THE DEVELOPMENT OF CROSS-CULTURAL PERSONALITY INSTRUMENT FOR THE SOUTH AFRICAN CONTEXT

 

Introduction

​

There is a growing interest in the measurement of personality variables in applied settings, such as in selection, placement,  therapeutic intervention and counseling in South Africa. In particular “personality in the workplace” has been widely studied in the last decade. Personality variables and issues related to their use, especially in work settings have generated a vast amount of interest, research and publications. The importance of personality to industrial, work and organizational psychology is now apparent with meaningful relationships between personality variables and criteria. Criterion-related validities for predicting work-related constructs reveal the importance of personality variables in understanding and predicting work performance (Van Aarde, Meiring &  Wirniek, 2017).

 

A further important issue relates to the fair application of personality measures to diverse cultural groups. Current personality instruments being used in South Africa have been imported from elsewhere (often Anglo-Saxon countries). Little effort has been invested in making these instruments suitable for the South African context and also being Employment Equity compliant (e.g., Employment Equity Act, 2013, Section 8 (a),(b),(c) & (d)). The SAPI project explores to what extent personality is universal and culturally specific within the South Africa contexts. In the SAPI project we follow the Global-local approach (GloCal) that is more likely to yield a comprehensive picture of personality by combining approaches informed by a thorough understanding of that language and culture (Daouk-Öyry, Zeinoun, Choueiri and Van de Vijver, 2016). The GloCal approach allows researchers to a) identify shared and unique components of personality across cultures b) ensure that the lexicon used is relevant to the culture and c) increase the ecological validity of stimulus materials in personality inventories.

 

Project Aims

​

Overall aim and research objectives

​

• The SAPI project aims to developing a single, unified personality inventory for South Africa that takes into consideration both universal and unique personality factors to be found across the various culture groups in South Africa.

• This personality inventory will be developed, standardized and  submitted for classification to the Psychometrics Committee of the Professional Board for Psychology (Health Professions Council of South Africa - HPCSA). In November 2014 the SAPI was submitted to the HPCSA for classification as psychometric test in development and have been classified as such ref: 18/11/178. 

 

Main research question

​

• How can we develop a new personality inventory in the South African context that can be applied fairly to all language (cultural) groups.

• In addition, the instrument should be applicable in all domains of assessment of normal (i.e. non-clinical) personality, such as job selection, and career counselling.

• Personality has been shown to be able to predict relevant psychological outcomes of work and personal performance. We are interested in the SAPI  predictive validity of various criteria domains.

 

Specific research objectives

 

• The project aims to develop a comprehensive questionnaire to assess personality among all South-African language groups. Comprehensiveness of the measure should be interpreted as covering all major aspects of personality as deemed relevant in a South-African context.

• Practically speaking this means that the project does not start from well-known conceptualizations of personality such as Costa and McCrae’s Big-Five or Eysenck’s “Giant Three”; rather, the project tries to start from everyday conceptualizations of personality as found in South African languages.

• On the one hand, we are interested to see to what extent the personality structure found in Western studies is applicable in the various South African groups and if the latent model can be replicated across the different languages.

• We are also interested in investigating the measurement invariance (MI) in a multi-group measurement model by exploring configural invariance (i.e., factorial invariance), metric invariance and scalar invariance.

 

Motivation for this study

​

The study by Meiring, Van de Vijver, Rothmann and Barrick (2005) clearly demonstrated that psychological instruments imported from abroad could have a limited suitability for South Africa. It was only after a series of item adaptations that acceptable psychometric properties were found. Similarly, the study by Boeyens and Taylor (1991) showed that even an instrument that was developed specifically for South Africa (the South African Personality Questionnaire), showed shortcomings in various items. In addition, there is a pressing societal need to develop more culture-sensitive psychological instruments. This development is in line with developments in other countries. For example, in the USA test standards are becoming increasingly important in the acceptance and usage of instruments (APA/AERA/NCME, 2014). The International Test Commision (ITC) (2016) released new guidelines for translating and adapting of test (second addition). Fairness to all cultural (language) groups is an important standard. In addition, test takers become more vocal about their rights and individual psychologists or psychology as a profession will be held accountable for the (im)proper usage of instruments. The Employment Equity Act creates a daunting task for psychology as a profession as it loads the burden of the proof in the profession. An additional reason for undertaking the study is that the development of such a personality inventory might contribute to the development of an indigenous personality psychology. It is expected that the personality inventory might become a useful research tool in the South African context.

 

The current situation with regard to personality testing in South Africa seems to suffer from two problems:

​

1.    Almost no questionnaires that have been shown to be adequate for the multicultural and multilingual South-African context.

2.   It is not at all clear whether the “imported” instruments are adequate to cover all aspects of personality that are relevant in South Africa.

 

Project Plan

 

The project has several stages. Fanny Cheung and co-workers from the Chinese University of Hong Kong have carried out a similar study, which was aimed at the development of a personality questionnaire in China (Cheung, Cheung, & Fan (2013). The approach used in this project was adapted by the current project team so as to maximize its adequacy for the South African context. The first stage was conceptual in that current literature was reviewed and a number of field studies of personality were conducted in the various languages. The information obtained from these studies were integrated so as to develop a unified personality structure encompassing various sub-clusters and facets. The personality structure was used as a framework to investigate the psychometric properties in the second stage of the project, which involved item generation and assessment and factorial validity. We are currently in the final stage where we are developing a test manual for the SAPI which will be submitted to the Psychometrics Committee as an instrument for classification. The SAPI is already registered as an instrument in development at the Professional Board for Psychology of the Health Professions Council of South Africa.

 

Research methodology

 

The overall structure of the project involves six stages:

​

1. The establishment of implicit perspectives of personality in all South-African language groups by analysing  interviews;

2. The synthesis of all these perspectives and the development of  a single instrument that has a core of items that are common for all groups and, if needed, a set of culture-specific items;

3. The administration of the instrument in all language groups and the establishment of equivalence of the instruments across these groups.

4. The refinement of the instrument to a suitable instrument that can be used in the South African context.

5. Validation of the instrument in the South African society by conducting validation studies in various institutions and business organizations.

6. Developing a South African representative norms.

7. Submit the instrument to the Psychometrics Committee of the Professional Board for Psychology of the Health Professions Council of South Africa for classification.

 

The project spanned two stages: a qualitative stage of conceptual model development and a quantitative stage of instrument development.

 

Stage 1:

​

The qualitative stage aimed to uncover the implicit personality conceptions in all 11 languages, as manifested in free personality descriptions. By employing the emic-etic approach, more than 1200 participants across all eleven official languages were interviewed in their native language and their responses were translated into English by language editors. These English responses were subsequently semantic-content analysed (see Nel et al., 2012). The researchers then identified the unique facets (using the emic approach) which were only found in certain languages, and also generated common facets (using the etic approach) which were found across the different languages. During this stage 188 facets were generated, and by the use of cluster analysis further grouped towards 37 sub-clusters. In the end, after many discussions and workshops which involved linguistic and cultural experts, nine broad clusters were identified (Nel et al., 2012; Valchev et al., 2013). These clusters were labelled Extraversion, Soft-Heartedness, Conscientiousness, Emotional Stability, Intellect, Openness, Integrity, Relationship Harmony, and Facilitating. This phase formed the groundwork for the development of an indigenous personality instrument for the South African context, namely the South African Personality Inventory (SAPI).

 

Stage 2:

​

Quantitative Instrument Development

​

Items were generated in English with input from the content of the free descriptions obtained by Nel et al. (2012). On average, at least 10 items were developed for each of the 188 facets; there were 117 to 482 items per cluster, and the total number of items was 2,574. The item generation rules were similar to those used for the FFPI (Hendriks, Hofstee, & De Raad, 1999). Items were formulated in the first person singular, used simple language and no negations, and specified concrete behaviors expressed with an object whenever possible (e.g., “I care for others” and “I help others cope with their problems”) (Hill, Nel, Van de Vijver, Meiring, & Valchev, 2013). The decision to use concrete behaviors was based on the finding that concrete expressions were favored by Blacks (Adams, Van de Vijver, & De Bruin, 2011; Valchev et al., 2013) and on literature pointing to improved cross-cultural replicability of psychological constructs when concrete behavior manifestations are used (Hendriks et al., 2002; Ramsay, Taylor, De Bruin, & Meiring 2008; see Hill et al., 2013, for further details on the item development).  The item selection was performed in pilot studies, where questionnaires for each cluster were administered separately to students and police service employees (samples included 439 to 1,023 participants per cluster; see Hill et al., 2013). Both psychometric and substantive criteria were used in iterative steps. Items were removed if they had extreme mean values (below 1.50 or above 4.50 on a 5-point Likert scale from 1 to 5), skewness (absolute value above 2), or kurtosis (above 4). Principal component analysis was conducted separately on facet and cluster level, focusing on the first component to construct homogeneous scales. Items were subsequently subjected to hierarchical factor analysis with Schmid-Leiman transformation (Schmid & Leiman, 1957). In all analyses, items with loadings of at least .30 on both the higher and lower level factors were retained; where that value did not yield sufficient distinction, .40 was used as a cutoff point. With respect to the substantive criteria, we selected items that (a) maximized construct representation, (b) minimized content overlap within and across clusters, and (c) were most in line with the formulation rules of behavior focus, simple language, and translatability.  Applying these criteria separately per cluster, the item set was reduced to 571.

 

These 571 items were translated by professional translators from English into all 10 other official languages, and the translations were checked by independent language experts (Fetvadjiev et al., 2015; Nel et al., 2016). The translators provided comments on the linguistic and cultural adequacy of the items for their respective language group; only a few items appeared hard to translate and were discarded. Finally, an attempt was made to avoid complex items; items longer than 10 words in any language and items employing abstract trait terms were removed, leading to a reduced set of 250 items.  The 250 items were administered to a large, multi-ethnic sample (Fetvadjiev et al., 2015).  Factor and internal-consistency reliability analyses were performed separately on the items from each cluster. Items that reduced reliability were removed in iterative steps, and only items with loadings over .30 (or .40) were retained. The factor structure was compared across groups, and items were replaced to obtain optimal factor replicability. The final item set contained 176 items grouped in 18 facet scales and an additional 12 social desirability items. In total the SAPI questionnaire consisted of 188 items. The final factor structure consisted of a six-dimensional structure (comprising a positive and a negative Social-Relational factor, Neuroticism, Extraversion, Conscientiousness, and Openness) (Fetvadjiev et al., 2015; Nel et al., 2016).  The SAPI (188 items) is available online for completion (click here).

 

The 188 item SAPI questionnaire showed adequate results in further studies, especially pertaining to construct, discriminant, concurrent and predictive validity (Bruwer, 2016; Combrink, 2017). Both confirmatory and exploratory factor analysis generated acceptable results in samples containing professionals and emerging adults. Pertaining to external validity, it was found that the six-dimensional structure had links with cultural intelligence (Ang, Van Dyne, Koh, & Ng, 2004), psychological well-being (Ryff, 1989), vocational interest (Pozzebon, Visser, Ashton, Lee, & Goldberg, 2010) and student-course engagement (Handelsman, Briggs, Sullivan, & Towler, 2005).

 

References

 

Adams, B. G., Van de Vijver, F. J. R., & De Bruin, G. P. (2012). Identity in South Africa: Examining self descriptions across ethnic groups. International Journal of Intercultural Relations, 30, 377–388.

​

AERA/APA/NCME. (2014). Standards for educational and psychological testing. Washington, DC: American Psychological Association.

Ang, S., Van Dyne, L., Koh, C. K. S., Ng, K. Y., Templer, K. J., Tay, C., & Chandrasekar, N. A. (2007). Cultural intelligence: Its measurement and effects on cultural judgment and decision making, cultural adaptation, and task performance. Management and Organization Review, 3, 335–371.

​

Bruwer, M. (2016). Assessing the nomological network of the South African Personality Inventory (Unpublished master’s dissertation). North-West University, Potchefstroom, South Africa.

​

Combrink, M. (2017). Assessing personality, vocational interest and student-course engagement among students (Unpublished master’s dissertation). University of Pretoria, Pretoria, South Africa.

​

Daouk-Öyry, L., Zeinoun, P., Choueiri, L., & Van de Vijver, F. J. R. (2016). Integrating global and local perspectives in psycholexical studies: A GloCal approach. Journal of Research in Personality. doi: http://dx.doi.org/10.1016/j.jrp.2016.02.008.

​

Cheung, F. M., Cheung, S. F., & Fan, W. (2013). From Chinese to cross-cultural personality inventory: A combined emic–etic approach to the study of personality in culture. In M. Gelfand, C.-Y. Chiu, & Y.-Y. Hong (Eds.), Advances in culture and psychology (Vol. III, pp. 117–179). New York, NY: Oxford University Press.

​

Employment Equity Act (2013). Employment Equity Act, No. 47 of 2013. Republic of South Africa. http://www.labour.gov.za/docs/legislation/eea/act98-055.html. Accessed on 28 September 2014.

​

Fetvadjiev, V. H., Meiring, D., Van de Vijver, F. J. R., Nel, J. A., & Hill, C. (2015). Psychological assessment the South African Personality Inventory (SAPI): A culture-informed instrument for the country’s main ethnocultural groups. Psychological Assessment, 27(3), 827-837. doi:org/10.1037/pas0000078.

​

Handelsman, M. M., Briggs, W. L., Sullivan, N., & Towler, A. (2005). A measure of college student course engagement. The Journal of Educational Research, 98(3), 184–192. http://doi.org/10.3200/JOER.98.3.184-192

​

Hendriks, A. A. J., Hofstee, W. K. B., & De Raad, B. (1999). The Five-Factor Personality Inventory (FFPI). Personality and Individual Differences, 27, 307-325. doi:10.1016/S0191-8869(98)00245-1

​

Hill, C., Nel, J. A., Van de Vijver, F. J. R., Meiring, D., Valchev, V. H., Adams, B. G., & De Bruin, G. P. (2013). Developing and testing items for the South African Personality Inventory (SAPI). SA Journal of Industrial Psychology, 39, 1-13. doi:10.4102/sajip.v39i1.1122

​

International Test Commission. (2016). The ITC Guidelines for Translating and Adapting Tests (2nd ed.). [www.InTestCom.org].

​

Meiring, D. Van de Vijver, F.J. R., Rothmann. S., Barrick. M (2005). Construct, item, and method bias of cognitive and personality tests in South Africa. South African Journal of Industrial Psychology, 31(1), 1 – 8.

​

Nel, J. A., Fetvadjiev, V. H., Rothmann, S., Van de Vijver, F. R. J., Meiring, D., & Hill, C. (2016). Developing psychometric instruments for a specific context: The South African Personality Inventory. In R. Ferreira (Ed.), Thinking innovatively about psychological assessment in a context of diversity (pp. 86 – 101). Cape Town, South Africa: Juta

​

Nel, J. A., Valchev, V. H., Rothmann, S., Van de Vijver, F. J. R., Meiring, D., & De Bruin, G. P. (2012). Exploring the personality structure in the 11 languages of South Africa. Journal of Personality, 80, 915-948. doi:10.1111/j.1467-6494.2011.00751.x

​

Pozzebon, J. A., Visser, B. A., Ashton, M. C., Lee, K., & Goldberg, L. R. (2010). Psychometric characteristics of a public-domain self report measure of vocational interests: The Oregon Vocational Interest Scales. Journal of Psychological Assessment, 92(2), 1–13. http://doi.org/10.1080/00223890903510431.

​

Ramsay, L. J., Taylor, N., De Bruin, G. P., & Meiring, D. (2008). The Big Five personality factors at work: A South African validation study. In J. Deller (Ed.), Research contributions to personality at work (pp. 99-114). Munich, Germany: Rainer Hampp Verlag.

​

Ryff, C. D. (1989). Happiness is everything, or is it? Explorations on the meaning of psychological well-being. Journal of Personality and Social Psychology, 57(6), 1069-1081.

​

Schmid, J., & Leiman, J. M. (1957). The development of hierarchical factor solutions. Psychometrika, 22, 53-61. doi:10.1007/BF02289209

Taylor, T. R., & Boeyens, J. C. A. (1991). A comparison of black and white responses to the South African Personality Questionnaire. Pretoria, South Africa: Human Sciences Research Council.

​

Valchev, V. H., Van de Vijver, F. J. R., Nel, J. A., Rothmann, S., & Meiring, D. (2013). The use of traits and contextual information in free personality descriptions across ethnocultural groups in South Africa. Journal of Personality and Social Psychology, 104, 1077–1091. doi:10.1037/a0032276

​

Valchev, V. H., Van de Vijver, F. J. R., Meiring, D., Nel, J. A., Laher, S., Hill, C., & Adams, B. (2014). Beyond Agreeableness: Social-relational personality concepts from an indigenous and cross-cultural perspective. Journal of Research in Personality, 48, 17–32. doi:10.1016/j.jrp.2013.10.003

​

Van Aarde, N., Meiring, D. & Wiernik, B.M. (2016 in review). The Validity of the Big Five Personality Traits for Job Performance: Meta-analyses of South African Studies. Special issue Career Development International.

​