Includes bibliographical references (leaves 69-73) / This study explored the relationship between absenteeism and commitment (affective, continuance and normative commitment) as directed towards the organisation, co-workers and the nursing profession. Job satisfaction, job involvement, career stage and the absence culture were examined as moderators of the relationship between absenteeism and commitment. The sample comprised of 227 public sector nurses (54% response rate) from 11 day clinics and hospitals within the Western Cape Metropolitan District Health Services of South Africa. Affective, continuance and normative commitment to the organisation were not directly related to absenteeism, although interactions between these commitment components were significant predictors of absenteeism. Affective commitment to co-workers did not significantly explain absenteeism. However, affective commitment to the nursing profession explained significant variance in nurse absenteeism both directly and through the moderation effects of the absence culture of the workplace and the profession. Job satisfaction, job involvement and career stage of the nurses did not moderate the absenteeism-commitment relationship. The contributions of the study are discussed and recommendations for future research are made.

Includes bibliographical references (leaves 82-86). / Assessment centres are used to make decisions for employee advancement and development within organisations. These centres comprise of a number of instruments which are used to establish competencies needed to perform effectively. The validity of this measure is yet to be fully ascertained as previous studies have shown that although assessment centres have predictive validity, they lack construct validity. The research study therefore set out to establish whether assessment centres used within financial institutions have predictive validity but lack construct validity. For purposes of this study, two sets of data each comprising of responses from 120 and 91 employees were used. This data was retrieved from a financial institution within the Western Cape. Contrary to previous studies, there was low predictive validity within this study. Some of the factors that could have contributed to the low correlation between performance and assessment ratings include a rather small sample that was used, having different criterion being measured in the evaluations than in the assessment centres, range restriction, complexity of assessing behaviour, and having a number of dimensions from which to derive an Overall Assessment Rating (OAR). However, similar to previous studies, this study also found low construct validity amongst the ratings. In addition, a further analysis indicated that both performance and personality measures are valid methods in predicting performance.

Includes bibliographical references.

This study explores how Postgraduate Diploma of Accounting (PGDA) students at the University of Cape Town consider potential employers when making employer-choice decisions. Kelly's Repertory Grid technique was employed during structured interviews with twelve (N = 12) PGDA students. Perceptions about potential accounting employers varied across the participants. Investec and Nedbank, both of which are Banks, were construed similarly for 11 out of 12 participants. On average, three of the Big Four auditing firms were mostly construed similarly, namely EY, PwC and Deloitte, with KPMG construed differently by the majority of participants. Nolands and Mazars were construed similarly for 8 participants. Transnet was perceived distinctly from the other firms by most participants. The most frequently elicited constructs were regarding progression opportunities, international exposure and ethical reputation. The following themes emerged from a thematic analysis of the participants' interview responses: (1) organizational attractiveness, (2) exposure gained during training, (3) work environment, (4) progression opportunities, (5) diversity policies, (6) brand awareness, (7) workplace flexibility and work-life balance, (8) the recruitment process, (9) corporate social responsibility, and (10) a felt moral responsibility to the employer. This study proposes that by understanding how students construe and perceive different sets of potential accounting employers, employers could improve their attraction and retention strategies. The findings of this study could also be of benefit to career counsellors and others tasked with advising and guiding accounting graduates.

The Psychology of Working Theory (PWT) focuses specifically on the work-based experiences of low-income workers. It is thus a suitable theoretical framework to predict and explain the work experiences of individuals performing domestic work, one of the largest work sectors in South Africa. The purpose of this study was to determine whether the theoretically derived predictors of decent work, in the PWT, could be empirically supported. Domestic workers, in Cape Town and Johannesburg, participated in a self-report survey (N = 139), which consisted of several measures. These participants were accessed through a variety of convenience and snowball sampling techniques. As expected, exploratory factor analysis revealed that the scales used to measure marginalisation, economic constraints, work volition and proactive personality were one-dimensional, and the decent work scale was fivedimensional (complementary values, access to health care, adequate compensation, free time and rest, and safe work conditions). Unexpectedly, the social support scale showed two dimensions (i.e. support from the community or friends and support from a special person or family). The findings supported the proposition that greater economic constraints were related to less decent work experiences, but marginalisation experiences were not related to the degree to which work was seen as decent, nor to work volition (mediator variable). The non-significant relationship between marginalisation, work volition and decent work changed when considering proactiveness as a moderator, however, this was only at low levels of proactiveness. None of the dimensions of social support served as moderators, though social support from the community or friends predicted work volition, and social support from a special person or family predicted the degree to which domestic workers experienced their work as decent. While there were mixed results, the findings of this study suggest that the PWT's antecedents and moderators may work differently in the domestic work sector. Future research should investigate this in the domestic work sector and other low-income samples.

Students at higher education institutions expect that their investment in education will be rewarded through positive employment outcomes. The dearth of research into graduates' personal circumstances which contribute to whether these expectations translate into reality was the starting point for this PhD thesis. Specifically, the thesis considered the role of identity development for success in the employment search. Erikson's and Arnett's theories of identity development and Côté's identity capital model were used as the theoretical basis to develop the Identity Capital Model of Graduate Employment (ICMGE). Erikson and Arnett proposed that gaining meaningful employment is a crucial task in an individual's development trajectory when moving from adolescence into adulthood. Côté's model explains under what condition this transition is likely to be successful: Individuals with greater access to resources, both tangible and intangible, are more agentic and thus in a better position to deal with identity formation challenges. The ICMGE thus proposed that graduates with more intangible identity capital, i.e. greater agentic personality, and greater access to tangible identity resources in the form of financial, human, social and cultural capital are more employable, which reflected in a greater chance of finding employment, a shorter time to find employment and higher quality employment. Given that in the South African context historically members of different racial and gender groups had unequal access to employment opportunities for which current employment legislation seeks to provide redress, race and gender were included as additional predictors of graduate employment. To test the ICMGE empirically, students' identity capital, race and gender were assessed via quantitative surveys, with data collected from N = 872 students in their final year of study at different higher education institutions in the Western Cape Province of South Africa. A year later, N = 508 of these participants provided data about their current employment situation in telephonic interviews. Contrary to expectations, not all forms of identity capital were related to one another. The strongest correlations emerged between financial and human capital, with weaker correlations with cultural capital. Social and psychological capital generally did not correlate significantly with other forms of identity capital. The level of identity capital differed by race, but no gender differences emerged. The ICMGE successfully predicted employment amongst 81.9% of the respondents. However, only race and cultural capital, in the form of type of secondary school and type of tertiary institution attended, and home language explained unique variance in the probability of gaining employment. Those who had attended former Model C or private schools, higher status tertiary institutions, were English speakers, and self-identified as white or coloured had a greater probability of gaining employment. Greater social capital, measured by the number of extracurricular activities participated in, was related to a lower probability of being employed. It is likely though that the indicator used for social capital was not appropriate, given that close to half of the employed respondents indicated having secured employment through social contacts. The quality of employment obtained was mostly predicted by race and agentic personality. Those who were more agentic in their approach to life, and those who identified as either white or coloured, had obtained higher quality employment. Financial capital and home language were the most relevant predictors of the time taken to gain employment. Those with greater financial capital and English or Afrikaans speakers spent longer looking for employment. Academic grades had little relevance in predicting whether or not graduates obtained employment. The study adds new knowledge to the graduate employability literature in that it shows that a theoretically derived graduate employability model can be applied to real-world conditions by predicting actual employment rather than a graduate's employment potential. The study also demonstrated the value of considering graduates' identity development and access to identity capital when considering their chances to secure employment, and in particular the quality of this employment. The ICMGE model only predicted small amounts of the variance in the employment variables, however. It is thus recommended that future research make use of instruments that are more sensitive to the intricacies of the different types of capital in larger and more representative samples.

This research examined work-family conflict among Hindu fathers in South Africa. Various authors have indicated that work-family conflict should be investigated in specific cultural contexts because the demands of work and family differ across various cultures. This research examined the relationships between work role overload, supervisor support, coworker support and work-family conflict. Additionally, this research also examined the moderating role of traditional Hindu cultural values (gender role ideology and family hierarchy orientation) in the relationships between work role overload, supervisor support, and work-family conflict. Secondary data from an earlier study that explored work-family conflict among Hindu parents in South Africa was used. The final sample consisted of 183 Hindu fathers involved in full-time work in South Africa. Exploratory factor analysis confirmed the bi-directionality of the work-family conflict construct. Hierarchical regression analyses showed that work role overload and co-worker support were significant predictors of work-to-family conflict (W2FC), whilst work role overload was a significant predictor of family-to-work conflict (F2WC). Results of the moderated regression analyses showed that gender role ideology did not moderate the relationships between work role overload, W2FC and F2WC. Family hierarchy orientation did not moderate the relationships between supervisor support, W2FC and F2WC. The findings of this research have practical implications for management and offer suggestions for future research.

While theorists and evaluation experts argue about the importance of evaluating training as a key business and training function, the practice around the globe has for many years been limited to the collection of reaction-level data (trainee satisfaction). Despite this trend, a chapter on training evaluation is still included in the most recent training and human resource development textbooks. Additionally, the practice of training evaluation still forms part of undergraduate and postgraduate human resource curricula. Yet we know little about the state of training evaluation in the South African context and its practical relevance in modern day corporate environments. This research therefore aimed to investigate a) the extent to which South African corporates are engaging in systematic training evaluation and what these practices are; and b) the extent to which South African corporates recognise the importance of, and use, training evaluation as a vehicle for organisational learning and change. Results indicate a disconnect between theory and practice in this area. South African corporates are highly committed to training due to their legal obligations and incentives to upskill employees but the same commitment for training evaluation does not exist. The corporates sampled are not interested or willing to perform training evaluations to determine the effectiveness of their training and development investments. These results raise the question of whether or not training evaluation, in its theoretically positioned conceptualisation, is still relevant in the highly demanding corporate environment. Based on the empirical findings, this thesis proposes a substantial shift in how training evaluation is conceived and conducted in order for organisations to engage meaningfully with the effectiveness of their training interventions and not waste opportunities for organisational learning and change.

This study investigated the model on the antecedents of proactive behaviour as identified by Den Hartog and Belschak (2012) within the context of entry-level graduate roles (n = 76). A survey was devised which included the use of a five-point Likert-type scale. It was then administered to graduates in entry-level roles in various industries in South Africa to measure the different variables stipulated by the model. When data was analysed, the results revealed that transformational leadership (inspirational), task-related role-breadth self-efficacy (RBSE), and people-related RBSE correlated significantly and positively with proactive behaviour. Transformational leadership (performance) and job autonomy obtained non-significant correlations with proactive behaviour. The results also revealed that job autonomy, task-related RBSE and people-related RBSE did not moderate the relationship between transformational leadership (inspirational or performance) and proactive behaviour. This meant that the display of transformational leadership did not lead to a significant increase in proactive behaviour in low autonomy, low RBSE situations or in high autonomy, high RBSE situations as hypothesised. The unique characteristics of entry-level graduate roles are highlighted by the study - the significance of this model on proactive behaviour in a general employee context potentially may not be relevant to a graduate context. The findings contribute towards research evidence on the development of proactive behaviour in entry-level graduate roles.

In the job interview literature, the positive effect of Frame-of-Reference (FOR) training on interviewer rating accuracy is well documented. However, how this training method increases rating accuracy is not well understood. The present study focused on rater individual difference characteristics as potential explanatory mechanisms for the effects of FOR training on accuracy. The researcher hypothesised that FOR training would enhance raters’ dispositional reasoning, rating motivation and self-efficacy, which, in turn, would increase their rating accuracy. A post-test only experimental research design was used in a sample of 32 students from a South African university. Participants were randomly assigned to the FOR training intervention or the no-training condition. Participants were required to rate three videotaped candidates on an interview competency and completed various individual difference measures. The FOR training intervention positively affected rating accuracy and findings suggest this effect may occur because of the influence of FOR training on dispositional reasoning, rather than FOR training enhancing rater motivation or self-efficacy. Study limitations and recommendations for future research are noted.