An extensive review of literature revealed three trends in assessing HR practitioners’ roles, responsibilities and competencies in the South African workplace. Studies to date adopted global or consultancy measures to assess HR practitioners’ roles and the application of HR practices in the workplace. Secondly, authors developed their own measurements to assess the application of global HR frameworks in the South African context. Finally, authors develop their own measures to assess specific areas of the HR field, such as competencies, ethics and professionalism. Only nine quantitative research studies could be found that focused specifically on the assessment of the roles, practices and competencies of HR practitioners in South African organisations. A critical overview of the assessment of HR roles, responsibilities and practices, as well as their applicability in the South African context, is provided hereunder.

Steyn (2008) and Walters (2006) used the Human Resource Role Assessment of Ulrich (1997) to explore the strategic role of HR professionals in the South African context. The survey consisted of 40 statements that measure four roles: strategic partner, change agent, employee champion and administrative expert, using operational phrasing and describes concepts, practices and activities of the HR function. In both studies, the measure yielded important information on the extent to which HR functions are allowed to play a strategic role in South African organisations. Although this measure is globally still the predominant measure of HR roles and competencies in the workplace, future studies should expand the measure to give an accurate reflection of the HR profession in the South African context.

Magau and Roodt (2010) developed a survey to assess the application of the Human Capital BRidge^TM framework (Boudreau & Ramstad, 2007) in the South African context. The 27-item questionnaire measures operational excellence, growing the company and securing the future and representing the company’s strategic themes. Although this survey yielded promising results, the authors acknowledged that the Human Capital BRidge^TM framework was not applied to the fullest extent to guide the measurement solutions to assist both HR and line managers to measure the contribution of human capital towards strategic objectives.

Schultz (2010) developed a measure to assess the effectiveness of HR competencies at a merged higher education institution. This questionnaire was developed based on the HR balanced scorecard of the institution and assesses three HR competencies: business knowledge, HR practices and personal and management skills. Although this study was narrow in its scope in terms of assessing the different types of HR competencies, it provided valuable insights into the need for HR competence assessment in the South African context.

Scheepers and Schuping (2011) applied the Human Resources Practices Scale of Geringer, Colette and Milliman (2002) to determine the effect of HR practices on the psychological contract in a mining company. The questionnaire consisted of 50 items and measured five HRM practices, namely communication, hiring practices, training and development, performance appraisal and remuneration practices. This questionnaire was extensively validated in more than 40 countries and showed promising results for use in both developed and developing contexts. Exploratory factors analysis showed several cross-loadings between the items of the five factors (Scheepers & Schuping, 2011). Furthermore, the instrument is limited in its assessment of the full scope of HR practices in the workplace. This survey should thus be further validated in the South African context and could be expanded to include a broader scope of HR practices.

Coetzee et al. (2014) used an organisational culture survey of Deloitte and Touche (2009) to assess the extent to which HR practices predict the commitment of engineering staff. This survey measures 13 organisational HR practices: change, communication, diversity and transformation, HR policies and procedures, innovation, job satisfaction, leadership, management style, performance management, recognition, rewards and remuneration, training and development and values and culture. Although this survey was sufficient for the purposes of that study, its scope is rather broad; there should be more focus on the specific roles and requirements of HR professionals in the South African context.

Other researchers developed measures to assess specific areas of the HR role in the South African context. Van Wyk (2006) developed a measure to assess the extent to which HR practitioners display knowledge and apply competencies regarding social responsibility issues such as HIV/AIDS in the workplace. The findings revealed a gap in the knowledge of HR practitioners at all organisational levels on managing HIV/AIDS. Van Vuuren and Eiselen (2006) developed a measure to assess the role of HR professionals in corporate ethics. The results showed that HR practitioners generally believe that they indeed have an ethics management competency and that they should be involved in ethics management. Van der Westhuizen et al. (2003) developed a measure to assess the status of HR management (HRM) as a profession in the South African context. The results of their study showed that HR practitioners do consider HRM to be a profession, but that it still needs to achieve true professional status. The above studies provided useful information on specific areas within the HR professional’s function and should encourage other researchers to conduct more detailed research on other functional areas of HRM.

To summarise, it is clear from the above discussion that a holistic and accurate measure of HR competencies within the South African context is lacking. Although most of the studies discussed above developed measures to assess various areas relating the HR function, none focused on the validation of the measures. These studies, however, provided the much-needed preliminary information, which led to the development of the SABPP’s South African HR Competency Model (SABPP, 2012). The SABPP identified 14 competency domains that HR professionals should be able to display in the South African workplace. These are: leadership and personal credibility, organisational capability, solution creation and implementation, interpersonal and communication skills, citizenship for the future, strategy, talent management, HR governance, risk and compliance, analytics and measurement, HR service delivery, HR and business knowledge, ethics, professionalism and duty to society.

Schutte (2015) used this model in conjunction with other models (Ulrich, 1997) as a guideline to develop an HR competency measure for HR professionals in the South African context. The measure consists of three dimensions, namely Professional behaviour and leadership, Service orientation and execution and Business intelligence. This survey, furthermore, measures 12 competencies associated with the three domains. The pilot study of this measure yielded promising results, with clear factor structures and high reliabilities for the three dimensions (Schutte, 2015). The present article focuses on the further validation of this measure in the South African context.

Not only is it important to establish reliable and valid methods of measurement with regard to HRM competencies, it is also important to consider the cultural diversity in a multicultural setting, such as South Africa. As the HRM Professional Competency Model is a new measure, an assessment of this nature should be concerned with construct equivalence. Assurance of measurement equivalency across different cultures is important for determining whether measures of a construct in one culture also exist in other cultures, to determine the degree of variability of the measures across several cultures and to determine whether the measures are universal or culture-specific (He & Van de Vijver, 2012). Furthermore, assessing measurement equivalence across countries and language groups provides information about the factorial invariance of an instrument and therefore allows the researcher a degree of confidence in using the instrument in two or more cultural settings (Van de Vijver & Tanzer, 2004).Figure 1 presents a model of the HR competency measure for HR professionals.

In light of the preceding section, the following research hypotheses and sub-hypotheses were formulated:

The research approach of the present study was quantitative in nature, as the aim of the research was to validate a HR competence measure. A non-experimental cross-sectional survey research strategy of inquiry was utilised in this study, based on the need for exploratory research on HR competencies in the South Africa workplace (Field, 2009).

A quantitative research approach was followed. A cross-sectional survey design was used to collect data and attain the research goals (Field, 2009). The present research mostly followed the DeVellis (1991) scale development process, as the aim was to develop and validate a diagnostic tool for HR professionals’ competencies in the South African context. The steps of the DeVellis scale development process, as well as the extent to which these were applied in the present study, are indicated in Box 1.

This article reports the results relating to Steps 7 and 8 of the DeVellis (1991) scale development process. The results of the research pertaining to steps 1–6 of the scale development processes are reported elsewhere (Schutte, 2015). This article reports on the final validation of the HR professional competency measure.

Permission was obtained from the necessary authorities prior to the administration of the questionnaire. Hard copy and electronic surveys were distributed to the relevant participants. Ethical clearance was obtained prior to the administration of the surveys. Confidentiality was maintained at all times.

Data were analysed with the aid of SPSS 22.0 for Windows. Descriptive statistics (i.e. means, frequencies, skewness and kurtosis) were applied. Construct equivalence was determined. The motivation for the application of these techniques was to obtain a structure in each language group that could then be compared across all language groups involved. Factor analysis is the most frequently employed technique for studying construct equivalence (Van de Vijver & eung, 1997). In the current study, both exploratory and confirmatory models could have been used. Since this is a newly developed instrument, the researcher used exploratory factor analysis.

Factors obtained in each group were compared (after target rotation). The agreement was evaluated by a factor congruence coefficient, Tucker’s phi (Van de Vijver & Leung, 1997). Values above 0.90 were taken to indicate essential agreement between cultural groups, whilst values above 0.95 indicated very good agreement. A high agreement implies that the factor loadings of the lower and higher levels are equal, up to a multiplying constant.

Exploratory factor analysis, using principal component analysis, was done on the 36 items of the Professional behaviour and leadership competency dimension of the questionnaire. The initial analysis revealed that four factors could be extracted, based on the eigenvalues for both language groups. Based on the results of the pilot study, a subsequent principal component factor analysis, using varimax rotation, was done to specify the four underlying factors. Varimax rotation was chosen as it resulted in the most appropriate underlying factor structure for the Professional behaviour and leadership competency dimension. The Organisational capabilities factor was deleted, because of problematic loadings (six items). Three more items were deleted to further refine the questionnaire. A subsequent factor analysis using varimax rotation revealed four underlying factors, which explained 82.974% of the variance for the English language group and 81.037% of the variance for the indigenous languages group. The factors for the English language group were as follows: Innovation (Factor 1), Leadership and personal credibility (Factor 2), Interpersonal communication (Factor 3) and Solution creation (Factor 4). The factors for the indigenous language group were as follows: Leadership and personal credibility (Factor 1), Interpersonal communication (Factor 2), Innovation (Factor 3) and Solution creation (Factor 4). The results revealed equivalence for both language groups in terms of the specified factors and items per factor. The original Citizenship factor was relabelled Innovation. The four factors per language groups are reported in Table 1.

The rotated component matrices of the four-factor solutions for English and indigenous language participants were then used as input for an exploratory factor analysis with target rotations. The following Tucker’s phi coefficients were obtained:

These coefficients compare favourably with the guideline of 0.90 and can therefore be regarded as acceptable for equivalence for both language groups.

The above results partially confirm Hypothesis 1. The factor analysis revealed four underlying factors for the Professional behaviour and leadership competency dimension, which is in contrast with the five factors initially specified for the measure. All factors showed acceptable Cronbach’s alpha coefficients and construct equivalence for both the English and the indigenous language groups.

An exploratory factor analysis, using principal components analysis, was performed on the 37 items of the Service orientation and execution competency dimension of the questionnaire. The initial analysis revealed that four factors could be extracted, based on their eigenvalues. A subsequent principal component factor analysis, using varimax rotation, was performed to specify the four underlying factors for the two language groups. Varimax rotation was chosen as it resulted in the most appropriate underlying factor structure for the Service orientation and execution competency dimension. Four items were deleted, due to problematic loadings. The four factors explained 81.660% of the variance for the English language group and 79.693% of the variance for the indigenous languages group. A further 11 items were deleted to further refine the questionnaire. The four factors for the language groups were Talent management (Factor 1), HR service delivery (Factor 2), HR metrics (Factor 3) and HR risk (Factor 4). The original HR analytics and measurement factor was relabelled HR metrics. Furthermore, the original HR governance factor was relabelled HR risk. The rotated component matrix is reported in Table 2.

The rotated component matrices of the four-factor solutions for the English and indigenous language groups were then used as input for an exploratory factor analysis with target rotations. The following Tucker’s phi coefficients were obtained:

These coefficients compare favourably with the guideline of 0.90 and can therefore be regarded as acceptable for equivalence for both language groups.

The above results confirm Hypothesis 2, namely that the Service orientation and execution competency dimension consists of four factors, Talent management, HR risk, HR metrics and HR service delivery. All factors showed acceptable Cronbach’s alpha coefficients and construct equivalence for both the English and the indigenous languages groups.

The factor analysis of the Business intelligence competency dimension was performed on the original 38 items, as specified in the discussion of the pilot study and the Organisational capability factor, which originally formed part of the Professional behaviour and leadership dimension. An exploratory factor analysis using principal components analyses was performed on the 44 items of the Business intelligence competency dimensions of the questionnaire. The initial analysis revealed that four factors could be extracted, based on the eigenvalues. A subsequent principal component factor analysis, using varimax rotation, was performed to specify the four underlying factors relating to the Business intelligence questionnaire. Varimax rotation was chosen as it resulted in the most appropriate underlying factor structure for the Business intelligence competency dimension. The four factors explained 78.535% of the variance for the English language group and 76.181% of the variance for the indigenous languages group. Fourteen items were deleted to refine the questionnaire further. The four factors for both language groups were labelled Strategic impact (Factor 1), HR business knowledge (Factor 2), HR business acumen (Factor 3) and HR technology (Factor 4). The results of the rotated component matrix are reported in Table 3.

The rotated component matrices of the four-factor solutions for the English and indigenous language groups were then used as input for an exploratory factor analysis with target rotations. The following Tucker’s phi coefficients were obtained:

These coefficients compare favourably with the guideline of 0.90 and can therefore be regarded as acceptable for equivalence for both language groups.

The above results partially confirm Hypothesis 3. The factor analysis revealed four underlying factors for the Business intelligence competency dimension, in contrast to the three factors initially specified. All factors showed acceptable Cronbach’s alpha coefficients and construct equivalence for both the English and the indigenous language groups.

The descriptive statistics of the measures are reported in Table 4. The results show excellent reliabilities for the factors of all the dimensions (Field, 2009). The results further show that the dimensions are negatively skewed. The mean scores indicate that the respondents viewed all the dimensions of the HR Professional Competence Model as important, with the factors Interpersonal communication, Leadership and personal credibility and HR business knowledge indicated as being the most important.