Neuropsychological Assessment of Hispanics

ANTONIO E. PUENTE ANO ALFREDO ARDILA

ln a volume on cultural aspects of clinical neuropsychology, a reader living in the United States, or for that matter the Americas, would certainly expect to find a chapter on Hispanics. Hispanics represent not only the fastest growing ethnic minority in the United States but represent a unique challenge to those who have a traditional view of ethnic minorities (i.e., ethnic minorities are essentially all African American). Hispanics represent an important challenge to those working in the field of clinical neuropsychology and it is the purpose of this chapler to explain not only those challenges but also their potential solutions. The chapter begins by defining Hispanic emphasizing what is traditionally considered to be "Hispanic," such as language differences. In addition, less obvious issues are explored as well including educational, economic, religious, and psychological variables. This approach provides a foundation for addressing potential culturd differences between Hispanics and other groups, including other ethnic minorities, in neuropsychological abilities and perfor- mance. The second section focuses on neuopsychological research on Hispanics and clinical knowledge. A review of the literature, including the problems and limitations of existing knowledge, is considered. The third section comprises the major portion of the chapter. This section addresses such issues as heterogeneity of Hispanics, translation of tests, data and norns, cultural and cognitive equivalence, and the use oftime. The fourth section focuses on potential solutions to these problems. This section includes the need for more research, personnel, understanding, public policy, and support from test publishers. The rationale for the preceding concerns is based notjust on sociopolitical foundations but on the consideration of clinical experience and the scientific literature. The chapter ends with a surnmary and direc- tion, both for short and long term possibilities. The authors' intention is to synthesize existing concepts and research. It is important to note that the existing literature is not extensive. Further, some of the previously published work that will be considered is anecdotal, clinical, or theoretical. The critical question underlying the neuropsychological assessment of Hispanics is whether English and Spanish tests are actually measuring the same variables.

ANTONIO E. PLIENTE . University of North Carolina-Wilmington, Wilmington, North Carolina 38406. ALFREDO ARDILA . Miami, Florida 33182. Hutdbook of Cross-Cultural Neuropsychology, edited by Fletcher-Janzen, Srickland, and Reynolds. Kluwer Aca- demic/Plenum Publishen, New York, 2000.

87 88 ANTONIO E. PUENTE AND ALFI{EDO ARDILA

Defining Hispanic working group that so. there are small yet img Merriam-Webster's Collegiate Dictionary defines the term Hispanic as relating to the justifi able generalizati people, speech, or culture of Spain and Portugal, or Latin America. The Diccionario de la Regionalisms on Lengua Espafiola (The Dictionary of the spanish Language, [Real Academia Espafrola, Spanish. This is valid 19841)' edited by the Royal Spanish Academy, defines the word Hispanic as perrainin; to (or forth. Spanish is the pr relative to) Spain or the nations of Latin America. Seemingly, there is a basic agreement in Hemisphere. Given th, terminology, even though according portugal to Merriam-Webster's Collegiate Dictionary, (some 400 million), re and Brazil would be included as Hispanic nations. Hispanic is usually defined in the United have its own slang terr States as a person whose (or, primary in some cases, secondary) language is Spanish. Further, cal variations are also there is an assumption that their heritage is of Spanish origin with their eventual roots being correspond to a singk traceable back to Spain. Few, if any other, variables are factored into the definition of this Regional variations ir ethnic group. These assumptions are misleading, maybe even incorrect. As a consequence, general Spanish is a w several important issues that help define Hispanic will be considered prior to addressing the standard English) that primary focus of this chapter, the neuropsychology of the Spanish spiaker. According to the U.S. Bureau of the Census (1999) the total Hispanic population of the Qsntsunding Variabl U.S. currently accounts for over 29 million people (ll%o of the total U.S. population). Of these Hispanics, 637o are of Mexican origin, l4.4%o are of Central and South Americ an oigin,l0.6Vo Hispanics are an are mainland Puerto Ricans (residents of Puerto Rico are not included in these statistic s),4.2Vo Americans. Unlike thr are of Cuban ancestry, arrd 7 .4Vo are classified as being of "Other origin. " The distribution of though mainly Caucas: the Hispanic population in the United States, however, is quite uneven. Mexicans are concen- and blacks. Thus, Hisl trated in the southwestern states, especially California and Texas. Puerto Ricans mostly live in might be more repres( New York. In south Florida it is estimated that over half of the Hispanic population is Cuban. For a number of Taking into account not only the significant number of Spanish speakers, but also the fact that achieved the level of, there is one Spanish-speaking associate state (Puerto Rico), the United States can be regarded, compared to other etlu to a certain extent, as a partially Spanish-speaking country. true for both Hispanics origin. There is growin Heterogeneity Spreen & SFauss, l99i Puente, 1994), that edu Hispanics are not a unified ethnic group. One could easily divide this group into brain-damaged Hispan individuals living or from the Iberian peninsula (i.e., Spain) and those living in or from the of educated, brain-dar Americas. Those from Spain tend in many respects to be more similar to other Europeans than (U.S. Bureau of the Ce to Latin Americans. It is anticipated that the commonality will be further enhanced with the as much as both their development of the European Union. In contrast, Hispanics in the United States have a form of African Americans. In the Spanish language, customs, and behaviors more akin with Latin American countries. average relative to ot Hispanics living in or from the Americas could be further subdivided into two groups; neuropsychological as (excepting North America Mexico) versus Mexico, Central and South America. Individuals When testing peol living in or from the United States and Canada are more likely to have some knowledge of psychology as in neurc English and the American way of living. This could include an understanding of standardized insfiuments, people w testing, the importance of time and time-based productivity, and competition in academic homogenous educatio: situations. Thus, although data is lacking in this area, Hispanics from the United States are neuropsychological tes more likely than their Mexican, Central, and South American counterparts to appear similar to years of education are Nonh American cohorts on standardized testing. education can be lou Individuals living in Latin America (i.e., Mexico, Central and South America) should differences are found also be further subdivided into subcultural groups. For the initial work on the translation of the educational effect repr Wechsler scales, a panel of experts from a variety of Latin-American countries subdivided (Ardila, 1998a; Osfior Latin Americans into the following categories: Caribbean (e.g., Cuba), Mexican, Central consequence, it is franl American (e.9., Panama), and South American (e.g., Argentina). It was the belief of the homogenous group. ASSESSMENT OF HISPAMCS

working group that some minor language differences existed between these groups. Further, there are small yet important differences in food, dress, and customs that prevent an easy and spanic as relating to the justifi able generalization across all subgroups. t The Diccionario de la Regionalisms or vernacular idiosyncrasies with respect to vocabulary exist in the use of :al Academia Espafrola, Spanish. This is valid with any large language including English, Chinese, Russian, and so panic as pertaining to (or forth. Spanish is the primary language spoken in at least nineteen countries within the Western : is a basic agreement in Hemisphere. Given the expansive geographic distribution and the huge number of speakers ate Dictionary, Portugal (some 400 million), regionalisms in vocabulary usage are common. Each country will also lly defined in the United have its own slang terms that are commonly used in conversational speech. Minor phonologi- uage is Spanish. Further, cal variations are also noted. In many Spanish-speaking areas S and Z as well as LL and Y reir eventual roots being correspond to a single phoneme. The affricate /ch/ sometimes becomes the fricative /sh/. nto the definition of this Regional variations in the production of other phonernes are also found. Nonetheless, in rect. As a consequence, general Spanish is a well unified language, and there is a standard Spanish (as there is also a d prior to addressing the standard English) that is easily understood by any Spanish speaker. speaker. spanic population of the Confounding Variables ;.S. population). Of these r American oigin, 10.6%o Hispanics are an ethnic group, like African Americans, Asian Americans, and Native in these statistics), 4.2Vo Americans. Unlike those groups, however, Hispanics are not a race. In fact, Hispanics, al- gin." The distribution of though mainly Caucasians, also include people of color, including Native Americans, Asians, n. Mexicans are concen- and blacks. Thus, Hispanics are multiracial and, depending on the area ofinterest, one race rto Ricans mostly live in might be more represented than another (Shorris, 1992). nic population is Cuban. For a number of reasons, ranging from economic to sociopolitical, Hispanics have not ers, but also the fact that achieved the level of educational attainment typically seen in the United States even when d States can be regarded, compared to other ethnic groups (Shonis, 1992). This decreased educational attainment holds true for both Hispanics living in and outside of the United States, regardless of their country of origin. There is growing evidence both in the neuropsychological literature (e.g.,I*zak,L995; Spreen & Strauss, 1998) in general, and with Hispanics in particular (e.g., Ardila, Rosselli, & Puente, 1994), that education plays an important role in the expression of brain function. Non- divide this group into brain-damaged Hispanic illiterates appeared neuropsychologically highly similar to a portion lse living in or from the of educated, brain-damaged persons (Ardila et al., 1994). Recent data in the United States 'to other Europeans than (U.S. Bureau of the Census,1999) suggest that Hispanics drop out of school almost three times ulher enhanced with the as much as both their white counterparts (i.e., Anglo- or European Americans) as well as ted States have a form of African Americans. In addition, Hispanics in the United States tend to start school later than t American countries. average relative to other groups, and exit much earlier (Shorris, 1992). ln the context of livided into two groups; neuropsychological assessment of Hispanics education has to be carefully understood. rth America. Individuals When testing people of low educational levels a significant error is frequently found, so in ave some knowledge of psychology as in neuropsychology: When normalizing psychological and neuropsychological standing of standardizec instruments, people with 10 years of education and below are most frequently considered a ompetition in academic homogenous educational group. Research has demonstrated that the educational effect on m the United States are neuropsychological testperformance is not a linear effect. Differences between zero and three parts to appear similar to years of education are usually highly significant; differences between three and six years of education can be lower; between six and nine are even lower, and so on. Virtually no South America) should differences are found between, for example, 12 and 15 years of education. This means the :on the translation of tle educational effect represents a kind of negatively accelerated curve, tending toward a plateau an countries subdivided (Ardila, 1998a; Osnosky, Ardila, Rosselli, Lopez-Arango, & Uriel-Mendoza,1998). As a uba), Mexican, Central consequence, it is frankly elroneous to consider people with fewer than 10 years of education a .t was the belief of the homogenous group. 90 ANTONIO E. PUENTE AND ALFREDO ARDILA

Religion and the understanding of the physical causes of diseases are similarly important become progressively mor variables' As a consequence, understanding and acceptance ofneuropsychological assessment 1992). Thus, Hispanics li' can be affected. However, the impact of religion on everyday life is extremelyvariable among eventually integrate their Hispanics. For many Hispanics religion is simply a ntn-existing issue in life or it is jusi integration is not easY an, formally maintained as part of the culture. Many other Hispanici on the other hand, view disease as something God has given them and could, with prayer and support, take back. Thus, appreciation of even Bilingualism the basic reasons for a neuropsychological assessment may be missing in this population of individuals. Psychologists Bilingualism in gene often misunderstand customs and traditions, especially as they pertain to psychological bilingualism in the Unite assessment. There are several important factors that are rarely appretiated 1e.g., depends upon the specifi Perez-Arce & Puente, 1996; Puente & Salazar, 1998). These include time, ielationship-to examiner, and competitiveness. groups represents one of Time in Hispanic cultures is often something to be cherished, appreciated, and not Spanish and English are n necessarily conquered. Speed may not be so importanfi Good products are usually is the official language, SP the result of a careful process, and quality and rush may be contradictory. iroductivity and speed are "spanglish" (mixture of often seen as less crucial by Hispanics, as enjoying life is more useful. Time is something, found. These combinatio for example, that one spends together with friends and family. In contrast, other Americans have very written), preference (likel intensive exposure and training in testing. For them, testing is a challenge and one of use (English, Spanish, is expected to perform his or her best and quickly. For Hispanics, it may be more important 1998b). The language sP to be courteous with the examiner and establish a good rafport rather than perform mixture of both. well. It may be significant to have the opportunity to talk anJ interchange ideas. That is, personal Several distinctions the relationship with the examiner may be more important than the results of the early and late bilinguali evaluation. Competitiveness is also viewed with suspicion. Cooperation and social ability are (Manuel-Dupont, Ardila by far more important. For example, in Hispanic cultures being "educated" implies having social skills, learned the second langui not necessarily educational attainment. In fact, it is not uncommon for an individual will be mediated through with a college degree to be thought of as uneducated because of arrogance and unpoliteness. The distinction among c Further, education in Latin American countries is not synonymous with the United extensively used. The co States. For example, a college degree in Latin America equals a master's degree in the speaker of each languag, United States' both in the scope of education and the number of y"urc of college required. bilingual can be classifier the ability to sPeak anc Acculfuration bilingualism appears unl Some variables are Another very critical, if not the most critical, factor in the neuropsychological assessment sequence of acquisition, of Hispanics is acculturation. This concept is broadly defined ur un ioOiuiOual's ability to guages, patterns of use understand and maneuver outside the culture that he or she was raised in and are most familiar with (Berry, language, and even indi' 1997). The underlying question is, What is the criterion? In other words, What is are only general variabl being measured? It could be argued that the ultimate criterion in neuropsychological measure- ment is whether the patient has the ability to understand and answer the question correctly l. The age, sequen enough so that the examiner finds the answer to be generally "normal." Thii criterion is quite degree of maste different from whether a brain injury actually exists. In other words, the issue ends up bling States initiallY I whether the patient can perform the task, and nothing more. Hence, what could be measured in the age of five o a neuropsychological evaluation may actually be acculturation of the patient. If the patient they have serio understands the value of time, and time is used to assess brain function (as is often the case), 2. Schoolinglangt then the uninjured individual will respond in the allocated time. If the first premise is not true, may be the mosl the patient comes across as being brain-damaged (Ardila et at.,lgg4). Hispanics might not butcommunical have understood that they had to respond as quickly as possible (something vLry "foreign" to to either langut them). To ignore this important concern will surely result in exaggerado; of ialse positives. friends). Degree of acculturation is quite variable among Hispanics in the United States. patterns 3. Personal and sc of behavior, beliefs, and values among Hispanics living in the United States, however, tend to tions. Some His ASSESSMENT OF HISPANICS 9I

; are similarly important become progressively more similar to the middle-class American standards over time (Shorris, ;ychological assessmenr 1992)' Thus, Hispanics living in the United States present mixed patterns of behaviors and tremely variable among eventually integrate their Latin values with the middle-class Amirican values. rsue Sometimes in life or it is just integration is not easy and conflict results. rn the other hand, view upport, take back. Thus, Bilingualism ,ment may be missing in

Bilingualism in general represents a rather cially as they pertain to complex phenomenon, but Spanish-English bilingualism in the United rarely appreciated (e.g., States is an extremely issue. BilinguJirrn in geieral depends upon the specific cultural context. "o-pi"* le time, relationship to The type of relationship between both culture groups represents one of the most the nething to be cherished, crucial points. For Hispanics in the United States, Spanish and English are not in conflict and the "rules rtant: Good products are of the gu.i" are clear enough: English is the official language, rtradictory. Productivity Spanish is the informal one. But, combinations of Spanish, English and "Spanglish" (mixture of Spanish and English is more useful. Time is due to borrowings and code-switchings) are found. These combinations occur with regard unily. In contrast, other to age of acquisition, proficiency (oral and written), preference (likely correlated with the degree ofcultural identification), For them, testing is a and patterns of use (English, Spanish, and any type of combination and ror Hispanics, it may be mixture can be imagined) iAtOitu, 1998b). The language spoken at home and at work ood rapport rather than can be English, spanish, or uny type of mixture of both. interchange ideas. That Several distinctions have been proposed for grouping bilinguals. rt than the results of the The distinction between early and late bilingualism seems to be the simplest one and t}re on and social ability are most extensively used (Manuel-Dupont, Ardila, Rosselli, Puente, 1992). ucated" implies having & Early bilingual means rhe individual learned the second language before the age of about l2.Later acquisition not uncommon for an of a second language will be mediated through the first language, and second language learning :ause of arrogance and will be incompleie. The distinction among coordinate, compound, and subordinate : synonymous with the bilingualism also has been extensively used. The coordinate bilingual is an early bilingual who can a master's degree in the function as a native speaker of each language. These distinctions, though useful, are rs of college required. insufficient, as a particular bilingual can be classified in more than one group. There are many factors capable of affecting the ability to speak and understand a second language, but also there is the factor that bilingualism appears under quite different contexts and diverse circumshnces. Some variables are considered crucial to pinpoint the degree ychological assessment of bilingualism: age and sequence of acquisition, method of acquisition, schooling language, contexts r individual's ability to of the two lan- guages, patterns of use of the two languages, personal and social attitudes in and are most familiar toward each language, and even individual differences in verbal abilities (Ardila, 1998b). ln other words, What is However, these are only general variables and many variations can be found. :sychological measure- ' question the correctly 1. The age, sequence, and method of acquisition are not necessarily correlated with the " This criterion is quite degree of mastery of each language. As an example, many Hispanics in the United the issue ends up being States initially learned Spanish in their native countries and used only Spanish until at could be measured in the age of five or six. Later, these people moved to the United States, and years later, 3 patient. If the patient they have serious diffrculty speaking Spanish whereas they speak English fluently. (as n is often the case), 2. Schooling language can be indeed a highly significant and decisive variable. In fact, it first premise is not true, may be the most crucial variable. Nonetheless, many children attend classes in English .). Hispanics might not but communicate among each other in Spanish. And in general, the degree of exposure fiing very "foreign" to to either language can be extremely variable (e.g., home language, TV, neighbors, rtion of false positives. friends). United States. Patterns 3. Personal and social attitudes toward the two languages can present significant varia- itates, however. tend to tions. Some Hispanics in the United States consider that what is really important is to 92 ANTONIO E. PUENTE ANDALFREDO ARDILA

learn proper English; Spanish mafters considerably less, if at all. others consider it syncrasies of the important to m1n$n qrgficiency in Spanish. Thus,-they expect their children not only iner's language. to learn to speak spanish, but also to read, to write, and wen to appreciate spanish language literature. of these three soluti A significant percentage of Hispanics falls between these two None anchors' Their location States Hispanic in this spectrum is probably ielated to the degree of cultural United identification, the type necessarily reflecting tht of links maintained with the native countries, their age, the community in which higher than observed. Ot they live, and so forth. Again, significant heterogeneity eiists in bilingual abilities. of bilingualism should be 4' Individual differences in the acquisition of the seconc ability to learn a second language has not been a frequent topic in bilingualism life, language used in tes literature (Kilborn, 1994). But evidlntt--y, very significant differ- ences have been observed inserted within the contt in the ability to learn and ur" oot only a first, but also a second language.

Linguistic idiosyncrasies may influence the results in psychological and neuropsy- chological testing. For_an Neuropsychological I example, spelling is not used ln testing in Spanish-or in^any language relying phonological on a writing, such as Italian or Russii. 1'tre overtrained ability to spell observed in English The focus of this sr speakers has some general linguistic consequences. The use of spelled abbreviations nately, some of the exi is common in English, but not in Spanish. No native Spanish speaker would read UCLA as U, C, orientation. or both ratht L, A, but /ukla./. Understanding abbreviated words such as MS or ADHD conducted and published is extremely hard for Spanish speakers, and an intermediate decoding process (,.mul- tiple sclerosis," "attention in question. For examP deficit hyperactivity disorder") is usually required. This strong tendency to "spell" Barcelona (Junque), Gr in English is observed even when saying numbe.s (fS is tt r"e, four, five)i Spanish speakers prefer (345 Latin America, research to cluster is three hundred iorty-nvey. English speakers do better in single digit span (about issues of culture and etht 7 digits) than spanish speakers iauoui s.2-cligits;-, and it could be conjectured that spanish ethnic group in Mexico speakers would do betteiin multiple-digit span (e.g., 34,76, etc.). Interestingly, As the Catalan grot using bilingual subjects, digit span was found larger (-closer to the English norm) when performing other regions (e.g., And in English, and shorter (closer to the Spanish norm) when perfonning this test in Spanish (Ardila et al., in press). America. the traditional Most Hispanics Neuropsychology (SLA in the United States are bilinguals to different degrees, and quite often bilinguals may be at a it does not appear to b significant disadvantage when tested in either laniuage (Ardila et al., in press). A bilingual can be considered Another group (Asocia notjust as the speaker of two different languages, but as the speaker of an extended April 1999 with the inau language (Grosjean, tssl;. nom hnguages can be active lan_ guages. As a matter of fact, the development of an al their functional language can be either Spanish or English, or a mixture journal. Neuropsicolo gi of both. Interference between the languages is expected to te high. To ise either Spanish ever, neuropsychology i or English testing materials and norms can penalize (Jnited States Spanish_Engtish bilinguals. than it is in the United I Unfortunately, a clear solution to this difficult situation is not yet available. developed the neuropsy Three procedures, however, could at least reduce the bilingwtism e6ect inpsychological and neuropsychological testing. Spanish neuropsycholo Nebraska Neuropsycho Have special l. norms for Spanish-English bilinguals. This solution does not seem easy, This multidisciplit taking into consideration the tremendous heterogeneity of United States bilingualism. chology is closely aligt The 2. examiner could be a bilingual mastering a similar type of biligualism.-Testing Society is a loosely forr could be performed in spanish, English, and any combination of both languagesl meetings held each yea Instructions and answers in either language or any mixture of both languages coutJUe tional Neuropsychologi acceptable; both English and Spanish norms could be used, preferring thi one favor- has served mostly as a I ing the subject. chological assessment, Scores 3. could be adjusted to neutralize the penalizing effect of bilingualism. The fredo Ardila. Rusen Er problem here is that we do not know how much the test scores should bL adjusted in Morote, Patricia Perez order to achieve reasonable validity. Of course, it depends on the test, the idio- organize this society. ASSESSMENT OF HISPANICS 93

at all. Others consider it syncrasies the of client's bilingualism, and the testing situation, including the exam- :ct their children not only iner's language. en to appreciate Spanish None of these three falls between these two solutions alone or combined seems easy to effect. Regardless, when testing United States to the degree of cultural Hispanic bilinguals, it should be emphaiized that current results are not necessarily reflecting countries, their age, the the real individual's abilities, and his/her real performance may be higher than observed. nt heterogeneity exists in Obviously, in any neuropsychological report on a bilingual, the degree of bilingualism should be reported. An estimate could bi arrived at by considlring the age of acquisition the second ;e has not been a frequent of language, schooling language, use of both languages in everyday life, language used in I, very significant differ- testing, and norms used. Caution regarding bilingual isiues needs io bL inserted within the rt only a first, but also a context of the report as well.

rological and neuropsy- Neuropsychological Assessment of Hispanics ; in Spanish-or in any r. The overtrained ability The focus of this section )nsequences. The use of of the chapter will be on a review of the literature. Unfortu- nately, some of existing r native Spanish speaker the literature is based on anecdotal or case studies, theoretical orientation, or both rather than comprehensive ted words such as MS or data gathering and analysis. Many studies are conducted published lecoding process ("mul- and outside the United States so their applicability to the United States is in question. For example, ly required. This strong in Spain, several programs of research exist including those in Barcelona (Junque), Granada (Perez), Madrid (cespedes, r (345 is tbree, four, five). Iruarrizaga, Tobal, & cano). In Latin America, research is often carried nglish speakers do better out by non-psychologists (including neurologists) and issues ofculture and ethnicity are not as central Z digits), and it could be to their concerns. Further, Hispanics are not an ethnic group in Mexico and Central and span (e.g., 34,76, etc.). South America, as they represent the majority. As the Catalan group in Barcelona rser to the English norm) remains relatively isolated from the rest of Spain, the other regions (e.g., Andalucia) n) when performing this of Spain remain relatively isolated from each other. In Latin America, the traditional neuropsychological group has been the Latin American Society of Neuropsychology (SLAN). They have journal degrees, and quite often attempted to publish a but for numerous reasons it does not appear to be widely distributed anguage (Ardila et al., in or indexed by the major abstracting services. Another group (Asociacion Latinoamericana de Neuropsicologia, fferent languages, but as ALAN) was formed in April 1999 with the inaugural meeting ages can be active lan- held in Cartagena, Colombia. An important reason for the developmeirt of an alternative organization for Latin Spanish or English, or a America appears to be the issue of the journal. Neuropsicologia, Neuropsiquiatriay r be high. Tb use either Neurociencias started publishing in 1999. How- ever, neuropsychology is viewed more from a State s Spanish- English multidisciplinary perspective in Latin America than it is in the United is not yet available. States. For example, in Barcelona, Spain, neurologists have primarily developed the neuropsychological n effect in psychological research and clinical programs. One of the first published Spanish neuropsychological tests, which came from Barcelona and resembled the Luria- Nebraska Neuropsychological Battery, was developed by a team led by a neurologist. Ltion does not se€m easy, This multidisciplinary pattern is in sharp contrast to the United States where neuropsy- ited States bilingualism. chology is closely aligned with psychology. For example, the Hispanic Neuropsychological ' of biligualism. Testing Society is a loosely formed group that meets at either one of the two main neuropsychological tion of both languages. meetings held each year: the National Academy of Neuropsychology (NAN) and the Intema- both languages could be tional Neuropsychological Society (INS). With bylaws, elected officers, and dues the group rreferring the one favor- has served mostly as a network of professionals, mostly Hispanic, interested in the neuropsy- chological assessment of the Spanish-speaking patient. Numerous individuals including Al- ct of bilingualism. The fredo Ardila, Rusen Echemendia, Josette Harris, Tedd Judd, Patricia Linn-Fuentes, Gloria es should be adjusted in Morote, Patricia Perez-Arce, Marcel Ponton, Antonio Puente, and Monica Rosselli helped s on the test, the idio- organize this society. 94 ANTOMO E. PUENTE AND ALtrREDO ARDILA

Unfortunately, however, out of the estimated - 4,000 neuropsychologists in the United . General guidelin States, it is our estimate (based on our experience and the membership of the Hispanic . Competence Neuropsychological Society) that fewer than 50 neuropsychologists could be considered . Responsibility either bilingual or bicultural. Further, almost all-of these n"uropry""noiogists (as evidenced by . Safety of the tes the membership of the. Hispanic Neuropsychological Society) L practitioners rather than . Confidentiality researchers' This paucity of academicians may help explain the paucity in the research literature. The general guidelines

1. Understanding Methodological Problems 2. Selection of thr 3. Consideration c It is not uncommon to assume that testing Hispanics involves simply the translation of a 4. Adequate prepe test into Spanish. The assumption is that the major problem in adapting the test is simply the language. 5. Adequate use o Indeed' this approach is typically usedto assess Hispani"r. Fo. the Minne- 6. Appropriate sct sota Multiphasic Personality Inventory (MMPI), which "*"-ple, is the most common test used by 7. Appropriate int neuropsychologists (Camara, Nathan & Puente, in press), has been translated into Spanish. 8. Appropriate co However, the Spanish is intended to be applicable to all Hispanic subgroups and there are no 9. Revision of the norms associated with the test. In this section, the numerous methodologicai problems facing a simple translation of tests will be discussed. The issues to be considered include translation, Appropriate transl: hanslators, sampling and norms, culturar meaningfulness, and cognitive equivalence. the translation of the L izaga, Cano, & Tobal, Tianslatian what Harris, Cullum, a individuals had to have The most common assurnption involving the application of psychological tess to Span- difficult to fi nd balance< ish speakers is that that all is needed for the test to be useo witir a Spanish speaker is a have the psychological translation of the text (Echmendia, Harris, congett, Diaz, & puente, 1997). Anecdotal evi_ resolution of discrepanr dence obtained from memben of the Hispanic Neuropsychological Society reveals that most of issues that will latet members have idiosyncratic tanslations that they have adapteJfor their own personal use or, Another problem in many cases, special tests are "fianslated" as the test is being administered, much like what previously, this ethnic would be done in an interview. Based on what our colleagues tell us at the annual meetings, case with the Luria-Ne standardized procedures are rarely used (in part because, for example, nonns are not avail- Puerto Ricans because r able). Numerous authors_have described these procedures but in thii case the suggestions by would have been adequ Brislin (1983) are used. While we consider the approach of Brislin to be traditional, there are only the those two trar several reasons pursuing for this approach. First, some of the early research started on the speakers. This required neuropsychology , of Hispanics used this conceptual framework. Secondly, the use of back- Standard Spanish was translation insures a safeguard of potential mistakes in the initial franslation. Howevet it is testing instruments. Th important to emphasize that back-translations should provide a general, rather than a specific of translating the Wecl check on the initial tanslation. Avoiding a back-translation plaies undue emphasizes on the A final issue is thr original hanslation. example, to have the s According to Brislin, there are several basic steps that should be used: digits is a very difficul syllables. Furthermore. 1. Initial translation Nebraska has numerou 2. Back translation no sense when transla 3' Resolution of differences between the original English version and the resulting Spanish version Translators More recently, Muniz and Hambleton (1996) published a summary of the International Test Commission's guidelines for the translation and adaptation of tests. The ethical guidelines are If a test is publish as follows: is necessarv is to use a ASSESSMENT OF HISPANICS 95 'chologists in the United . General guidelines for professional behavior rbership of the Hispanic . Competence .sts could be considered . Responsibility (as ologists evidenced by . Safety of the test materials practitioners rather than . Confidentiality paucity in the research The general guidelines for appropriate use of tests include:

1. Understanding the testing situation 2. Selection of the appropriate tests 3. Consideration of potential biases imply the ftanslation of a 4. Adequate preparation for the testing situation ting the test is simply the 5. Adequate use of the tests For example, the Minne- 6. Appropriate scoring methods st common test used by 7. Appropriate interpretation r translated into Spanish. 8. Appropriate communication of findings bgroups and there are no 9. Revision of the test logical problems facing a lered include translation. Appropriate translation and adaptation requires much time and expertise. For example, in nitive equivalence. the translation of the Luria-Nebraska Neuropsychological Battery (Puente, Cespedes, kuar- izaga, Cano, & Tobal, 2000) care had to be taken to use personnel versed in both languages; what Harris, Cullum, and Puente (1995) have called balanced-bilinguals. In addition, those individuals had to have expertise not just in Spanish but also in psychology. While it is not 'chological tests to Span- difficult to find balanced-bilinguals, it is difficult to find individuals who are truly bilingual and t a Spanish speaker is a have the psychological knowledge necessary to understand the subtleties of testing. Also, the e,1997). Anecdotal evi resolution of discrepancies of the original and initial translation require special understanding Society reveals that most of issues that will later be addressed under the rubric of cognitive equivalence. heir own personal use or, Another problem of a serious nature in working with Hispanics is that, as discussed inisfered, much like what previously, this ethnic group actually includes different subgroups (Shorris, 1992). As in the s at the annual meetings, case with the Luria-Nebraska, the initial translation would have worked well with Cubans and ple, norms are not avail- Puerto Ricans because a Cuban completed it. The second Eanslation, completed by Spaniards, s case the suggestions by would have been adequate for Castillian speakers. The final translation attempted to blend not o be traditional, there are only the those two translations but also use words and phrases that were generic to Spanish y research started on the speakers. This required avoiding specific terminology that would be applicable to one culture. condly, the use of back- Standard Spanish was used, and this is indeed the best solution in developing and translating 'anslation. However, it is testing instruments. The problem of translating became even more evident in the initial efforts ral, rather than a specific of translating the Wechsler intelligence scales into Spanish (Puente & Salazar, 1998). rndue emphasizes on the A final issue is that a translation is not just a translation. Sometimes the intention is, for example, to have the similar number in Spanish represent a number in English. Translating be used: digits is a very difficult task. Eight in English is one syllable whereas in Spanish ocho is two syllables. Furthermore, sometimes a literal translation simply does not make sense. The Luria- Nebraska has numerous proverbs and phrases, such as "the golden egg," which literally make no sense when translated. Hence, care must be taken to address the criterion in question. ersion and the resulting

Translalors of the International Test fhe ethical guidelines are If a test is published in English, then one could easily and incorrectly assume that all that is necessary is to use a tanslator. This simple approach is riddled with unforeseen difficulties. 96 AIITOMO E. PUENTE AND ALF.REDO ARDILA

Translators are not necessarily well versed in psychology or medical principles. As a conse- quence, very literal but nonsensical Learning Test. Loewens translations are produced. Thus, the fatient views the question as literally not making have also worked to de any sense. Further, the translator might speak Spanish but be unfamiliar with Maybe it is unreas< Spanish culture. Next, there is the issue of time. lian item is to be timed, timing may get convoluted with neuropsychological test figuring out the translation. Rapport is decreased when a third party acts as an intervening typically available. Ho variable. Hence, the patient may view the evaluation and evaluator in a less trusting values should be consir manner. Finally, subtleties wili be missei if a translator is used. This could range from immediate nonverbal cues to more complex language responses, which are bound to be "translated out" when the final response is provideJ. Cultural Meaningfuln'

When administerit Norms and Sampling cultural context, the iss addressed. Items devel< lated to another culture. Neuropsychological tests require a reference value. This reference value is typically in the form of a normative simply unfamiliar and table. This is one reason why flexible batteries are often considered to be problematic. mal), whereas others l In other words, how does a unique set of tests compare between patients. interpreted as an oven Whereas neuropsychological tests have traditionally been weak in this regard, Golden, puri- sch, speakers because it is u and Hammeke (1979) provided norms for the Luria-Nebraska whereas both Reitan (Reitan in the Picture Arangen & wolfson, 1994) and more recentry Heaton and colleagues (Heaton, Grant, & Matthews, Cultural familiarity anr 1991) have provided useful norms for the Halstead-Reitan nattery. Almost all tests published by the major publishing houses now include normative data. Equivalencc Perusal of all of these normative sources, including two recent compendium books Cognitive (Mitrushina, Boone, & D'Elia, 1999; Spreen & Strauss, l99g), suggests thai while age and education (though Assuming that al not always) are taken into account, ethnic status almost always-is not considered. becomes whether Eng As discussed eadier, this omission may reflect the overall belief that ethnic status "Translation," the quer or culture is not important in understanding brain function and dysfunction. However, as andocho-one versus argued in this chapter, in the rest of this volume, and in other scholarly publication outlets, this test depends not on wt omission introduces unnecessary error into the evaluation procedure. from a cognitive one. I If this premise is held to be true, then most neuropsychological tests provide faulty reference understanding must be values. Samples that are used for normative tables rarely include ethnic status. This for each item, scale, a is indeed the case with the two historically used batteries in neuropsychology, the Halstead- measure a variety of Reitan and the Luria-Nebraska, as well as all the tests that frJquently used -ort by recognition and recall. neuropsychologists (from a large-scaled survey completed by camara,"t" Nathan, & puente, in press). is to determine what is One could argue that because ethnicity is not reported in the sampling procedure, the then proceed accordin norms could be considered invatd. difficulty. For examplt Another related issue is that most neuropsychological tests do not have norms in Spanish. whereas in Ardila, Ror Few commercially available tests in Spanish contain Spanish-speaking norms. The Wechsler frequency of occurrenr scales have been translated and have nonns. However, the norms *" iro- puerto Rico only. requiring digits in the Ardila, Rosselli, and Puente (1994) include in their book norms for over a dozen different tests. single-digit number o However, with most of the tests, the sample is limited to individuals living in Colombia. estingly, when a letter Further' most of the individuals representing the normative sample are over the age of 50. occasionally adjectiver However, one particularly positive aspect of this set of norms is the inclusion of both literates grammatical words in and illiterates. Ponton and colleagues (Ponton, Satz, Herrera, & OrEz, 1998) have recently published script). The reason for norms for a screening test. The nonns, obtained from residents in Los Angeles, may prove internal representation to be more promising in that a variety of ethnic groups are included. Howevir, as witl constructed. It may be other normative samples, no nonns are found for the different Hispanic subgroups proposed in grammatical elements this chapter. This pattern is found also with other studies, such aJ Harris, Cullum, and puent" the answer to these qut (1995) which reported preliminary results of a Spanish translation of the California Verbal the necessary framew ASSESSMENT OF HISPANICS 97 ,al principles. As a conse- Learning Test' Loewenstein and colleagues (Loewenstein, rus, the patient views the Rubert, Arguelles, & Duara, 1995b) have also worked to develop norms for Spanish speakers, ight speak Spanish but be especially for older populations. Maybe it is unreasonable to expect norms for If an item is to be timed, different Hispanic zubgroups wien 1t; most neuropsychological tests are not adequately translated Spanish, :is decreased when a third into and (2) norms are not typically available. However, the goal ethnic-group e evaluation and evaluator of sampling and normative reference values should be considered as new sfudies rslator is used. This could are being formulated. sponses, which are bound Cultural M e anin gfulne s s

When administering and./or translating tests developed in a specific culture to a different cultural context, the issue of equivalence in cultural meaningfuh-ss and relevance should be addressed. Items developed in a cultural context do not have the same relevance when trans- lated to another culture. The Boston Naming Test ence value is typically in is a good example. Several of the figures are simply unfamiliar and meaningless for Hispanics (e.g., the beaver ies are often considered to is a North American ani- mal), whereas others have a quite different degree )mpare between patients. of familiarity (e.g., igloo, frequently interpreted as an oven by Hispanics). Igloo is a very this regard, Golden, Puri- easy and familiar word for Englisi speakers because it is used at the elementary school for learning ska whereas both Reitan the sound /i./). Several stories in the Picture Arrangement subtest of the WAIS are perceived tgues (Heaton, Grant, & by Hispanics as stange and rare. Cultural familiarity and meaningfulness can result in differences in performance. n Battery. Almost all tests data. cent compendium books Cognifive Equivalence 3gests that while age and Assuming that all the preceding variables are held constant, tus almost always is not then the real question becomes whether English and Spanish tests are measuring the same thing. ill belief that ethnic status In the section "Translation," the question of translation equivalence was addressed and the example of eight lysfunction. However, as and ocho-one versus two syllables-was presented. When a test is translated the value ly publication outlets, this of the test depends not on whether a translation is viable from a language lre. perspective but, instead, from a cognitive one. Before a tanslation is completed or gical tests provide faulty before a test is administered a clear understanding must be had of what is to be measured. This requires that the underlying factor nclude ethnic status. This for each item, scale, and test be understood. As Anderson (1996) has indicated, sychology, the Halstead- most tests measure a variety of things-ranging from immediate attention to verbal articulation most frequently used by to recognition and recall. Rarely does a test measure just one cognitive domain. The ara, Nathan, & Puente, in obvious step is to determine what is the most important or salient variable that needs to be measured and : sampling procedure, the then proceed accordingly. In addition, one must then be careful that the item is of equal difficulty. For example, one of the most common tests for fluency in English is the ot have norms in Spanish. R A, S, whereas in Ardila, Rosselli, and Puente (1994) norms are provided for this test, ing norms. The Wechsler because tle frequency of occurrence for these letters is not the same in English as in Spanish. In the tests re from Puerto Rico only. requiring digits in the WAIS and WISC, the question is whether the task is to remember a ver a dozen different tests. single-digit number or a single-digit number with a specified number syllables. uals living in Colombia. of Inter- estingly, when a letter fluency test is administered to Hispanics, they mainly report nouns, and e are over the age of 50. occasionally adjectives and verbs, but do not say grarnmatical words. English speakers report inclusion of both literates grammatical words in the letter fluency test (M. Rosselli & A. Ardila, unpublished )rtiz, 1998) have recently manu- script). The reason for this difference is unclear, but it may suggest a subtle difference in the lents in Los Angeles, may internal representation of words, i.e., what is understood as a word and how semantic nets are cluded. However, as with constructed. It may be conjectured that this difference relates to the clearer distinction ric subgroups proposed in between grammatical elements existing in Spanish relative to English. Outcome studies will determine arris, Cullum, and Puente the answer to these questions. For now, clinical experience and theoretical speculation provide of the California Verbal the necessary framework from which to consider these important questions. 98 ANTONIO E. PUENTE AND ALFREDO ARDILA

Interviewing education and sociocu (Ardila, Unfornrnately, a comprehensive leagues Brydr review of the psychological literature revealed that no empirical articles were published adolescent and adult A on interviewing Hispanics-in neuropsychological evalua- tions' (1990) reponed norma The few articles that exist are based on clinical, anecdotal, or theoretical perspectives. Examination. This typt For example, Puente and Perez (2000) have provided some background for structuring of the Spanish Speaker interviewing. Considering the complexities associated with cultural issues and the possibilitf hensive publication of missing key elements to in the interview, a structured approach will increase the likelihood of addressing the important domains that were tra issues in question. However, it is important to appreciate the importance abilities. The test resr of establishing rapport and explaining the purpose of the evaluation to Hispanics. discussed, both of thes, This need is based on prior limited encounters with mental or medical professionals. r'urttrer, interesting to note tha mental concerns may be private matters which are family and religious focused. Thus, opening could not be reproduce up to a stranger is abnormal. And when opening up involves talking about personal-issuei (such publish extensively wi as memory), the interview will hold special challenges and diftculties for the profes- Another successfi sional. Frequently, intellectual or cognitive testing rn"y U" perceived by the puti"nt u, series of studies durinE aversive. In Latin America, highly educated people usually dislike, and try io avoid, testing. Loewenstein, Duara, 1 Intellectual testing may be even be perceived as humiliating ,,Ar1 and disrespectful of privacy. and colleagues have I you taking me for stupid?" is frequently the implicit message of some Hispanici when elderly Spanish speak rejecting testing. People with little education, on tle other hand, may be afraid and embar- subtests along with thr rassed when tested. In consequence, testing can be much more effective if performed in a capacity. They have al flexible and informal way rather than using a rigid and highly standardized situation. assessing neuropsycho that the Fuld object M Testing speakers. Taussig and Pont As previously mentioned, there are few research programs associated with neuropsy- translation of the WA chological assessment. A review of the neuropsychological testing literature reveals four reported good differenr cities; Denver, Los Angeles, Miami, and wilmington, North carolina. In Denver, Harris has Ponton, Satz, Herrera, focused on the development of a Spanish version of the California Verbal Learning Test. In subjects ranging from Los Angeles, Ponton and Taussig, in separate and collaborative efforts, have published Screening Battery for important information on Spanish speakers. In Miami, the work of Ardila and Rosselli as well variety of neuropsych, as Loewenstein and colleagues (primarily with the elderly) has been ongoing for over a A series of unreli decade. In Wilmington, Puente and colleagues (primarily from universities in Madrid and involving verbal fluen Granada, Spain) have worked on a variety of issues including the translation and adaptation of effect of bilingualism the Wechsler (especially scales with Ardila and Harris, from Colorado). This section reviews Spanish translation (n the existing literature and ongoing research efforts on the neuropsychological testing of the California Verbal Lear Spanish speaker. speakers participated ir One of the firstpublications in the literature (Dergan, 1987) reflected the ongoing status of significant differences Hispanic neuropsychological assessment. In an article in Avances en psicologica Clinica Perri, Naplin, anr Latinamericana, Dergan describes an initial translation of the Luria-Nebraska Neuropsy- Verbal Learning and N chological Battery' Although the translation violates many of the principles previously out- categories of 4 words lined in this chapter and no empirical data are reported, this article could be considered the first Spanish-speaking adul in the field. During the 1980s, a series of articles by Ardila and colleagues began to appear in that verbal learning an the literature (e.g., Ardila & Rosselli, 1989; ostrosky, canseco, euintana, Narvarro, & Ardila, ized instruments. Hou 1985). Ardila's primary collaborators have been Monica Rosselli in Colombia and Miami. and than might initially ap1 Ostrosky in Mexico City. However, it was not until 1989 that L. Bernard first reported that Olazaran, Jacobs, and "blacks" and Hispanics of average intellectual functioning but from poor academic back- digit string might have grounds (and no reported brain impairment) exhibited scores that were in the impaired range. WAIS-R digit span. M A series of studies was conducted by Ardila and colleagues examining the relationship of ported cross-ethnic diffr ASSESSMENT OF HISPAMCS 99

education and sociocultural status on brain functioning. As early as lggg, Ardila and col- leagues (Ardila, Bryden, & ostrosky, 1989) reported iterature revealed that no on the incidence of handedness in adolescent and adult Amazonian jungle residents. uropsychological evalua- Also, Rosselli, Ardila, Florez, and Castro (1990) reported normative data on a ' theoretical perspectives. Spanish translation of the Boston Diagnostic Aphasia Examination' This type of work resulted in :kground for structuring the publica tion of Neuropsychological Evaluation of the spanish speaker by Ardila, Rosselli, issues and the possibility and Puente (lgg4).The book was the first compre- hensive publication to present a variety of neuropsychological increase the likelihood of tests across several cognitive domains that were translated into Spanish, ranging lortant to appreciate the from tests of attention to visuopraxic abilities' The test results were stratified : evaluation to Hispanics. according to age and education. As previously discussed, both of these variables were associated al professionals. Further, with neuropsychological performance. It is interesting to note that some of the tests and test ts focused. Thus, opening materials le.g., wectrsler Memory Scale) could not be reproduce{ in- the book due to copyright ng about personal issues restrictions. The group has continued to publish extensively with their primary ifficulties for profes- base of oferation being Miami. the Another successful research program, also in Miami, has eived by the patient as f,een that of Loewenstein. In a series of studies during the 1990s (Loewenstein, Arguenes, and try to avoid, testing. Arguelles, & Linn-Fuentes, 1994; Loewenstein, Duara, Arguelles, & Arguelles, ,spectful of privacy. "Are 1995a; Loewensiein et ar., r995b) Loewenstein and colleagues have focused on developing neuropsychological f some Hispanics when insrruments applicable to elderly Spanish speakers. They have reported ny be the wAIS Btck Design ano oijit Symboi afraid and embar- subtests along with the Mini-Mental State Examination were strong predictors .ective if performed in a of functional capacity' They have also been supporters of the use rdardized situation. of culturally riu"eo or limited tests for assessing neuropsychological functioning. For example, Loewenstein et al. (1995a) reported that the Fuld object Memory Evaluation was a useful tool for detecting dementia in Spanish speakers. Thussig and Ponton have also published another series of studies. socialed Using a customized with neuropsy- translation of the wAIS-R, Taussig and colleagues (Taussig, Mack, & Henderson, 1996) g literature reveals four reported good differentiation of controls to mildly-to-moderately demented Spanish speakers. n. In Denver, Harris has Ponton, satz, Herrera, ortiz, et al. (1996) have provided normative data on 300 Hispanic Verbal Irarning Test. In subjects ranging from the age of 16 to 75 for the newly developed Neuropsyctrotogica efforts, have published screening Battery for Hispanics. Gender, age, and education urroiiution, were seen rdila and for a Rosselli as well variety of neuropsychological tests. een ongoing for over a A series of unrelated articles have been published during the second half of the 1990s iversities in Madrid and involving verbal fluency and learning. Harris, Cullum, and iuente (1995) investigated the slation and adaptation of effect of bilingualism on verbal learning and memory, using a scientifically constructed lo). This section reviews Spanish translation (most existing Spanish forms are not scientifically constructed) of the ohological testing of the California verbal Learning zl4 Test. A total of Hispanics bilingual and.22monolingual English speakers participated in this study. When groups were assessed in their dominant languagi, no :ted the ongoing status of significant differences were found. en Psicologica Clinica Perri, Naplin, and Carpenter (1995) reported on the development of the perri Test of ia-Nebraska Neuropsy- verbal Learning and Memory (in Spanish). The test consists of a i6-word list composed of 4 inciples previously out- categories of 4 words each together with a 40-word recognition list. Normative data for 100 ld be considered the first Spanish-speaking adults ranging from 15 to 70 years of age were presented. Hence, it appears tgues began to appear in that verbal learning and memory can be measured with appropriaiely translated and standard- ana, Narvarro, & Ardila, ized instruments. However, it is important to note ttrat verUat *".ory is more complicated olombia and Miami, and than might initially appear and that normative data are inconclusive at tlis point. For example, :rnard first reported that Olazaran, Jacobs, and (1996) Stern reported that the differences in the number of syllables per m poor academic back- digit string might have been responsible for an observed decrement in Spanish speakers on the :e in the impaired. range. wAIs-R digit span. More recently, Kempler, Teng, Dick, Taussig, andbavis (199g) have re- Lining the relation-ship of ported cross-ethnic differences ftetween Vietramese and Spanish speaken) for fluency measures. ANTOMO E. PUENTE AND ALFRED0 ARDILA

other studies are apparently in progress. For example, Munges (1996) reported a work in a HisPanic net progress Development of at the University of California at Davis. The study focrised on the development of a needed to understand the neurc with strong psychometric characteristics !1""y in both English and spanish across 12 majority culture (on anY continr different cognitive dom-ains. The pittiran, ongoing work of Stricks, Jacobs, Sano, and Stern in increasing th (1998) assist not onlY probably best reflects the current status of the field. In their attempt to obtain norms for evaluation of the SPanish sPeake English and Spanish-speaking elderly, a battery of neuropsychological tests was administered neuropsychology. It is anticipate to almost 1,000 participants. Their findings indicate, as do many of-the studies reviewed in this point for expanding neuroPsYch chapter, that age, education, and language are associated with neuropsychological test perfor- cal. SpecificallY, a more comPrel mance' What is important to note, however, is that there is increasini evidenci to suggest that serve as an initial steP for qur nonverbal tests are not culturally reduced. Indeed, it has long Ueen nletA true that tests that are function in evolutionary focus.' translated are high in cultural confound. However, increasing evidence has been recently glimpses into the kinds of quest reported that identifies similar problems, possibly even at ulo*", level, in tests that arl example, Does a commonalitY nonverbal. psychological "g," exist? What life-span? Is neuroPsYchologic Interpretation cultural traits? And, if so, could cognitive acculturation? Could Most of the research reviewed, whether theoretical, clinical, or empirical, focused on the an individual's cognitive and e appropriate development of neuropsychological instruments. However, interpretation is the knowledge be the staning Poj important element of the standardized neuropsychological assessment. condition? After all, with over Taking information out of context invariably will result in increased error variance. more than a strong methodolog: Taking psychometric data without regard to the issues previously addressed is incorrect and Conceivably, neuroPsYchologS unethical. For example, care must be taken to understand the limiis of the test, the patient, and these heretofore unanswered g the evaluator in determining what the psychometric findings really mean. Hence, the clinician There is little question tha has to work doubly hard to appreciate the unusual nuances that are being faced with the Most members of APA s Div Hispanic neuropsychological patient. However, the psychometric results must be framed in a proportions not tYPical of othel complex biopsychosocial context (see Puente & McCaffrey, Lgg2). This would include not mix of ethnic minorities in the 1 only an appreciation of the limits of the test but an understanding of the purpose of the ethnic minorities in the field w< examination (e.g., learning disability placement) and the context of the patieni aithe time of dia, Harris, Congett, and Pue the evaluation (e.g., how many years in the United states). Arnold, Montgomery, castaneda, important to understand cultt and Longoria (1994) reported on the association of acculturation to performance on the Clearly, one way to resolve thi Halstead*Reitan Neuropsychological Battery. The authors reported that acculturation levels bilcultural. UnfortunatelY, thr correlated with the results of the following: Tactual Performance Test, the Seashore Rhvthm Neuropsychological SocietY ir Test, and the Halstead Category Test. point, is not very active. Also' As argued previously (Puente &Perez,2000), the issue in question may not necessarily Hispanic is a fellow of the divi be the neuropsychological status of the patient. The question may be whether the patient (one way to estimate HisPanic cognitively acculturated into the culture from which the principles are being derived lor the While the senior author of th foundation of the evaluation. As discussed earlier, -uoy oeutopsychological tests are timed. minority affairs, its imPact h For Hispanics, time is not the same thing as for North Americans (Shorris, 1992). Hence, one present, there is little to sugg would expect to see a greater percentage of false positives in situations when the psychometric Another Possible alterna results are not interpreted in a broad biopsychosocial context. suggested, is to increase awan the need for multicultural unc neuropsychologY as well. Tht Suggestions and Potential Solutions with the first one occurring t Society has met irregularlY at Neuropsychologists are beginning to address the problems that arise from the ever- the number of non-HisPanics expanding population of Hispanics in both the United States and Latin America. However, due to the lack of a regularl' Hispanics as an ethnic group pose challenges that could serve as a foundation for conceptualiz- Tests for SPanish sPeakt ing the understanding of other ethnic rninorities both in and outside the United States. chologists. Outside the work ASSESSMENT OF HISPANICS l0r

| (1996) reported a work in Development of a Hispanic neuropsychology would serve but a small amount of what is d on the development of a needed to understand the neuropsychological function of minority individuals living in a h and Spanish across 12 majority culture (on any continent). In this section, several issues are presented that might r, Jacobs, Sano. and Stern assist not only in increasing the knowledge necessary to address the neuropsychologi-al ttempt to obtain norms for evaluation of the Spanish speaker but to lay the framework for larger issues involving cal tests was administered neuropsychology. "utt*A It is anticipated that this foundation would, in turn, later serve as a starting he studies reviewed in this point for expanding neuropsychological knowledge beyond the clinical to the more theoreti- rpsychological test perfor- cal' Specifically, a more comprehensive understanding ofculture and neuropsychology should g evidence to suggest that serve as an initial step for questioning the possibility of the role of niuropsychological reld hue that tests that are function in evolutionary focus. The work of Sperry Qgg4),Wilson (1995), and others pro-uia" idence has been recently glimpses into the kinds of questions that heretofore neuropsychology has not considered. For er level, in tests that are example, Does a commonality across cultures in neuropsychological functioning, or neuro- psychological "g," exist? What role does culture play in shaping brain functioning across the life-span? Is neuropsychological function nothing more than the cognitive expression of cultural traits? And, if so, could neuropsychological analysis be nothing more than measuring cognitive acculturation? Could neuropsychological knowledge serve to understand more than empirical, focused on the an individual's cognitive and emotional status after brain injury? Could neuropsychological ever, interpretation is the knowledge be the starting point for a more comprehensive understanAing of the human rent. condition? After all, with over a century of research, psychology has yet to produce much increased error variance. more than a stong methodology. "Why" questions have lagged way behind "h;w" questions. ddressed is incorrect and Conceivably, neuropsychology, especially from a cultural context, could serve to address rfthe test, the patient, and these heretofore unanswered questions. nean. Hence, the clinician There is little question that neuropsychology is a very traditional psychological specialty. are being faced with the Most members of APA s Division 40, clinical Neuropsychology, are white men, and in :sults must be framed in a proportions not typical of other APA divisions (Puente & Marcotte, in press). Considering the t. This would include not mix of ethnic minorities in the population at large, it would seem that increasing the numbir of ng of the purpose the of ethnic minorities in the field would be an obvious process for the field to encourage. Echemen- 'tle patient at the time of dia, Harris, Congett, and Puente (1997) reported that neuropsychologists do believe it is Montgomery, Castaneda, important to understand cultural issues, especially when dealing with Spanish speakers. r to performance on the Clearly, one way to resolve this is to increase the number of individuals who are bilingual and I that acculturation levels bilcultural. Unfortunately, there is limited evidence that this is occurring. The Hispanic est, the Seashore Rhythm Neuropsychological Society is comprised of about 50 members and the organization, at this point, is not very active. Also, a review of all the members of Division 40 revealed only one stion may not necessarily Hispanic is a fellow of the division and the total number of individuals with Hispanic surnames y be whether the patient (one way to estimate Hispanic membership) approximately LVo of theclose to 4,000 members. are being derived for the While the senior author of this chapter helped organize a Division 40 committee on ethnic hological tests are timed. minority affairs, its impact has been highly limited. The situation needs quick remedy. At horris, 1992). Hence, one present, there is little to suggest that the pipeline of Hispanic students is being reversed. ns when the psychometric Another possible alternative, as the responders of the Echemendia et al. (1996) survey suggested, is to increase awareness of the issue in question. APA Ethical Guidelines emphasize the need for multicultural understanding and, presumably, this would obviously translate into neuropsychology as well. The authors of this chapter have presented two workshops at NAN, with the first one occurring during the early 1990s. Also, the Hispanic Neuropsychological Society has met inegularly at NAN, APA, and INS meetings. However, in almost all instances hat arise from ever- the the number of non-Hispanics attending these meetings is always very, very small, in large part America. However, -atin due to the lack of a regularly published newsletter or a very active group of officers. mdation for cogceptualiz- Tests for Spanish speakers need to be developed and made widely available to neuropsy- rtside the United States. chologists. Outside the work by Ardila (e.g., Ardila, Rosselli, & Puente, 1994), the few things 102 ANTONIO E. PUENTE AND ALtrREDO ARDILA

that are published are research reports with minimal normative data (except for ponton et al, just 1997). The major test publishing house ttran majority individr in Spain (TEA) has made several tests (e.g., stroop) available but almost none of the tests developed countries represl have nlrms, The first neuropsychological test develop- ment project the company sponsored, text, neuropsychology has the Luria-Nebraska N"*ffi"nological Battery, will be published in the early part of 2000 (with an N of over 300). n*ing the past decade, the Psychological corporation has been interested in developing sp*irr, translations of the wechsler scales (already described in this chapter). References under the dlrection of the senior author, a yorkgroup eventually was convened which included several of ;;;r;;rent researchers both in-neuropsychology Anderson, R. M. (1994). Practiti and school psychology (including one of the puerto authors of the Rican tudila, A. 0998a). A note of car adaptation of the wrsc). After a numbeior years oflre[minary planning, a translation was gender and ethnicity: A con eventually developed. After the painstaking work on diveloping thi ;rhtion, a tryout phase Ardila, A. (1998b). Bilingualism: was initiated' Due to the.low number tudila, A., Ardila, O., Bryden, of responses and potentialludgetary reasons, the project handedness. N europ sycholo has been placed on hold for the foreseeabli 1 future. The commerciily available tests that are currently tudila, A., & Rosselli, M. (1989). available with norms (e.g., WAIS and WSC) are very outdateo and included norms ogy, 5, 307 -320. essentially only from puerto Rico. psychotogica Recently, the corporation has begun to Ardila, A. Rosselli, M., & Puentt distibute in the United states a brief neuropsychological t"i b"tt"ry Lorun as NEURopSI Plenum. (ostrosky-Solfs, Ardila Rosselli, Ardila, A., Rosselli, M., Ostrosk & 1997) developed in Spanish anj normalized in Mexico. syntactic comprehension in Amold, B. R., Montgomery, G. T. the united states, there is no question thafserious problems ar! being faced by neuropsycholo- selected Halstead-Reitan nr gists wishing to evaluate_Hispanics. some practical suggestions inclide: l) whenever possible, Bernard L. C. (1989). Halstead- refer to a bilingual and bicultural neuropsychotogist;-i; if possible, use appropriately trans- adult males from poor acad lated tests; 3) if possible, use Berry, J. W. (1997). lmmigratior applicable norms; 4; if necessary, use atranslator and preferably (1983). a hained and unbiased one; 5) understand the limitation, Brislin, R. W. Translation of th" evaluation procedure; and Handbook of cross-c 6) identify perceived limitations in the @ds.) context of the report and contextualize the overall Camara W., Nathan, J., & Puente findings within the limitations of the evaluation procedure. Dergan, J. J. (1987). La bateria ne At this point, tle situation does not appear to have many potential remedies in the 5.2-36. immediate future. Whereas there is clearly Diccionario d.e la lengw Espaft a dimonstrated need tt field considers this a -o Echemendia, J., Harris, J.C. C critical issue, there is little to suggest that the traditional " R. approaches to testing Spanish practices with Hispanics: A speakers will be changed. Hence, translators, idiosyncratic translations, and no norms will Golden, C. J., Purisch, A,D., & E continue for the foreseeable future unless the field Legins to address more aggrcssively the for clinical and experiment issues in question. Grosjean, F. (1989). Neurolingu Language, 36,3-15. Hanis, J., Cullum, M., & Puente Conclusion ad:;lts. Journal of the Inter Heaton, R., Grant, L, & Matthe chological Battery. Demogt Tlvo issues should be Assessment evident from reading this chapter. one, there is an increasing need logical Resour to understand the uniqueness Kempler, D., Teng, E. L., Dick, I associated with the neuropsychological evaluation of the-Span- verbal fluency. Joumal of t ish speaker. Demographic characteristics indicate that the iargest!*rnic minority in the united Kilborn, K. (1994). Learning a States by the eady part of the twenty-first century will be Hispanics. In some populous states, H andbo ok of p syc ho lingui : such as California, Texas, and Florida, Hispanics will be the majority very soon. The second Itzak, M.D. (1995). Neuropsyc issue is that there is no evidence Loewenstein, D. A., Arguelles,' to suggest that neuropsychology is prepared to meet that pressing need. In fact, using chological assessment of th Division 40 membership and publishJi tests with norms as rough Loewenstein, D. A., Duara, R., r guidelines, the field is nowhere near adequately addrlssinjthis problem. whether the urru*!- the detection of mild demet tion is that culture is not important to brain function or wtrittrer ii is that culture could be easiiy J. 300-307. held comtant with idiosyncratically translated tests with no norns is not clear. Regardless, the Loewenstein, S. A., Rubert, M. resolution of this problem prediction would also further the generalizability of neuropsychology beyond of functional c the clinical realm. Archives of Clinical Neurc Such knowledge would go a long way in helping addresi t1" porriuitity oru neuropsychological Manuel-Dupont, S., Ardila, A.,I "g" and in the development of a neuropsycn6togy that is usable to more @ds.), Handbook of neurc ASSESSMENT OF HISPAMCS 103 ta (except for Ponton et al, just than majority individuals from well-developed nations. Majority individuals several tests (e.g., Stroop) of well- developed countries represent no more than 10Vo of the world population. global rsychological test develop- In this con- text' neuropsychology has been almost exclusively directed to the study of world rsychological Bat0ery, will minorities. )uring the past decade, tle panish translations of the References :tion of the senior author, a prominent researchers both Anderson, R. M. (1994). Practitionerb guide to clinical neuropsychology. New york: plenum. ruthors of the Puerto Rican Ardila, A. (1998a). A note of caution: Normative neuropsychological test performance: Effects of age, education, planning, a franslation was gender and ethnicity: A comment on Saykin et al. (1995). Applied Niumpsychology,j, 5l-53. (1998b). l translation, a tryout phase Ardila, A. Bilingualism: A neglected and chaoric area. Aphasiology, 12, l3l-134. Ardila' A.' Ardila, O., Bryden, lgetary reasons, the project M. P., & Ostrosky, F. 0989). Effects of culrural background and education on handedness. Neuropsychologia, 27, 893-997. ally available tests that are Rosselli, (1989). Ardila" A.' & M. Neuropsychological characleristics ofnorm alaging. Developmental Neuropsychol- rtdated and included norms ogy, 5,307-320. Corporation has begun to fudila A. Rosselli, M., & Puente, A. E. (1994). Neuropsychological evaluation of the Spanish speaker. New york: ery known as NEUROPSI Plenum. Ardila, A., Rosselli, M., ostrosky-soliis, md normalized in Mexico. F., Marcos, J., Granda, G. & soto, M. (in press). Memory abilities and syntactic comprehension in Spanish-English bilinguals. Applied Neurcpsychology. rber of Spanish speakers in Arnold' B. R.' Montgomery, G. T., Castaneda, I., I-ongoria, R. (1994). Acculturaltion and performance of Hispanics on tg faced by neuropsycholo- selected Halstead-Reitan neuropsychological tests. Assessment, l(3), Z3g_44g. (1989). lude: l) whenever possible, Bernard' L. C. Halstead-Reitan Neuropsychology Test performance of Black, Hispanic, and White young adult males from poor le, use appropriately trans- academic backgrounds. Archives of Clinical Neuropsychology, 4, 267 -274. Berry, J. W. (1997). Immigration, acculruration and adaption. psychology,46, a translator and preferably Applied 5_6g. R. (1983). Brislin' W Translation and content analysis oforal and written material. In H. C. Triandis and J. W Berry evaluation procedure; and (Hs.) Handbook of cross-cultural psychology methodology. Boston: Allyn & Bacon. I contextualize the overall Camara W., Nathan, J., & Puente, A. E. (in prcss). hofessional usage of psychological tests. prcfessional psychology. Dergan' J. J. (1987). La bateria neuropsicologica Luria-Nebraska. Avances en Psicological Clinica Latinoanuricana, potentii remedies in the 5,2-36. Diccionario de la lengua Espaftola (1984). (Vig€sima Ed., Tomo II). Madrid: rd the field considers this a Real Academia Espaffola. Echemendia, R. J., Harris, J.C. Congett, S. M., Diaz, M. L., & Puente, A. E. (1997). Neuropsychological training and oaches to testing Spanish practices with Hispanics: A national survey. The clinical Neuropsychologist, II, zzg-243. lations, and no norms will Golden, C. J., Purisch, A,D., & Hammeke, T. A. (1979) The Luria-Nebrasl

Merriam-websterb collegiate Dictionary (1996). (6th ed.) Springfield, MA: Merriam-webster. Mitrushina, M. N., Boone, K. 8., & D'Elia, L. F. D. (1999). Handbook of normative data for neuropsychological asEessment. New York: Oxford University hess. Munges, D' (1996). process The of development of valid and reliable neuropsychological assessment measures for English- and Spanish-speaking elderly persons. In G. Yeo & D. Gallagher-Thompson (Eds.), Ethnicity and the dementias (pp. 33-46). Washington, D.C.; Taylor & Francis. Muniz, J., (1996). & Hambleton, R. K. Directrices para la traducion y adaption de los tes rs. papeles del psicologo,66, 63-70. Neuropsycl Olazaran J., Jacobs, D. M.' & Stern, Y. (1996). Comparative study of visual and verbal short-term memory in English and Spanish speakers: Testing a linguistic hypothesis. Joumal of the Intemational Neuropsychological Soiiety, 2, 105-110. Interv, ostrosky, F., and Ardila, A., Rosselli, M., l,6pez-Arango, G., & uriel-Mendoza, v. (199g). Neuropsychological test performance in illiterates. Archives of Clinical Neuropsychology, l13, U5_660. Ostrosky-Sofis' F., Ardila, A., & Rosselli, M. (1997). NELIROPSI: Evaluacidn neuropsicoldgica breve en espaftol. with Nativr Manual, instructivo y protocolo de aplicaciAn [NEUROPSI: A brief neuropsychological evaluation in Spanish. Manual, instructions, and application protocoll. M6xico: Bayer de M6xico. Ostrosky, F., Canseco, E., Quintanar, L., Navarro,8., & Ardila, A. (1985). Sociocultural effects in neuropsychological JEFF KIN assessment. Intemational Journal of Neuroscience, 27, 53-66. Perez-Arce, P., & Puente, A. E. (1996). Neuropsychological assessment ofethnic-minorities: The case for assessing Hispanics living in North America. In R. J. Sbordone & C. J. Long (Rts.), Ecological validity of neuropsy- chological testing. Deh^y Beach, FL: St. Lucie. Perri, B. M' Naplin, N. A., & Carpenter, G. A. (1995) A Spanish auditory verbal leaming and memory test. As se ssment, 2, 245 -253. Introduction Ponton, M. O., Sau, P., Herrera, L., &Ortiz, F. (1998). Normative data scatified by age and education for the Neuropsychological Screening Battery for Hispanics (NeSBHIS): Initial report. Joumal of the International Neuropsychological Society, 2, 96-IC/. Much of the attention give Puente, A. 8., Cespedes, J. M., Iruanizaga I., Tobal, J. M., & Vindel, A. C. (2000). The Luria_Nebraska Neuropsy_ research literature is minimi chological Battery and culrurally dissimilar individuals. In c. J. Golden & L. warren (Eds.), lfte LNNB 20th disappointing number of nel anniversary handboolc Volume I: A guide to clinical interpretation and use in special settings. Los Angeles: Westem Psychological Services. 1970s on subjects such as Puente, A. E., & McCaffrey, R. J. (Eds.) (1992). Handbook of neuropsychological asiessmenr. New york: plenum. lateralization (Scott, Hynd, Puente, A. E. & Perez, M. (2000). Psychological assessment of ethnic-minorities. In G. Goldstein & M. Hersen (Eds.), Gestalt (Taylor & Thweatt, Handbook of psychological assessment (pp. 522-551). Boston: Allyn & Bacon. 1979). In the early 1980s sn Puente, A. E., & Salazar, G. (1998). Assessment of minority and culturally diverse childrcn. In A. Prifetera & D. scales and hemispheric spe Saklokske (Eds.) WBC-III: Clinical use and interpretation (pp.227-248). New York: Academic hess. Mishra, Lord, & Sabers, 191 Reitan, R. & Wolfson, D. (1984). The Halstead-Reitan Neuropsychological Battery. Tucson, AZ: Neuropsychology Press. are more reflective of bioet Rosselli, M., Ardila, A., Florez, A., & Casuo, C. (1990). Normative data on the Boston Diagnostic Aphasia Examina- status (Ferraro, Bercier, & tion in a Spanish-speaking population. Journal of Ctinical and Experimental Neuropsychology, 12,313-322. (Anderson, Bastida, Krame Rosselli, M. & Ardila, A. verbal fluency in Spanish and English. unpublished manuscript. Liu, & Schinke, 1995; Myer Shonis, E. (1992). Latinos. New York: Norton. Manson, 1995). Sperrl', R. W. (1994). Holding course amid shifting paradigms. In W. Hartman and J. Clark (Ed s.), The metaphysical are foundatiow of nodem science: Issues of causality (pp.99-123). San Francisco: Institute of Noetic Sciences. Native Americans Spreen, O. & Strauss, E. 0998). A compendium ofneuropsychological tests (2nd,ed.). New York: Oxford University although well over one-hal Press. numbers are comparatively Stricks, L., Pittman, J., Jacobs, D. M., Sano, M., & Stern, Y (1998). Normativa data for a brief neuropsychological pling (Myers, Kagawa-Sing battery administered to English- and Spanish-speaking community-dwelling elders. Jounral ofthe Internationa! tion is available about the h Ne uropsyc holo gical Society, 4, 3ll-318. the 60Vo Indi: Taussig, M. I., Mack, W., & Henderson, V. W. (1996) Concunent validity of Spanish-language versions of the Mini- of American Mental State Examination, Mental Status Questionnaire, Information-Memory-Concentration Test, and about how the other 407o ol Presentation-Memory-Concentration Test: Alzheimer's disease patients and non-demented elderly comparison. limitations of data gatherin Journal of the International Neuropsychological Society, 2, 286-298. Taussig, M. I., & Ponton, M. (1996). Issues in neuropsychological assessment for Hispanic older adults: Cultural and JEFI'KING . Native Arnerican linguistic factors. ln G. Yeo and D. Gallagher-Thompson @ds.), Ethzicity and the demenrias. Washington, D.C.: Colorado 80208. ELAINE I Taylor Francis. & Colorado 80921. U.S. Bureau of the Census (1999). United States population. Washington, D.C.: Author. Handbook of Cross-Culural Neu Mlson, Z. O. (1995). Naturalist. New York: Warner Books. demic / Plenum Publishers, New