4% American Indian, .5% Asian, and about 1% other. Except the American Indian group, other ethnic groups were under-represented in the normative sample compared with U.S. Census norms (Kwan & Maestas, 2008).
Although the Rorschach findings have been often questioned as admiss- ible evidence in court, it is one of the most frequent used tests besides the MMPI-2 and the MCMI-III in child custody evaluations (Ackerman & Ackerman, 1997; Erard, 2005; King, 2013). In a recent survey of the quantity and content of legal citations between 1996 and 2005 where the Rorschach had been used, Meloy (2008) found that only 2% of 150 cases were chal- lenged by attorneys. Meloy contended that the Rorschach continues to have authority in forensic settings. However, the cross-cultural validity of the Rorschach has been seriously questioned (Allen & Dana, 2004). The latest normative sample consisted of 450 nonpatient adults with a majority of Whites (83%) (Exner & Earberg, 2005). The ethnic groups, which included 9% Black, 7% Hispanic, and 2% Asian, were underrepresented compared to U.S. Census norms at the time of standardization.
To address ethnic group differences in Rorschach performance, Rorschach researchers developed international reference samples for the Rorschach Comprehensive System from 17 international countries (Shaffer, Erdberg, & Myer, 2007). Shaffer et al. found that the adult samples were similar in their protocols and recommended incorporating international composite reference values when the Rorschach was used in other countries with adults (King, 2013). Recently, international nonpatient adult norms have been developed in a new interpretive system called the Rorschach Performance Assessment System (R-PAS) (Meyer, Viglione, Mihura, Erard, & Erdberg, 2011). As R-PAS is a new measure, its use with ethnic populations in the custody context requires further exploration.
Given the under-representation of ethnic sample groups in the norma- tive samples, evaluators should question both external validity and differen- tial validity of these major personality tests when they are used in child custody evaluations with ethnic individuals. With this in mind, the MMPI-2, the MCMI-III, and the Rorschach have their roles in child custody evalua- tions. They have been extensively studied and have established reliability and validity in research. It is reasonable to conclude that these tests have contributed to necessary information (e.g., whether a mother’s depression or a father’s schizoid personality might affect parenting).
ETHICAL CONCERNS AND PRACTICE RECOMMENDATIONS
A range of professional organizations have promulgated ethical standards and professional guidelines that govern the practice of psychological testing
112 E. Y. Chiu
in clinical and forensic settings. Examples of these standards and guidelines include the APA Ethics Code (APA, 2002), Guidelines for Child Custody Evaluations (APA, 2009), Model Standard of Practice for Child Custody (AFCC, 2006), Specific Guidelines in Forensic Psychology (APA, 2013b), Standards of Educational and Psychological Testing (AERA, APA, & NCME, 1999), and Guidelines for Test User Qualifications (Turner, DeMers, Fox, & Reed, 2001). These standards cover many aspects of the use of psychological testing and assessment including the qualifications of the test user, the use of psychological tests, test fairness, test construction, and test security. All of these guidelines aim at promoting the highest standards in the use of psycho- logical tests in a variety of settings. The following discussion covers APA’s (2002) key ethical domains in using psychological testing with ethnic families in child custody evaluations. These domains include competence of the test user, fairness in testing, and test interpretation.
Competence of the Test User
According to the APA Ethics Code (2002), psychologists are ethically respon- sible for ensuring their competence in providing delegated services to their consumers. Competence here is defined as having sufficient knowledge, skills, and experience for the intended purpose. As such, a competent test user has suitable and adequate knowledge, skills, and experience in administering, scoring, and interpreting psychological tests in child custody evaluations. There are many ways of defining competence in the forensic mental health community. This article focuses on the trilogy of competence in clinical and forensic assessment: technical competence, contextual competence, and cultural competence.
According to the APA Ethics Code (2002), psychologists use psychological testing ‘‘only within the boundaries of their competence, based on their edu- cation, training, supervised experiences, consultation, study or professional experiences’’ (APA Ethics Code, Standard 2.01a). Psychologists should have technical knowledge of test theory and the psychometric properties of selec- ted tests, and actual experience in administering, scoring, and interpreting the tests. The AFCC (2006) suggested in the Standards of Practice that in child custody evaluations, ‘‘child custody evaluators not trained in and experi- enced in the selection and administration of formal assessment instruments and not reasonably skilled in data interpretation shall not conduct testing’’ (AFFC, 2006, p. 17). Such practice guidelines are consistent with the APA Ethics Code on the competence domain. The test users should have acquired specific qualifications with appropriate training and experience on psycho- logical testing.
Ethnicity and Psychological Testing 113
For those who work in the forensic setting of child custody evaluation, tech- nical competence may not suffice. The APA emphasizes that psychologists acquire knowledge about the specific contexts in which the tests are used (APA, 2001, 2009, 2013a, 2013b). These contexts may include but are not lim- ited to employment, education, vocational and career counseling, health care, and forensic assessment. I call this contextually relevant knowledge and skills contextual competence. In a child custody evaluation, psychologi- cal tests are used to collect information about the parent’s capacity and the needs of the child so that recommendations of child custody arrangements can be substantiated. Psychologists and other test users should be familiar with the statutory and administrative procedures of the local family court system in which psychological testing is used in the custody evaluation. For example, psychological tests may not be required in every custody case. Recently some states in the United States have begun using brief focused assessments for child custody disputes in which psychological testing might not be needed. Procedures for reporting test results may also vary across court systems. Some states provide test users with nondisclosure privileges to safeguard test materials from wrongful release (Kaufmann, 2009).
Given the increasing ethnic and cultural diversity of American families, cultural competence is not an optional skill in child custody or other forensic evaluations. The APA (2002) expects psychologists in the practice of psycho- logical testing to have ‘‘an understanding of factors associated with age, gender, gender identity, race, ethnicity, cultural, national origin, religion, sexual orientation, disability, language, or socioeconomic status . . . for effec- tive implementation of their services’’ (APA Ethics Code, Standard 2.01b). The test taker’s cultural background will affect the test performance. For example, Hamby and Gray-Little (2000) found that lower partner income and being Black increased self-labeling of domestic violence. The assessment of intimate partner violence is one of the critical concerns in child custody evaluations (Saunders, Tolman, & Faller, 2013).
In addition, there is the potential for bias in the interpretation of test results especially when there are differences in values between the evaluator and the test taker. Labeling theory holds that the existence of social distance between an agent of control and a rule breaker (or an outsider) will lead to an overestimation of pathology in the rule breaker by the agent of control (Becker, 1963). In the context of child custody evaluations, if the behavior of the parent seems unfamiliar or unintelligible to the rules held by the evaluator, it is more likely to be diagnosed as problematic or abnormal (Gray-Little, 2009; Gray-Little & Kaplan, 1998). In Western psychology, for
114 E. Y. Chiu
example, emotional expression is perceived as a sign of healthy well-being while emotional suppression is considered an indicator of an individual’s defensiveness and difficulties in relationship with others. Emotions, however, are expressed differently across cultures. A recent study found that Chinese students who were psychologically healthy reported suppressing their emotional expression in order to preserve interpersonal harmony with others (Wei, Su, Carrera, Lin, & Yi, 2013). The emotional suppression of a parent from a culture of emotional self-control might be misperceived as emotional unavailability to the child by a Western-trained evaluator, who would thus misinterpret the parent’s emotionally availability to the child.
The child custody evaluator’s knowledge of the test taker’s culture can influence the interpretation of test results. For example, an elevated scale 6 score on the MMPI-2 is considered paranoia or insecurity toward others. In the context of child custody evaluations, the MMPI-2 scale 6 can be used to indicate the sensitivity of a parent to a child’s positive comments about the former spouse (Gordon, Stoffey, & Bottinelli, 2008). This might lead to a per- ception that the former spouse is attempting to alienate the child’s bonding from the other parent. The attachment with the child can be affected when the parent is insecure toward others (Pickering, Simpson, & Bentall, 2008).
However, such clinical impressions and interpretations can be biased when the evaluator has limited knowledge of the parent’s cultural back- ground. The literature on cross-cultural mental health has addressed the manifestation of ‘‘healthy cultural paranoia’’ or ‘‘cultural mistrust’’ in ethnic groups, particularly for those who have experienced racism and=or oppression (Ridley, 1984; Whaley, 2001). Groth-Marnat (2009) observed that MMPI-2 scores might reflect an individual’s personality traits, but they can also be an expression of cultural experience. Elevated scores on the scales associated with paranoia, anger, and frustration may be the result of racial discrimination. Asian and Black individuals are found to have higher elevated scores on the MMPI and the MMPI-2 scale 6 (Cheung, Song, & Zhang, 1996; Cheung, Zhang, & Song, 2003; Whatley, Allen, & Dana, 2003). An Asian parent with a history of traumatic migration may appear guarded during the testing and express mistrust toward authorities. For example, a Vietna- mese refugee parent who has a history of war and traumatic migration experiences may be very hesitant to talk much about the family and his or her immigration history. Being silent is often a way of coping with previous traumatic pain experience (Chiu, 2008). A Black parent who is psychological healthy can be culturally guarded against racism from a white evaluator (Ridley, 1984). This nondisclosing test attitude can be ‘‘normal’’ to many eth- nic groups, and it should not be quickly labeled as pathological guardedness and paranoia in child custody evaluations. Psychological assessment with culturally and ethically diverse groups requires a cultural understanding and sensitivity to the test taker’s cultural background that will influence the interpretation of test data. The APA advocates for cultural competency in
Ethnicity and Psychological Testing 115
the researcher’s and practitioner’s practice with ethnically, linguistically, and culturally diverse populations: ‘‘Psychological service providers need a sociocultural framework to consider diversity of values, interactional style, and cultural expectations in a systematic fashion. They need knowledge and skills for multicultural assessment’’ (APA, 1993, p. 45).
Psychologists and other test users should strive for technical, contextual, and cultural competence in using psychological testing in child custody evalua- tions with ethnic families.
1. Psychologists should comply with the APA Ethics Code (2002) on under- taking ‘‘ongoing efforts to develop and maintain their competence’’ (APA Ethics Code, Standard 2.03). To achieve cultural competence, ongoing efforts should include enhancing knowledge and skills working with ethnically diverse families. Consultation with cultural experts should be sought when needed. Striving for professional competence, and in parti- cular, cultural competence, is a lifelong learning process.
2. Psychologists and other test users should honestly assess their own knowledge and biases of the test taker’s cultural perceptions and values regarding parenting and the parent-child relationship. Custody experts are ethically obligated to provide a nonbiased and nondiscriminatory evaluation.
3. Psychologists and other test users must recognize their own strengths and limitations in psychological testing with ethnically diverse populations. When evaluators have questions about the test taker’s cultural values and practice in parenting and parent-child relationship, consultation with cultural experts is highly recommended. When the limits of competence are reached, referrals should be made to culturally and linguistically appropriate professionals for psychological testing and=or evaluation. In some situations when an interpreter is necessary to assist with the testing process, the interpreter must not only speak fluently the language of the test taker, but also be well trained in the translation in the assessment and child custody context. Using interpreter in assessment process poses practical challenges to the evaluator as well. Evaluators should be trained how to use the interpreter in a meaningful way. It should be noted that the validity and reliability of the test would be compromised because the use of interpreter deviates from the standardization of the test.
Fairness in Testing
Fair testing practice refers to providing and using tests that are fair to all test takers regardless of cultural, ethnic, and linguistic background (AERA, APA,
116 E. Y. Chiu
& NCME, 1999; Fremer, Diamond, & Camara, 1989). This is an important ethical issue in psychological testing with diverse ethnic populations. Fairness refers to the testing process in which ‘‘the test takers are given a comparable opportunity to demonstrate what they know and how they can perform in the area being tested’’ (Fremer et al., p. 1065). The APA (2002) requires that individuals benefit from equal quality of psychological test measures, procedures, and interpretation regardless of cultural background (APA Ethics Code, Standard 9). The standard states that psychologists should ‘‘use assessment instruments whose validity and reliability have been established for use with members of the population tested’’ (APA Ethics Code, Standard 9.02b). Population here refers to the normative sample of gender, age, ethnicity, languages, nonclinical, or clinical clients, and other population characteristics. In other words, applicability of test results based on the standardization in a given population might not be transferable to a different population. The APA (2002) further underscores that ‘‘psychologists use assessment methods that are appropriate to an individual’s language preference and competence, unless the use of an alternative language is relevant to the assessment issues’’ (APA Ethics Code, Standard 9.02c).
In child custody evaluations, the Minnesota Multiphasic Personality Inventory-2 (MMPI-2; Butcher et al., 1989) and Millon Clinical Multiaxial Inventory-III (MCMI-III; Millon et al., 2006) are the most widely used objective psychological tests to assess a parent’s personality and emotional functioning that has an effect on parenting. The validity and reliability of these two tests have been established in populations including White and other major ethnic groups in the US. However, as previously discussed, the samples of some ethnic groups were inadequately represented in the normative sample for test standardization.
The test validity is even more diminished when these English-based tests are administered to a parent who speaks little English. To complete the test, the MMPI-2 and the MCMI-III require the test taker to have at least 6th grade and 8th grade English comprehension levels, respectively. Many ethnic immi- grant populations may not meet such English proficiency requirements. A majority of the members of Hispanic and Asian groups are foreign-born immi- grants and English is not their primary language. Administration of the test to individuals from non-English-speaking backgrounds is a significant challenge to the reliability and validity of the measure. The APA Ethics Code (2002) states that it is appropriate to administer a test according to ‘‘an individual’s language preference and competence’’ (APA Ethics Code, Standard 9.02c). However, the child custody evaluator often selects the tests based on the test taker’s language preference but not language competence, which are not equivalent (Fisher, 2009). Furthermore, individuals may feel embarrassed to reveal their limited proficiency in English. This is particularly true of ethnic immigrant populations whose primary language is not English. My clinical experience confirms that many educated Asian immigrants (Master’s or higher
Ethnicity and Psychological Testing 117
degree) in the United States still have difficulty comprehending some test items, particularly in the MMPI-2. Many MMPI-2 items that describe person- ality characteristics or emotions also may contain idiomatic expressions that are unfamiliar to non-native English speakers. Inaccurate understanding of the items may bias the test results.
Both the MMPI-2 and MCMI-III have been translated and effectively adapted into different cultural contexts (Butcher, 1996; Cheung, 2009; Craig & Olson, 2005; Kwan & Maestas, 2008). Ongoing research has tested the validity and reliability as well as the linguistic and psychological equivalences of the translated measures. Child custody evaluators may not be well informed about these translated versions of instruments and their psycho- metric properties. For example, the MMPI-2 has been translated into Asian languages including Chinese, Japanese, Korean, Thai, Vietnamese, and Hmong (Butcher, Cheung, & Kim, 2003, Cheung et al., 1996). The translated versions of the MMPI-2 also include revisions to some items to make them more culturally appropriate for particular ethnic populations. In addition, individuals from different cultural backgrounds may differ in their perceived meaning of some item content. The test adaption must ensure the conceptual and functional equivalences of the test. Although a well-translated and adapted version of the measure can be found, the norms developed in the original standardized samples may not be transferable and applicable to other cultural and ethnic groups (Cheung, 2009). A specific norm may be needed for a particular ethnic group. For example, there is now a Chinese version of the MMPI-2 with its own norm for Chinese populations (Cheung et al., 2003).
Test users should keep themselves updated and informed about empirical and scientific evidence of a test’s usefulness and applicability to an individual’s cultural, ethnic, and linguistic backgrounds. However, psychologists and other child custody experts may not be well informed regarding the strengths and limitations of the English-based and translated instruments.
Test users are ethically obligated to provide a nonbiased and nondiscrimina- tory psychological testing in child custody evaluations.
1. When selecting and administering English-based instruments with ethnic parents or children, both the test taker’s language preference AND competence should be properly evaluated.
2. If the test taker does not have the English comprehension level required by the test manual, the examiner should determine if a translated and adapted version of the test is available.
3. The test user should carefully examine the psychometric properties, the conceptual and functional equivalences, and the applicability of the
118 E. Y. Chiu
translated instruments with ethnically diverse populations. Butcher (1996) and Craig and Olson (2005) provided a detailed strategy for translation and adaptation of the MMPI-2 and MCMI-III in international use.
4. When language-specific psychological tests and=or their norms are not available, the APA Ethics Code (2002) states that psychologists should ‘‘clarify the probable impact of their limited information on the reliability and validity of their opinions, and appropriately limit the nature and extent of their conclusions or recommendations’’ in the child custody evaluation report (APA Ethics Code, Standard 9.02b).
5. When suitable instruments cannot be found in a child custody evaluation with ethnic parents and children, the evaluator might consider giving up the use of psychological tests in the custody evaluation and relying more on other data collection methods, including interviews and observations.
It is imperative to ensure the accuracy of test interpretations; this is critical for diagnostic impressions, custody arrangements, and legal recommendations. Some test interpretation issues have been addressed in the aforementioned dis- cussion. When interpreting test results, psychologists are ethically obligated to:
take into account the purpose of the assessment as well as the various test factors, test-taking abilities, and other characteristics of the person being assessed, such as situational, personal, linguistic, and cultural differences, that might affect psychologists’ judgments or reduce the accuracy of their interpretations. They indicate any significant limitations of their interpre- tations, (APA Ethics Code, Standard 9.06)
Elevated scores may be indicative of cultural characteristics of personality rather than actual psychopathology. For example, research has shown that Asians may have elevated scores on MMPI and MMPI-2 Scales F, L, K, Hs, D, Pa, Sc compared to Whites (Butcher et al., 2003; Cheung et al., 1996). Cheung et al. (1996) found that Scale 2 (depression) and 8 (schizophrenia) were elevated about 10T-score points in Chinese normative samples as com- pared with U.S. norms. The elevated scores of Pa and Sc scales in the MMPI and MMPI-2 have been considered indicators of a parent being at risk for mal- treating the child (Ezzo, Pinsoneault, & Evans, 2007; Paulson, Afifi, Thomason, & Chaleff, 1974). Based on such an elevated profile, the evaluator might then suggest reducing or even terminating the parental rights in the child custody evaluation. However, an alternative explanation for the elevated findings could be consistent within the Chinese socio-cultural context, in which emotional self-control, healthy paranoia, and restraint are considered positive and protective. Therefore, both test interpretation and custody recommenda- tions should take the peculiar cultural contexts into consideration.