Organisations worldwide are currently facing increasing competition, price pressures and slowergrowth rates and in order to be successful, for many organisations, this prompts the need to constantly change to survive. Having the right employee fit-for-purpose is imperative to the success of an organisation faced with such challenges. The use of psychological assessment is a key enabler for the appropriate selection and development of employees in the workplace to meet the constant changes in the internal and external economic and social environment (Bartram, 2004; Paterson &Uys, 2005).
However, the nature and value of current day psychological assessment needs to be assessed as it faces many challenges, particularly in a South African context that is influenced by political injustices of the past. Psychological test use in South Africa currently faces many challenges including the adaptation of tests in a multicultural context, language, age and gender barriers and other measurement challenges (Foxcroft, 2004; Foxcroft, 2006).
These challenges introduce criticism of the fairness and ethical practices of psychological assessment and highlight the need for enhancing fairness in assessment and developing equitable and unbiased psychological tests (Foxcroft, 2011; Paterson & Uys, 2005).
In order to respond to these challenges, it is important to gain an understanding of how psychological assessments have developed over the years so that meaningful changes can be made ((Foxcroft, Roodt & Abrahams, 2005).
It is important to reflect on the history and development of psychological assessment as it helps explain how and why it is currently practiced, it allows for critical analysis of psychological assessment, identifying the strengths and weaknesses of tests used today and provides lessons to be learned from the weaknesses and injustices of past assessment practices so that psychologists can develop new assessment tools and techniques and ensure fair and ethical social practices of psychological assessment (Gregory, 2000; Moerdyk, 2009).
The perceptions of psychological assessment being unfair and biased are slowly changing as improvements are made in the development of tests and in South Africa this is closely linked to the developments and changes in legislation and the professional bodies that govern the control and use of psychological assessments (Mauer, 2000; Paterson & Uys, 2005).
The objective of this discussion is to evaluate the nature and value of psychological assessment practices in the South African work context by critically reflecting on the historical developments that have shaped and influenced current psychological assessment, focusing onhow test use have been transformed by legislation, government and society and the challenges that exist for fair and unbiased psychological assessment in South Africa. Finally, the ways in which fairness and bias in testing is currently being addressed and the steps in developing further ethical practices of psychological assessment will be discussed.
This discussion will also include an overview of psychological assessment STUDENT NO: 44674481 4 and clarification of central concepts such as psychometrics, testing, measurement and evaluation, fairness and bias.
Foxcroft & Roodt (2005, p. 4), define psychological assessment as a “… process-oriented activity aimed at gathering a wide array of information by using assessment measures (tests) and information from many other sources.. ” in order to make inferences about human behaviour.
Shepard & Spalding (as cited in Setshedi, 2008) note that psychological assessments make use of standardised instruments, which, combined with other sources of information, are evaluated to inform some intervention.
Psychological assessment refers to the overall multidimensional process that uses tools (tests) and techniques for measurement and evaluation of human behaviour (Foxcroft & Roodt, 2005). Psychological assessment can be distinguished from some related terms commonly used, such as, psychometrics, testing, psychometric testing, measurement and evaluation.
Measurement refers to the ‘how much’ in an assessment. When doing an assessment, psychologists use measurement to ascribe values or numbers to a phenomenon against certain criteria or standardized norms, following predefined rules (Moerdyk, 2009). Measurement therefore helps in gathering information to add to the holistic assessment of a phenomenon. 1. 2. 2. Evaluation and Assessment Evaluation refers to “…interpreting or attaching a judgemental value to an assessment…” (Moerdyk, 2009, p. 4).
In other words, evaluation refers to reaching conclusions about the measurement outcomes (Moerdyk, 2009). Thus evaluation forms part of the process of the overall assessment. STUDENT NO: 44674481 5
Psychometrics is the subfield of psychology that refers to the use of scientific methods, theory and techniques used in psychological measurement (Foxcroft & Roodt, 2005). Psychologists base their assessments on these scientific ways of measurement in an attempt to learn more about or predict behaviour. Thus psychological assessment is informed by the study of Psychometrics (Foxcroft & Roodt, 2005).
Testing refers to the use of a tool (for example, psychometric tests, scales, or instruments) to inform decision making (Foxcroft & Roodt, 2005; Moerdyk, 2009; Patterson & Uys, 2005). Tests are not used in isolation but as part of the bigger process of assessment in order to make fair and informed decisions. Patterson & Uys (2005) argue that assessment is the overall process which includes testing, but more than testing, it implies the holistic evaluation based on test results and other information from other assessment methods.
Psychometric testing refers to testing based on psychometric theory and practices and is mainly used in industries for recruitment, selection and counselling purposes (Gregory, 2000). According to van der Merwe (2002), psychometric testing can be a useful tool in making decisions about management of employees and if used appropriately with insight and sensitivity it can become one of the most effective ways in predicting behavior that is not surfaced during screening interviews.
In South Africa psychometric testing is regulated by the Employee Equity Act of 1998 which prohibits the use of psychometric tests or assessments unless it proves to be scientifically valid and reliable, is fair and free from bias (Foxcroft et al, 2005).
Psychologists use techniques and tools in assessment in attempts to gain a better understanding of human behaviour, to describe certain phenomenon, as well as topredict future behaviour. Psychological assessments are also used to identify and plan interventions to change behaviour and in decision making (Foxcroft & Roodt, 2005; Moerdyk, 2009). Some of the main areas of assessment are cognitive, behavioural and personality assessment (Moerdyk, 2009).
Organisations benefit from psychological assessments as it aids in making clear and informed decisions in managing STUDENT NO: 44674481 6 human resources. Bartram (2004, p. 238) highlights the value of assessments in personnel selection enabling “…organisations to act both tactically and strategically to increase their effectiveness”. Psychometric tests make it possible for organisations to measure attributes that can’t easily be assessed at face value but that may be crucial in job performance, such as personality and leadership assessments (Moerdyk, 2009).
Assessments are also used to measure work performance to identify training needs (Moerdyk, 2009), management and leadership assessments to inform development plans (Bartram, 2004), as well as assessments to improve employee wellness (Moerdyk, 2009). Assessments can also inform career counselling needs, career and organisational development plans, and research (Moerdyk, 2009).
From the above it is clear that there is a need for psychological assessments in the workplace for organisations and individuals to benefit from human resource management.
However psychological assessment practices over the years have drawn attention to bias in testing and fairness and ethical issues in assessment. The developments of psychological testing and assessment practices in South Africa will be further discussed to show how historically assessments have been unfair and bias and how this has been addressed in current practices. First it is important to clarify what is meant by fairness and bias in testing and assessment.
Bias can be defined as the “systemic error in measurement or research that affects one group (e. g. race, age, and gender) more than another” (Moerdyk, p. 261). In the work context this means that the psychometric tests that are administered to people need to measure what it’s supposed to measure and it needs to be consistent and systematic in measurement across different groups. Bias in testing can be controlled through the use of statistical procedures or other objective measures (Foxcroft & Roodt, 2005; Moerdyk, 2009). Fairness refers to the equitable nature of the psychological assessment processes and procedures and the lack of bias in testing methods and interpretation.
Fairness is both subjective and contextual as can be seen in South Africa where the use of affirmative action is used in human resource practices to address discrimination in past assessment practices. To be fair to previously disadvantaged groups it has become necessary to select and develop employees that score lower on tests. This raises the debate of whether the previously advantaged groups are now being unfairly treated.
There is the option of treating both groups fairly, however it does not resolve the unfair, bias and discriminatory practices of the past (Moerdyk, 2009). STUDENT NO: 44674481 7 Both fairness and bias are important in psychological assessment in South Africa and are governed and regulated at the individual (practitioner) level, through professional bodies as well as through legislation. Fairness and bias in assessment has been a challenge in South Africa and has been shaped and influenced particularly by the political ideologies through time.
These challenges in fair and unbiased testing and the development of psychological test methods inSouth Africa will now be further explored.
The early developments of psychological assessments in South Africa were introduced in the twentieth century and followed closely with the developments internationally (Painter & Terre Blanche, 2004).
The psychological tests that were introduced at the time in the country were largely influenced by the political context (Foxcroft et al, 2005). South Africa was a British colony and thus the psychological tests used were influenced by the political ideologies of the British rule and focused on assisting in the regulation of differential treatment (Louw, 2002). Most of the tests that were administered were for whites only and were based on either adaptations of international tests (e. g. Stanford-Binet adaptation) or were developed specifically for use in South Africa (e. g. South African Group Test).
Other examples include the study by Fick, who measured intellectual abilities across different race groups using tests that was standardized for whites and concluded that whites had superior intellect. This study influenced the Bantu Education system but the limitations of the study were also highly criticised (Foxcroft et al, 2005). Thus in this era of British colonial rule, tests in South Africa were biased, discriminatory against other races and psychological assessments ignored the influence of other social factors on test performance such as poverty, language, culture, etc. (Claassen, 1997; Foxcroft et al, 2005; Louw, 2002)
In the early 1900s the political rule in South Africa was based on laws of segregation (apartheid) which believed that blacks and whites were different and should be treated differently (Claassen, 1997; Louw & van Hoorn,1997). Thus, most of the tests that were used supported the regulation of legislature that encouraged such thinking (Foxcroft et al, 2005; Louw & van Hoorn, 1997). Between the two world wars, research started to focus on understanding the “Native” and
psychology aimed to support the government and apartheid rule (Louw & van Hoorn, 1997). After the Second World War, assessments focused on suitable placements of blacks in the workplace, and tests such as the General Adaptability STUDENT NO: 44674481 8 Battery (GAB) were used. In this era, tests were used without adapting to South African norms, on whites, such as the Otis Mental Ability Test (Foxcroft et al, 2005). Between the 1960’s and 1990’s institutes were formed (National Institute for Personnel Research, Institute for Psychological and Edumetric Research) as well as changes in legislation were made to regulate the administration of psychological assessments.
The Health Professions Act 56 of 1974 regulated that only psychologists were allowed to administer psychological assessments. Radical changes in the socio-political situation started in the 1980’s and apartheid rules started to soften. Different races started to compete for positions in the workplace and this raised concerns about the test measures that were being used. Tests such as the General Scholastic Aptitude Test (GSAT), the Ability Processing of Information and Learning Battery (APIL-B) and the Pencil and Paper Games were introduced to address these concerns.
The Pencil and PaperGames was made available in all 11 languages in South Africa to address bias and application in a multicultural context. These tests however we criticised as it was based on inappropriate norms. As the apartheid regime came to an end, psychological assessment was held under much scepticism and negative perceptions were developed because of the discriminatory nature and use of tests at that time (Foxcroft et al, 2005; Van de Vijver & Rothman, 2004).
After the abolition of the apartheid era, South Africa became a country based on democracy and radical changes were made in political policy and legislation to address the wrongs of the past (Foxcroft et al, 2005).
Psychological assessment was held under much criticism and the changes in constitution and legislation lead to a major shift in the approach of testing in South Africa. Some of the issues that were highlighted in the practices of testing during the apartheid era were the discriminatory use of tests, the lack of applicability in a multicultural context creating bias, as well as other ethical issues such as the use of inappropriate norms and standards (Foxcroft et al, 2005).
The changes in politics will now be further discussed to show how it addressed the above issues and regulates fair, ethical and unbiased assessments in current society.
Chapter 2 of the South African Constitution addresses the Bill of Rights, which calls for democracy based on dignity, equality and freedom. Furthermore, Section 9 of the Bill of Rights states that no persons may be subject to unfair STUDENT NO: 44674481 9 discrimination or unfairly discriminate unto others (Mauer, 2000).
This means that no tests may be used to discriminate against any individual, group or institution and called for fair assessment approaches. Psychological assessments therefore are required to aim to eliminate any bias in testing.
Thus a dynamic and multicultural assessment approach is necessary which encourages the development of culture-specific and culture-informed psychological practices for the various cultural groups in South Africa (Van de Vijver & Rothman, 2004).
The Labour Relations Act (66 of 1995) (LRA) protects individuals against unfair labour practices and discrimination in the workplace (Mauer, 2000).
This has an impact on the ethical considerations of organisations when employing psychological assessments. Joiner (2000) provides guidelines for fair and ethical practices in psychological assessment which ensures that psychologists take responsibility in assessments protecting the rights of individuals, treating individuals fairly, and not misusing the information from test results. Both the constitution and LRA address unfair discrimination and allow for fair discrimination if it can be proven that the discrimination is fair. 2. 3. 3. The Employment Equity Act (55 of 1998)
The Employment Equity Act (55 of 1998) (EEA) enforces fair, unbiased and equitable practices in psychological testing by regulating the requirements of tests. The Act clearly states that psychological testing and other forms of assessment is prohibited unless it can be scientifically proven to be valid and reliable, can be applied fairly and is unbiased to any employee or group. The impacts of this act means that psychological tests need to be cross-culturally applicable (Foxcroft et al, 2005).
Foxcroft (2004) highlights the need for more tests to be developed in South Africa that can be used in a multicultural context. Organisations need to take cognisance of the above legislation that regulates fair and unbiased practices of assessment and should aim to practice testing ethically by ensuring that the rights of individuals are protected, the processes of assessment is fair, confidentiality of individuals is protected, individuals are given feedback to assessments, and the outcomes of assessment is not harmful to the individuals (Moerdyk, 2009). In line with the LRA the EEA also allows for fair discrimination in employment, specifically affirmative action if used in accordance with the Act.
It also allows for fair discrimination in employment based on criteria that are inherent to the job which means that the assessment tools used to assess individuals need to assess evidence of these criteria. 2. 3. 4. Professional Bodies Other bodies such as the Health Professions Council of South Africa (HPCSA), the Professional Board of Psychology STUDENT NO: 44674481 10 and the Board for Personnel Practitioners govern the practice of psychological assessment in South Africa to ensure fair and ethical approaches in psychology (Moerdyk, 2009; Wassenaar, 1998).
Psychological tests can only be administered by individuals professionally registered with the HPCSA, which provides professional guidance and advice for the use of psychological assessments (form 208). The HPCSA regulates the training of professionals, the development of psychological tests and the use of psychological tests. 2. 4. ENHANCING THE DEVELOPMENT OF FAIR AND UNBIASED ASSESSMENT MEASURES 2. 4. 1. Steps in Test Development To address some of the issues in fairness and bias in testing, Foxcroft (2004) has identified some critical steps in developing tests, particularly for a multicultural context.
These include: (a) Defining the purpose of the test and the intended target population This involves outlining what is being tested, how the outcomes of the tests will be used and specifying that the test can be used in a multicultural setting. This implies that the test developer needs to identify aspects of the constructs that are common and unique across the groups being tested. The test developer should also identify the characteristics of the target population (such as educational status) and also identify the characteristics that may impact on test performance (such as language) (Foxcroft, 2004).
One of the major challenges in assessment in South Africa at the moment is language. In South Africa there are eleven official languages, and although English is the commonly accepted business language, psychologists need to ensure that in administering tests language is not a barrier in test performance. Translation of tests into different language is complex and makes the validity of constructs vulnerable (Foxcroft, 2006). (b) Defining the construct and creating a set of test specifications to guide item writing This step refers to identifying the specific observable measures that will be tested.
This can be derived from a job analysis which highlights the key knowledge, skills, abilities and other attributes required to perform a job successfully. It is important in a multicultural test that the norms and values across the different cultural groups for a specific construct are identified upfront so as to eliminate the construct bias. The constructs being measured need to also be of value and have meaning to the test user. Language is often a barrier in the meaning of a construct across multicultural groups and research also indicates that in South Africa some groups attach political meanings to some constructs which can impact on test performance.
Developing the content and specifications of the test is dependent on the methods used to develop the test, which can be theory-based (this means that the theory used needs to be applicable in a multicultural context), empirical (this implies that for a multicultural context the criterion being measured need to only discriminate on the specific criterion and not on other variables) or criterion-referenced (this implies that the development of the cross STUDENT NO: 44674481 11 cultural criterion need to be developed by a panel representing the different cultural groups).
The different specifications and dimensions of tests should then be tabulated (Foxcroft, 2004). (c) Choosing the test format and item format, specifying the administration and scoring methods In choosing the best method to present a test it is important that the tools (paper vs. computer), the format in which it is presented (multiple choice, diagrams, etc) and the response methods (verbal, written, etc) take into account the capability and familiarity across the different cultural groups.
The administration and the scoring methods should also be fair across cultural groups (Foxcroft, 2004). Moerdyk (2009) identifies seven key steps in test development. The first step conceptualising involves identifying what phenomenon is trying to be understood. The second step is identifying the observable measures of the phenomenon (operationalising). The third step involves quantifying the observable measures. This requires content validity, distracters and a response set.
The next step involves a pilot session of the test. By administering the test to a pilot group that is a representation of the final test group, the opportunities for improving the test can be identified. Once the test is administered to the pilot group the next step is to conduct an item analysis and the correlation of items identified. The item analysis can then be used to revise the test and compile the final test (what is known as validity shrinking). The last step is to develop a set of norms that can be used to measure the responses against for interpretation (Moerdyk, 2009).
Lastly, the test needs to be published and the HPCSA provides guidelines for test development to ensure fair and ethical standards are met. De beer (2006), research findings indicate that for a South African multicultural and multilingual context the need for dynamic assessment is important as it accounts for differences across cultural groups as well as identify opportunities for further development. Other issues that are challenging to psychologists are the influence of other factors such as age, gender, socio-economic status, environment and cultural differences on test performance (Moerdyk, 2009).
Some considerations for the future of assessment in South Africa are the use of technology advances in testing (artificialintelligence), the need to develop more culture specific tests, and the need for fair and ethical practices of psychometric testing (Moerdyk, 2009). 2. 4. 2. Core Characteristics of psychometric test Some of the core characteristics that psychometric tests should adhere to can be summarized as follows: ? It is based on standardized procedures and methods of assessment; ?
It makes use of norms, comparing individuals performance against a category or norm group; ? It is scientifically proven to be valid (i. e. , it measures what it says), and reliable (the consistency and accuracy of STUDENT NO: 44674481 12 measuring instruments); ?
It can be applied in various institutions (health, education, occupation, etc. ) and the measures can be cross- culturally adapted with minimal test bias; ? It can be fairly administered to all individuals groups and organisations (Foxcroft & Roodt, 2005). In South Africa psychometric testing is regulated by the Employee Equity Act of 1998 which prohibits the use of psychometric tests or assessments unless it proves to be scientifically valid and reliable, is fair and free from bias (Foxcroft et al, 2005). 3.
From the above discussion it can be concluded that an understanding of the developments of psychological assessment is imperative to understanding the value in current society and in highlighting aspects to focus on change. In the past, psychological tests developed were used internationally in psychological assessments despite the differences in culture and language. This produced issues of bias and unfairness, as can be seen in misuse of assessments in the apartheid era in South Africa (Foxcroft et al, 2005). For a long time psychological tests in South Africa were perceived as unfair and unjust (Sehlapelo & Terre Blanche, 1996).
However, with the political changes in government and legislature ( such as the Employment Equity Act 1998) , the introduction of governing bodies (such as HPCSA) and the changes in ideologies (a move toward a dynamic approach in assessment), psychological assessment has progressed toward a more fair and ethical practice that is slowly changing these perceptions (Mauer, 2000). Challenges still do exist, however, in producing tests that can be applied in such a diverse country that has 11 official languages, and a vast number of cultural groups as well as moving toward testing using the advances in technology.