|Foxcroft, C. D. (2002). Ethical
issues related to psychological testing in Africa: What I have learned
(so far). In W. J. Lonner, D. L. Dinnel, S. A. Hayes, & D. N.
Sattler (Eds.), Online Readings in Psychology and Culture (Unit
5, Chapter 4),
Center for Cross-Cultural Research, Western Washington University,
Bellingham, Washington USA.
This material is copyrighted by the author(s), who have kindly extended to the Center the right to use the material as described in the Introduction to this collection and the form entitled "Agreement to Extend License to Use Work."
UNIT 5, CHAPTER 4
ETHICAL ISSUES RELATED TO PSYCHOLOGICAL TESTING IN AFRICA: WHAT I HAVE LEARNED (SO FAR)
Cheryl D. Foxcroft
University of Port Elizabeth
This article attempts to introduce readers to what I have learned regarding ethical issues in psychological testing in Africa. The core ethical consideration of how to best cater for cultural and linguistic diversity so as to be sensitive to test-takers' cultural backgrounds and values during the testing process is highlighted. Thereafter, specific issues to consider and address before using tests in an assessment or research context are presented. This is followed by a discussion on the ethical issues that an assessment practitioner needs to deal with during and after testing. Throughout, examples relevant to the multicultural African context are provided to illustrate specific ethical issues.
Africa, which covers 20.28% of the land surface of the Earth, is the second largest continent and is currently comprised of 54 countries, with 650 million inhabitants. Lying on both sides of the equator, it is a continent of contrasting climates, geographical features, and plant and animal life. The peoples of Africa are of diverse ethnic origins and cultural backgrounds, speak a large variety of languages, and differ greatly in terms of their religious and political orientations, clothing, gestures, attitudes toward child-rearing and family life, the extent of their formal education, and levels of literacy. The inhabitants of Africa largely live in rural areas, with only three out of every 10 people living in cities. Being part of the third world, the human development index (HDI) of the vast majority of African countries is low. Conditions of extreme poverty and severe malnutrition are not uncommon, while the gap between the rich and the poor continues to grow. Readers are encouraged to expand their knowledge of the African continent by visiting the website of the Africa Institute (www.ai.org.za), which provides valuable information regarding the political, socio-economic, international and developmental issues facing contemporary Africa.
Given that Africa is a continent of great contrasts, any attempt to present African issues in a collective way runs the risk of failure. Keeping this in mind, this paper will map out the core ethical issues related to psychological testing that I have experienced (so far) as a psychologist living and working in Africa. What I have learned has come from assessing children, adolescents, and adults in Africa (and primarily in South Africa); dialoguing with colleagues, students, educationalists, employers, and key community informants; and from the writings of cross-cultural scholars in Africa and elsewhere in the world. Nonetheless, I am aware that although the map of the territory that I will provide you with should help you to conceptualize the ethical considerations encountered when practicing psychological testing in Africa, the map is essentially my map of the territory. It should thus only be seen as providing an outline, which needs to be enlarged, expanded and contextualized when assessment practitioners and researchers operate in a specific country.
The Core Ethical Consideration
As a helping professional, the psychological assessment practitioner's core ethical responsibility can be summed up as being that nothing should be done during testing and assessment to harm the client. The relationship between the assessment practitioner and the test-taker in many ways represents a power relationship, in which the practitioner holds most of the power, which leaves the client in a vulnerable position (Foxcroft, Roodt, & Abrahams, 2001). Given the imbalance of power, the assessment practitioner should ensure that the test-taker does not perceive the testing process and its outcome as being negative and disempowering because of unfair or unethical testing practices. It is the responsibility of assessment practitioners to ensure that they follow ethical practices that result in test-takers perceiving psychological testing to be fair and constructive (helpful). The key ethical testing practices that assessment practitioners should follow have been highlighted in the International Guidelines for Test Use developed by the International Test Commission (ITC; 2001). The Guidelines can be found on the ITC's website (www.intestcom.org). According to the ITC Guidelines, ethical testing and assessment practices require that the assessment practitioner ". . . will use tests appropriately, professionally, and in an ethical manner, paying due regard for the needs and rights of those involved in the testing process, and the broader context in which the testing takes place" (2001, p. 7).
The ITC Guidelines indicate further that the goal of ethical testing practices will be attained by practitioners who have the necessary competencies, spanning the entire process of testing; a sound knowledge of psychometrics and testing; and an understanding of the broader social, cultural, political, and legal context in which testing is used and the manner in which such factors might affect test results, their interpretation, and the use to which they are put. It is particularly this latter aspect that assessment practitioners in Africa need to pay special attention.
The core ethical consideration facing those who conduct psychological testing in the culturally and linguistically diverse African continent relates to how best to cater for this diversity so as to be sensitive to test-takers' cultural backgrounds and values during the test selection, administration, interpretation and reporting phases of the testing process. In this regard, Korman (1973) asserted that:
The provision of professional services to persons of culturally diverse backgrounds by persons not competent in understanding and providing professional services to such groups shall be considered unethical; ... it shall be equally unethical to deny such persons professional services because the present staff is inadequately prepared; ... it shall be the obligation of all service agencies to employ competent persons or to provide continuing education for the present staff to meet the service needs of the culturally diverse population it serves. (p. 18)
Consequently, a challenge facing those involved in psychological training and continuing professional education in Africa is to seek ways of developing assessment practitioners and researchers with an increasingly multicultural awareness and worldview. Furthermore, more practitioners from indigenous communities, who have often not had access to psychological services, need to be trained. In this way, the provision of ethical psychological testing services will be facilitated and enhanced.
Against the backdrop of the core ethical consideration of how to cater for multicultural diversity in psychological testing in Africa, the remainder of this paper will expand on some of the specific ethical issues related to the various phases of the testing process.
Ethical Issues to Consider Before Using Tests
Immersing Yourself in the Test-taker's World
Extracts from the following poem by Kizito Muchemwa, an African poet, serve to provide much food for thought for psychological assessment practitioners:
They came into the wilderness clich‚s
Talismans they cherished as shields against poisonous madness ...
They saw no familiar hills and heard no familiar songs.
Holding onto their fetishes they defy time and distance ...
They surround themselves with jacarandas and pines
Build concrete walls around their homes
I hope next time they will import snow, change
The seasons to homour their eccentric whims. (in Chapan & Voss, 1986, p. 237)
Just as the tourists in the poem who defiantly changed the unfamiliar surroundings to become something more familiar to them, but very unfamiliar to the indigenous people, assessment practitioners need to guard against doing the same. Imagine how people living in a deep (remote) rural area might react with anger and suspicion if an unknown person arrives in their village one day and orders them to do certain strange things. Would the outcome of such testing be valid, ethical and constructive?
One thing that I learned early on in my career was that to perform an assessment in a valid, ethical and fair way in a multicultural context, you have to acquire knowledge of the test-taker in relation to his or her cultural, family, linguistic, educational and socio-economic background and heritage. Furthermore, such knowledge should be acquired before you decide whether psychological testing is possible and what type of test tasks might be appropriate. You should never presume that you know how best to assess some aspect of human or cognitive functioning, or how to interpret test performance without first having immersed yourself in the lived world of the test-taker. This implies adopting an emic approach in which human behavior is examined using criteria related to a specific culture as opposed to using behavioral criteria that are presumed to be universal (etic approach).
How can assessment practitioners immerse themselves in the world of their test-taker? One way is to use community and family genograms to gain insight into the cultural factors that underlie individual and family development (Ivey, Ivey, & Simek-Morgan, 1997). In the community genogram, test-takers are asked to develop a visual representation of their community and to then generate positive stories and experiences related to different groups in the community (e.g., family, neighborhood, peers, school, religious). While genograms are normally used in counseling contexts, they can also be fruitfully employed in testing contexts to provide the assessment practitioner with a vehicle through which the community, family, and cultural background of the test-taker can be understood. Readers are referred to Ivey, Ivey, and Simek-Morgan (1997) for a more detailed explanation of community and family genograms.
Another way to immerse oneself in a certain cultural group is to venture into the real world to gain first-hand experience. Visit learning institutions, community centres, work settings, homes, and so on, where appropriate. Speak to key informants in communities. All with the aim of starting to see the community through its glasses instead of your own. You also need to be aware of your own prejudices and preconceived ideas about certain community or cultural groups and you need to take care that they do not incorrectly color the picture of the community that you are beginning to form.
One of my more humbling experiences was when my colleague and I visited schools in the rural areas not too far from the city where we lived. The purpose of the visit was to see whether the teachers wanted to be trained to use a developmental screening test so that the results could be used to help them tailor their instruction to the needs of their learners. We had the notion that the almost non-existent pre-school facilities in rural areas together with impoverished living and economic conditions would impact negatively on children's development. Our preconceived idea turned out to be very wrong. At the first school that we visited, the school principal asserted that there was little doubt in his mind that if the performances of rural and urban children were compared, rural children would be found to be superior. Although we were skeptical regarding this assertion, in the years to come the data that we collected from the developmental screening test clearly showed that the rural children performed significantly better than urban children on the test! Despite poor preschool facilities and impoverished living conditions, rural schools provided a more stable learning environment than urban schools at a time in our country's history when the disruption of education, particularly in urban areas, was seen as one way of protesting against the discriminatory practices of the Apartheid regime in an attempt to bring about political change. This explanation would have escaped us had we not dialogued with key community informants about the characteristics of urban and rural school environments and the factors impinging upon the provision of education.
Whether you use genograms, interviews,
or community visits to immerse yourself in the lived world of your test-takers,
language could prove to be a barrier, as indeed it could be at any stage during
the testing process. Given the diversity of languages and dialects spoken by the
peoples of Africa, the chances are high that the test-takers might not speak the
same language as the assessment practitioner. Consequently, interpreters may
have to be used. Where interpreters are used, they should be trained in the
nature of interviewing and psychological measurement, and should have a clear
understanding of their role. If such training is not undertaken, interpreters
could invalidate the information obtained by, for example, adding their own
opinions, prompting test-takers regarding an appropriate answer, or even asking
questions in a leading way. Nonetheless, even with training, interpreters are
likely to be handicapped by the fact that each of the languages being used are
likely to have concepts for which there are not equivalent concepts in the
other, making accurate translations difficult.
If you work in a specific community for a while, you should try to learn certain important phrases, greetings, and greeting rituals (e.g., handshakes). This will demonstrate to the community that you are serious about building a relationship with them and your efforts will be appreciated. In turn, this will assist you to establish a trust relationship with them and will dispel any suspicious thoughts that they might hold regarding your intentions.
Considering When it is Appropriate to Use a Test and When Not
Testing is a Westernized activity that cannot always be transported to non-Westernized cultures. Psychological testing was brought to Africa in the colonial era, and is not something that is indigenous to Africa and its peoples. Given differing levels of literacy and education, 'taking a test' is not something that is necessarily within the everyday experience world of many people in Africa. According to Nell (1997), the extent to which a test-taker is testwise has a significant impact on test performance. Consequently, if an assessment practitioner wants to follow ethical testing practices, the extent to which it is possible to even consider administering a test is a decision that needs to be reached early on. In some instances, using other forms of assessment (e.g., behavioral observation) and obtaining information from key family members and community informants (e.g., teachers, religious ministers/priests, a respected elder person in the community), may be preferable to administering a test to someone who has no test-taking experience. Not only will this eliminate the anxiety that taking a test could have for such a person, but it will also provide the assessment practitioner with more valid and authentic information about the person.
Alternatively, an assessment practitioner could try to lessen the impact of a lack of testwiseness on the part of test-takers by ensuring that they are thoroughly prepared prior to testing. By using practice examples completed under supervision, or even complete practice tests, test-takers can be prepared for both the nature of the test tasks as well as for the demands of the assessment context in general (Foxcroft, Roodt, & Abrahams, 2001). Nonetheless, there are instances where supervised practice examples are not sufficient to reduce the negative impact of a lack of testwiseness on test performance. This is especially true when young children are tested and when a computer-based test (CBT) is used. For example, my experience has been that when you test Grade 1 children who have had limited or no exposure to pre-school education, practice examples do not sufficiently prepare the children for the test, especially if you test them early in the first term. One of the reasons for this is that having to pay attention to fairly complex instructions and respond to them, using a pencil or a crayon to draw or write, and even assembling something unfamiliar, are such foreign experiences that the test results do not accurately reflect their ability or potential. In such instances, testing should be conducted later in the school year, and consideration should be given to administering a practice test before the 'real' test.
When it comes to CBT, which is increasingly being used in Africa at present, research suggests that an introductory tutorial to familiarize test-with the computer, together with practice examples, have been found to adequately prepare test-takers with low levels of computer familiarity. However, many people in Africa generally lack technological sophistication, which means that a brief tutorial and practice examples may not sufficiently prepare them to perform according to their ability on a CBT (Foxcroft, Watson, Greyling, & Streicher, 2001). When you want to use a CBT, you should first determine the test-taker's levels of technological sophistication and computer familiarity. Test-takers with low levels of computer familiarity should rather be given the option of taking a paper-based test in the interests of following ethical testing practices.
Following the Correct Protocol to Gain Permission and Consent to Test
When psychological testing is to be conducted as part of a research project, permission often needs to be gained from appropriate government and community bodies. This process can take a long time, even when assessment practitioners already have a working relationship with such bodies. Relevant stakeholders will want to scrutinize the test material to see that there is nothing that could be considered to be offensive and insensitive. Furthermore, assessment practitioners will need to motivate why the testing needs to be undertaken and how the community will benefit from the project. The test and research results need to have practical value and application for the community group if they are to perceive it to be constructive, and thus ethical (Durojaiye, 1979). For this to happen, assessment practitioners need to negotiate with relevant stakeholders what this will entail (e.g., teachers will be provided with the results of learners in their class and a workshop will be held to empower teachers to tailor their instruction according to the results). In addition, a timeframe for providing the test and/or research results needs to be agreed upon. Usually, for the results to be of practical value, they need to be provided as soon as possible after the testing has been completed, which might require that the final research report should be compiled at a faster rate than is normally the case.
Whether the testing forms part of a research project or as part of psychological service provision, obtaining informed consent is central to conducting the testing in an ethical way. Nonetheless, obtaining informed consent when children are to be tested, or those who have no experience of testing, poses many challenges. Parents might seek work in the large cities, while leaving their children in the care of their grandparents or another member of the extended family in a rural village or in another urban area. Normally, it is the ethical responsibility of the assessment practitioner to obtain consent from the child's parents or legal guardians. However, in this instance, would it be appropriate to gain permission from the child's grandparents or extended family member? Should the parents be contacted telephonically or in writing? There are no easy answers to these questions and the situation is compounded further if the parents, grandparents or family members have no experience of testing and are thus unsure as to what they are consenting to. The current main caregiver is probably the most appropriate person to contact first. The caregiver may have no problem giving consent. If not, then the caregiver could direct you to the most appropriate person to contact.
Trying to explain testing in as concrete and colloquial terms as possible is probably the only way of trying to ensure that people with no experience of psychological testing are sufficiently informed to consent to the testing. The more that you have immersed yourself in the culture, the better you will be able to explain the concept of testing in a way that community members can relate to.
Where to Test
Given that psychological testing is unfamiliar to many people in Africa, assessment practitioners should also think carefully about where the best place is to conduct the testing so as not to place a further strain on test-takers. For people from a deep rural area, transporting them to a big city and taking them to a strange office or test room may make the test session very anxiety provoking, which could invalidate the results and leave them with a negative perception of psychological testing. In the interests of ethical testing practices, it would be better to test such a person in familiar surroundings in their community or village. However, this too will pose certain challenges to the assessment practitioner. Finding a quiet place without distractions where the assessment can take place is not easy and is sometimes impossible.
When to Test
Usually, when a client comes to a psychologist for help, appointment dates are arranged that suit the psychologist. However, when professional services are delivered in multicultural contexts, sensitivity needs to be shown toward the client by, for example, making allowances for traditions (e.g., dates of initiation camps and ceremonies) and special religious observances when setting dates for psychological testing. Furthermore, setting a date to conduct testing in a rural, farm school during the fruit-picking season, for example, is futile. Most people in the area, young and old, will be employed to pick fruit, and school attendance dwindles noticeably during this time.
Choosing Appropriate Tests
Psychological tests are closely linked to the context in which they are designed. A test is usually designed in a certain context (society, culture) for a specific purpose, and the way in which test performance is interpreted is linked to behavioral criteria, norms, or cut-scores developed in the context where the test was developed. Tests developed for use in multicultural contexts would have to include items and tap constructs common to all the cultures in which it is to be used. However, very few multicultural tests have been developed in the African continent. The majority of tests that are in use have been developed in a mono-cultural context, either in Africa but more likely in the United States (US), United Kingdom (UK), or Europe. These tests cannot simply be applied to people form other cultural backgrounds without investigating the possibility of test bias and without strong consideration being given to adapting and re-norming them. Herein lies one of the biggest challenges facing ethical testing practices in Africa. To be honest, very few bias studies have been undertaken and little effort has been put into adapting mono-cultural, Westernized tests for the many cultural groups in Africa. It is thus essential that we start raising consciousness among assessment practitioners regarding how unacceptable (unethical) it is to use tests before having undertaken bias studies, to use tests without adapting and re-norming them, and to use them without establishing 'local' psychometric properties. Below are a few suggestions of aspects that should be addressed when deciding on the appropriateness of a test and its possible adaptation for a certain group.
Construct equivalence. An important question that needs to be asked is whether a test developed for a certain culture measures the same construct in another culture? One way to investigate this is to compare the factor structures of the test in the two cultural groups. In addition, it may be useful to use focus groups where experts can unpack the meaning and description of a particular construct in their culture. If there is no evidence of factor structure equivalence and/or there is qualitative evidence from the focus groups that the way in which the construct is defined or described in the two cultural groups differs in some ways, consideration should be given to adapting the test before it is used in another cultural group. The ITC test adaptation guidelines developed by Hambleton (1994), in collaboration with international experts, should be consulted and followed when adapting a test developed in one culture for another cultural group.
Nature of test content or tasks.
Other than measuring the appropriate construct, the test content or tasks need
to be culture-appropriate. For example, if playing with colored wooden/plastic
blocks is not a familiar activity for children in a particular cultural group,
then a classic block design task, found in many individual intelligence tests,
would be inappropriate. Similarly, trying to administer a paper-based verbal
(written) test would be inappropriate for members of a deep rural community who
have only been exposed to oral and not written language.
Creativity, ingenuity and knowledge of the behavioral criteria associated with a construct in a specific culture are required when adapting test content or test tasks. Here are a few creative adaptations that I have come across. Given the richness of the designs found in African beadwork, the reproduction of a design using beads could be used instead of a block design task or less complex bead patterns could be used in seriation and sequential memory tasks. Arithmetical problems can be represented in terms of counting heads of cattle, oranges, beads, or stones, or in terms of quantities related to cooking for a certain number of people. A child who has never been exposed to using a pencil or a crayon could be asked to draw a figure with a stick in wet sand or to construct a human figure from clay instead of drawing a man or a woman. When adapting tasks that involve the use of color, special care needs to be taken. In some cultures (e.g., the Xhosa culture), the color vocabulary is limited and the name of a common object whose color is a prominent attribute is sometimes used when referring to a color name. For example, in isiXhosa the word for "blue" and "green" is the same, so to more accurately describe the color of something that is blue or green an object is added to the word such as "like the sky" or "like the grass". In this instance, using both colors in a test item could be problematic for the test-taker and one of the colors should rather be eliminated.
The process of adapting a test task should not be undertaken independently by the assessment practitioner. Local cultural experts, linguists, and anthropologists, among others, should be consulted as they could provide valuable information that could guide the adaptation process. In addition, the adapted content, instructions, and tasks should be carefully piloted and, if needs be, refined before the adapted test is used.
Furthermore, the psychometric properties of the adapted test have to be established before it can be used with confidence. It should not be assumed that since the original test's psychometric properties were sound that it is not necessary to investigate the psychometric properties of the adapted test. In addition to establishing the validity and reliability of the adapted test, the psychometric equivalence of the original and adapted tests needs to be established if it is to be asserted that the two tests measure the same constructs across the two cultures. This can be achieved by investigating (a) item equivalence through the application of various techniques to determine differential item functioning (DIF) (Hambleton, 1994) and (b) factor structure equivalence using confirmatory factor analysis methods, for example.
You might be tempted to ask whether it
is only for the purposes of pursuing cross-cultural research studies that it is
important to establish the psychometric equivalence of the original and adapted
tests? The answer is "No". In the multicultural societies that exist
in virtually every African country, people from various cultural backgrounds
have to compete for the same jobs, education and training opportunities, and so
on. If psychological tests are used to guide personnel selection and educational
training decisions, it is vital that the tests used are culturally and
linguistically appropriate and that, should different versions of the test exist
for different cultural groups, the different versions have been empirically
demonstrated to be equivalent. Otherwise questions will be raised regarding bias
and fairness, the similarity of selection standards being applied, and so on,
which will in turn raise serious issues around the ethical and fair use of
Language issues and test translation. Issues related to language in multicultural testing are as complex as issues around the cultural appropriateness of a construct and the tasks used to tap it. From an ethical perspective, test-takers have the right to be assessed in the language of their choice, which is normally their home language. To be able to allow test-takers to exercise this right, however, will present the assessment practitioner with a number of challenges. While I will highlight some challenges related to translating test instructions and verbal items, these challenges apply equally well to instances where a test is being developed in an indigenous African language.
When attempting to translate test instructions and verbal test items into the indigenous language of choice of test-takers, one common difficulty encountered is to find bilingual experts to undertake the translation and to judge its accuracy. The translator needs to understand the idiosyncrasies and subtle nuances of both languages as well as the social meanings attached to words and phrases. Furthermore, where an item contains proverbs or folklore relevant to the culture for which the test was originally designed, the translator cannot simply perform a word-for-word translation. The item will first have to be adapted by, for example, finding a proverb with a similar meaning in the culture for which the test is being adapted and translated.
Another common difficulty that arises during test translation is that equivalent indigenous terms or words cannot always be found in the language into which the instructions and verbal items are being translated. To get around this problem, a fairly long description of something for which no term or word exists has to be provided. For example, when one of my students was researching the gender stereotyping of careers in young children, a difficulty arose when trying to translate "Can a boy/girl become a veterinarian?" into isiXhosa as there is no specific word for "veterinarian" in this language. A description of what a veterinarian does instead of a single word is usually used by Xhosa-speaking children. Thus, the community experts and linguists advised us to use the descriptive phrase "a person who makes animals better". It is not only the difficulty of finding appropriate words and terms in the indigenous language that needs to be overcome during translation but translators also need to be sensitive to local dialects and nuances in the use of language. Thorough pre-testing of the instructions and test items is thus essential to identify what words and phrases are unfamiliar to a sample of the target group of test-takers so that the language used can be refined before the actual administration of the test.
When the equivalence of the source and target versions of the test has to be evaluated, forward- and back-translation methods can be used for this purpose. Furthermore, the two versions of the test can also be administered to bilingual test-takers, or source and target language monolinguals. However, researchers might struggle to find bilingual test-takers, especially in rural villages. Source and target language monolinguals are easier to obtain, but the matching of the two groups of monolinguals could prove to be challenging, given the possible diverse backgrounds of the groups.
At the start of this section, I made the point that test-takers have the right to be tested in the language of their choice. For many test-takers this would mean being tested in their home language. However, especially in the big cities and in more Westernized areas, children and adults are often educated in English, which for many represents their second or third language. Especially where tests tap previously learned knowledge (e.g., certain subtests on intelligence tests), it may be more ethical and valid to test such children and adults in the language medium in which they have received their instruction, or even to use bilingual assessment. The point to stress is that in the interests of ethical testing practices, the assessment practitioner should gain information about both the test-taker's home language and the language of instruction used during schooling, before reaching a decision on which language would be appropriate to use during testing or whether bilingual testing would be the most appropriate.
Self-report questionnaires and self-reflection tasks. Westernized and non-Westernized people differ in terms of the relative importance that is given to individual needs as opposed to group needs. Westernized society emphasizes individualism whereas in non-Westernized societies collectivism (a sense of "ubuntu") is stressed more. The implication of this is that the Westernized child learns to introspect (reflect), to be aware of personal needs, and to strive for personal growth and development. In contrast, the non-Westernized children learn to place their personal desires below that of the needs of their community, and awareness of and participation in the welfare of their society is seen as being more important than pursuing personal happiness. A self-report questionnaire or an introspection task will thus pose different challenges for individuals from individually based and collectively based societies, and the latter individuals will be at a distinct disadvantage when confronted by such questionnaires and tasks. In this regard, Tollman and Msengana (1990) comment that:
The evaluation of responses based on introspection is an integral requirement when behavioural assessment is made, yet the ability to focus on the 'self', to explore inner feelings and needs, would be highly developed in a society emphasizing 'Personal Happiness', but not in a society concerned with 'Group Happiness'. (p.24)
The possibility that self-report questionnaires and tasks that require introspection or self-reflection on the part of test-takers are biased against people from non-Westernized societies thus needs to be kept in mind by assessment practitioners and researchers when they select psychological tests to assess a certain aspect of behavioral functioning.
Speed versus power tests. Assessment practitioners and researchers need to carefully consider whether to use a speed test (where time limits are imposed and items are all of similar difficulty) or a power test (where no time limits are imposed, but the items get progressively more difficult). With test-takers who are not testwise, using a power test might be less anxiety provoking, as the fact that they are not pressurized for time means that they have one less thing to worry about. Furthermore, "being ruled by the clock" and being very precise about time is something that is found in Westernized cultural groups. Non-Westernized people may perceive time differently and may be less driven by time. Consequently, being asked to perform something as quickly as possible within a given time limit may be unfamiliar to them and they are also less likely to be intrinsically interested in and motivated by such tasks. The outcome of this will be that an unreliable picture of their functioning will be obtained as regards the aspect of behavior that was measured.
Familiarity with the laws governing testing. In certain African countries, the use of psychological tests is under statutory control. In such an instance, it would be unethical for someone who is not a psychologist, for example, to use a psychological test. Assessment practitioners and researchers who are uncertain whether there are laws governing who may and may not use psychological tests in a certain country, should contact national professional psychology associations and statutory bodies to obtain this information before they embark on testing.
Ethical Issues to Consider During Test Administration
Be Aware of the Impact of the Administrator on Test Performance and Try to Minimize It
In multicultural contexts, if the cultural background and the language spoken by the assessment practitioner are different to that of the test-taker, they could have a negative impact on test performance (Hambleton, 1994). For example, Kamara (1971) found that the conservation skills in children from Sierra Leone lagged behind their developmental stage when the administrator did not speak the children's home language, but their performance was age appropriate when tested by administrators who spoke their home language.
Asking for Help During Group Testing
It is standard procedure during group and individual testing to tell test-takers that they can ask for help if they need to. Especially in multicultural contexts, where the different cultural groups do not have an understanding of each other, and where the test administrator and test-takers are from different cultural backgrounds, it is unlikely that test-takers will ask for help. Why? The test-takers might feel that they are displaying their ignorance by seeking for help, which is something that they do not feel comfortable doing in front of someone from another cultural group. Consequently, they do not ask for help, which could impact negatively on their test performance.
Clarity of Instructions and Use of Practice Examples
Given the language barriers described earlier that test-takers are likely to encounter, and the problem of a lack of test-wiseness in general, it is vital to ensure that test-takers clearly understand what is expected of them. This might mean repeating the test instructions or allowing test-takers to ask questions of clarification. The use of supervised practice examples aid in trying to ensure that test-takers have understood the nature of what is expected of them and how they should respond. Nonetheless, as pointed out earlier, practice examples should not be seen as being able to entirely counter the problem of test-takers not being test-wise.
Ethical Issues to Consider After Using Tests in an Assessment Context
Using Appropriate Norms
Age norms. In non-Western societies, keeping exact track of children's ages is not necessarily the norm, which poses difficulties in developmental assessment in particular. The assessment practitioner may thus have to estimate the child's age based on collateral information (e.g., current school standard, onset of menstruation) from a variety of sources. Often parents will tell you that their child was born at the time of the great flood, or not long after the sky went dark during the day (a total solar eclipse) and so on. By searching through historical records, you might be able to gain some estimate of the year in which the child was born. When interpreting performance on developmental tests, knowing the child's chronological age is of utmost importance. Where the age has to be estimated, the assessment practitioner has to make allowances for the fact that the estimate may be incorrect and thus what appears to be a developmental delay or superior development might in fact not be so.
Lack of appropriate local norms. Ethical test practices require the use of locally developed norms to interpret test performance accurately. However, as mentioned previously, many tests developed and normed in Europe, the UK, and the US are used by assessment practitioners in Africa. Often the norms developed elsewhere in the world are applied "with caution" and little or no attempt is made to develop local norms. This could lead to incorrect decisions being made on the basis of the results, and problems being over- or under-identified as a result of inappropriate norms being used. One of the reasons why local norms are often not developed is that there is a lack of test development and applied measurement expertise in Africa. This aspect needs to be addressed by professional societies, training institutions, and test development organizations as a matter of great urgency. Furthermore, professional societies should establish centralized databases for various popular tests, which can be populated by the data gathered by assessment practitioners on an ongoing basis, and which could provide the basis for developing national and local norms.
Being Sensitive to Background Factors
In multicultural contexts, test results must be sensitively interpreted in relation to background factors such as cultural and family background, level and quality of formal education, literacy levels, and level of test-wiseness. Failure to do this could lead to incorrect interpretations of and explanations for test performance, as illustrated in the example that I recounted earlier related to children in rural schools performing better on a developmental screening test.
Physical Problems and their Impact on Test Interpretation
In a continent ravaged by HIV/AIDS and where the majority of people do not have access to adequate medical care, clean water, and sanitation, the impact of the physical status of the test-taker on test performance during the interpretation phase should not be overlooked. For example, lack of adequate sanitation and access to clean water heightens the possibility that children might experience worm infestations, which negatively affects their attention span, speed of information processing and responding, and school performance in general. Similarly, a test-taker who is hungry is unlikely to perform according to his or her potential during testing. The more knowledge the assessment practitioner has of the environment in which the test-taker lives and works, and of the test-taker's physical status, the greater the possibility that this knowledge and information can be factored in during test interpretation.
Providing Meaningful Feedback
It is the assessment practitioner's ethical responsibility to provide feedback on the testing, be it to the individual, school, community organization, employers, and so on. However, given the low levels of literacy on the continent, oral feedback may be more meaningful than written feedback. What is of cardinal importance is that the results are conveyed so that their practical application is clear. If the results indicate that there are a group of test-takers who performed poorly on a problem-solving test, then the feedback needs to suggest how problem-solving can be developed in these individuals. In this way, the link between test results and development will be clearly established in the minds of those to whom the results are conveyed. This implies that the testing would have been perceived to be of practical value, which is one of the hallmarks of ethical testing practices.
As I look back over the things that I have learned (so far) regarding ethical issues related to testing in Africa, it interests me that much of my learning relates to the ethical issues that need to be considered, worked through, and addressed before testing can take place. These issues are far greater in number than the ethical issues that I have encountered during or after testing. The learning that I have gained from this reflection is that being well prepared and being sensitive to the test-taker's community and cultural background lies at the very heart of following ethical testing practices in multicultural contexts, in Africa and elsewhere in the world.
About the Author
Cheryl Foxcroft obtained her bachelor's, masters, and doctoral degrees in Psychology at the University of Port Elizabeth UPE), which is situated on the southernmost tip of Africa. She is a registered psychologist who has a particular interest in psychological assessment and researching measures used in multicultural contexts. After graduating, she has lectured at the University of Port Elizabeth for two decades in the areas of psychometrics, test development, psychological assessment, neuropsychological assessment, and research and data analysis methods. Currently, she heads up UPE's Admissions and Placement Assessment Programme, which is responsible for assessing the new student intake to the university so as to facilitate accountable and fair admissions, placement and developmental decisions. Nationally, Cheryl has served on the Psychometrics Committee of the Professional Board for Psychology and she currently serves on the Council of the Psychological Society of South Africa. Internationally, Cheryl currently serves on the Council of the International Test Commission and is a member of the International Association of Applied Psychology. Cheryl has presented more than 25 papers at international psychology and assessment conferences all over the world, and has presented numerous papers at national conferences, workshops, and symposia. She is the co-editor of a recent South African textbook on psychological assessment and contributed 6 chapters to the book. She is also the author of 19 articles, which have mainly focused on assessment, and that have been published in South African and international psychology and higher education journals.
Chapman, M. & Voss, T. (1986). Accents: An anthology of poetry from the English-speaking world. Craighall, South Africa: AD. Donker.
Durojaiye, M.O.A. (1979). Ethics of cross-cultural research viewed from third world perspective. International Journal of Psychology, 14, 137-141.
Foxcroft, C.D., Roodt, G., & Abrahams, F. (2001). The practice of psychological assessment: Controlling the use of measures, competing values, and ethical practice standards. In C.D. Foxcroft & G. Roodt (Eds.), An introduction to psychological assessment in the South African context. Cape Town: Oxford University Press.
Foxcroft, C.D., Watson, A.S.R., Greyling, J., & Streicher, M. (2001). CBT challenges relating to technologically unsophisticated test-takers in multicultural contexts. Paper presented at the European Congress of Psychology, London, July 2001. (Obtainable from first author at email@example.com).
Hambleton, R.K. (1994). Guidelines for adapting educational or psychological tests: A progress report. Bulletin of the International Test Commission, 10(3), 229-244.
International Test Commission (ITC). (2001). International guidelines for test use. International Journal of Testing, 1(2), 93-114. (Also obtainable from www.intestcom.org)
Ivey, A. (1995). The community genogram: A strategy to assess culture and community resources. Paper presented at the American Counseling Association Convention, Denver.
Ivey, A.E., Ivey, M.B., & Simek-Morgan, L. (1997). Counselling and psychotherapy: A multicultural perspective (4th ed.). Boston: Allyn and Bacon.
Kamara, A. (1971). Cognitive development among Themne children of Sierra Leone. Unpublished doctoral dissertation. University of Illinois at Urbana-Champaign.
Korman, M. (1973). Levels and patterns of training in psychology. Washington, DC: American Psychological Association.
Nell, V. (1997). Science and politics meet at last: The insurance industry and neuropsychological norms. South African Journal of Psychology, 27, 43-49.
Tollman, S.G. & Msengana, N.B. (1990). Neuropsychological assessment: Problems in evaluating the higher mental functioning of Zulu-speaking people using traditional Western techniques. South African Journal of Psychology, 20(1), 20-24.
Questions for Discussion
1. What is the main ethical challenge facing assessment practitioners in Africa?
2. You are asked to test the cognitive functioning of 6-year-old children in a remote rural village in Africa. What challenges will you encounter as you attempt to perform the assessment in as valid and as ethical a way as is possible?
3. Is it possible to test a non-westernized person who has no experience of testing or writing school examinations? What can you do to give such a person the opportunity to become more familiar with testing? What alternative methods could you use to evaluate the behavioral/cognitive functioning of such a person if testing is not an option?
4. How can assessment practitioners or researchers gain knowledge of the lived world and cultural background of a test-taker who is from a different community and language group than themselves?
5. Obtaining informed consent brings with it certain special challenges for assessment practitioners or researchers who work in Africa. What are these challenges?
6. You are required to convince a community that has no experience of psychological testing that such testing can be of value (benefit) to them. Briefly indicate the main points that you would include in your argument.
7. "Translating test instructions and verbal items, and adapting test items and tasks is easy to achieve in Africa". Support or refute this statement.
8. What steps will you follow if you have to adapt a test task (verbal or non-verbal) for a specific community?
9. What are some of the factors that impact on test interpretation, especially in Africa?
10. What should be done in the training of assessment practitioners and researchers to enable them to develop the necessary knowledge and sensitivity to work in an ethical way in multicultural and multilingual contexts?