BURNOUT AND ENGAGEMENT IN FOURTH-YEAR SOCIAL WORK STUDENTS: WHY WE SHOULD BE CONCERNED

This article focuses on levels of burnout as well as engagement experienced by fourth-year undergraduate social work students attending a South African university. By means of quantitative procedures, data were obtained from 43 final-year social work students who completed a self-administered questionnaire. Respondents presented with above-average levels of exhaustion, professional efficacy, vigour, dedication and absorption, and below average levels of cynicism. Non-parametric procedures revealed significant differences in the levels of burnout related to population group, place of origin and household income status. Recommendations focus on monitoring the wellbeing of students and mobilising support for students who appear more vulnerable to burnout and low academic engagement.


INTRODUCTION
Burnout was considered a social problem long before the phenomenon drew the attention of researchers (Maslach, Schaufeli & Leiter, 2001).The emotional demands that are characteristic of many helping professions such as nursing, psychotherapy, medicine and social work can result in individuals experiencing distress and developing burnout over time (Acker, 2012;Chang, Eddins-Follensbee & Coverdale, 2012;Iglesias, Vallejo & Feuntes, 2010;Puig, Yoon, Callueng, An & Lee, 2014).Local research by Rothmann and Malan (2003) found high levels of burnout and cynicism, and low to average levels of a sense of personal efficacy, in a sample of 107 social workers.International studies indicate that, even in their undergraduate studies, the demands of social work training can result in burnout among social work students (Harr & Moore, 2011;Ngai & Cheung, 2009;Piatkowska, 2014), especially in individuals with limited resources (Alarcon, Edwards & Menke, 2011).The implications of burnout among students may indeed influence attrition rates and academic success (Alarcon et al., 2011).The higher risk of burnout among social work students and the negative effects of burnout on students' academic performance should be a cause for concern for teaching staff involved in the training of social work students.
Burnout can impact on every aspect of an individual's life, with adverse effects on interpersonal and familial relationships as well as a generally negative attitude towards life (Papathanasiou, 2015).Moreover, costs related to personal fulfilment and organisational productivity, such as depression and absenteeism, are high in individuals who experience burnout (Leiter, Hakanen, Ahola, Toppinen-Tanner, Koskinen & Vaananen, 2012).Psychological reactions to burnout often include hopelessness, anxiety, anger, frustration and tension (Kozak, Kersten, Schillmoller & Nienhaus, 2013;Piatkowska, 2014).Additionally, the development and subsequent continuation of burnout has serious health implications that could result in the development of psychosomatic illnesses (Kozak et al., 2013;Papathanasiou, 2015;Piatkowska, 2014).Prolonged emotional stress related to burnout is sometimes associated with the use of psychotropic medication such as antidepressants (Leiter et al., 2012).Burnout is frequently related to the personality trait of neuroticism, which exists in individuals who have an increased vulnerability to stress.In contrast, the trait of extraversion is related to the opposite of burnout, namely engagement, and acts as a buffer against stress, allowing a more adaptive response to environmental stressors (Langelaan, Bakker, van Doornen & Schaufeli, 2006).
An online search for published work on burnout and engagement revealed that much has been published on these phenomena, especially in the medical and helping professions.A substantial body of literature on burnout among students was identified; however, publications focus predominantly on students in the health sciences.International studies on burnout in undergraduate social work students in the USA and Hong Kong found that these students reported higher levels of burnout, and that emotional exhaustion was positively linked to idealism among social work students (Harr & Moore, 2011;Ngai & Cheung, 2011;Piatkowska, 2014).
Anecdotal reports by undergraduate social work students in the authors' work environment seemed to emphasise a central theme, namely the high demands of simultaneously managing the requirements of academic and practice training.Moreover, a literature search on local electronic platforms, such as Sabinet, showed that local research on burnout in student populations relate to, among other things, students' general wellbeing (Mokgele & Rothmann, 2014); burnout and engagement in student leaders (Pienaar & Sieberhagen, 2005); and burnout and mood state among student rugby union players (Grobbelaar, Malan, Steyn & Ellis, 2011).Although international studies on burnout among social work students could be found, Piatkowska (2014) notes that such studies are scarce.Furthermore, no local studies could be found on burnout in social work students.To fill the knowledge gap, this article describes the levels of burnout and engagement among fourth-year social work students attending a South African university.In addition, the article describes correlations between levels of burnout and engagement with respect to demographic variables, in particular population group, place of origin and household income status.
Research has identified various predictors of burnout among students, including pressure related to fulltime courses, expectations from lecturing staff in different subject fields, and anxieties regarding academic performance (Tomaschewski-Barlem, Lunardi, Lunardi, Barlem, Da Silveira & Vidal, 2014).In addition, group characteristics such as competition, a lack of time for leisure activities, and professional concerns about the future have been found to precede burnout among students (Bonafé, Maroco & Campos, 2014).Coffey, Samuel, Collins and Morris (2014) and Collins, Coffey and Morris (2010) indicate that professional education and training in social work involve academic as well as practice learning components, which could bring about higher levels of stress than generated in other more generic academic programmes.Seen broadly, emotional exhaustion is present in social work students and driven by, amongst other things, over-commitment and idealism, while protective factors result in high levels of compassion satisfaction.
Recent trends in research on burnout reflected an expansion of the concept's traditional meaning and scope of influence.There has been a shift in research towards the opposite of burnout, namely work engagement.Instead of focusing exclusively on the negative pole (burnout), researchers extended their focus of interest to the positive pole of workers' wellbeing (engagement).Research on burnout has also expanded to all parts of the globe, including Africa (Schaufeli, Martinez, Marques-Pinto, Salanova & Bakker, 2002).Furthermore, research has shifted focus from the predictors and outcomes of burnout to the numerous variables correlated to burnout.Lastly, research supports the conceptualisation of burnout and engagement (as well as their respective dimensions) as polar opposites (Demerouti, Mostert & Bakker, 2010;González-Romá, Schaufeli, Bakker & Lloret, 2006).

KEY CONCEPTS
While burnout is characterised by low energy levels, fatigue, cynicism and diminished accomplishment, engagement essentially relates to mental resilience, enthusiasm and immersion in one's work (Alarcon & Edwards, 2010;Maslach et al., 2001).

Burnout
Until recently, burnout was thought to occur in employed individuals only; hence the traditional description of 'job' burnout.However, burnout is present among numerous student cohorts worldwide (Mostert, Pienaar, Gauche & Jackson, 2007).Although there is no standard definition of burnout (Maslach et al., 2001;Piatkowska, 2014), the concept generally refers to a reaction to occupational stress (González-Romá et al., 2006).As a psychological syndrome, burnout has been explained as the response to prolonged emotional and interpersonal stressors.Burnout consists of three dimensions, namely emotional exhaustion (fatigue), cynicism (depersonalisation) and inefficacy (reduced personal accomplishment) (Maslach et al., 2001), as discussed below.
 Emotional exhaustion commonly manifests as low energy and fatigue.It refers to feelings of depleted physical and emotional resources, which prompts the individual to distance him/herself emotionally and cognitively from his/her work.Emotional exhaustion is considered to be the most widely reported and analysed dimension of burnout.Some researchers postulate that emotional exhaustion is the only central feature of burnout, as it is the most obvious manifestation of the condition (Maslach et al., 2001).
 Depersonalisation is the experience of cynicism and psychological distancing from interpersonal relationships (Maslach et al., 2001;Piatkowska, 2014).More specifically related to the helping professions, it refers to an emotional detachment between the helper and the recipients of services, where the helper remains aloof from the experiences of the client.Through cognitive distancing, a cynical or indifferent attitude develops, considering others as impersonal objects of one's work (Maslach et al., 2001).Some researchers therefore argue that it is not only emotional exhaustion, but also depersonalisation, that in fact represent the core components of burnout (Purvanova & Muros, 2010).
 Inefficacy refers to reduced personal or professional effectiveness and the experience of diminished accomplishment (Piatkowska, 2014).This third dimension has a complex relationship with the other two components of burnoutit is recognised that chronic exhaustion and cynicism tend to erode a personal sense of accomplishment and effectiveness, leading to increased levels of inefficacy (Maslach et al., 2001).
In the present study burnout among social work students will consequently be considered as elevated levels of emotional exhaustion from study demands, a detached and cynical attitude to one's work, and decreased levels of personal accomplishment and competence as a social work student (Schaufeli et al., 2002).

Engagement
Although the concept of engagement in work contexts has many definitions, it generally refers to a positive connection with one's work (Alarcon & Edwards, 2010).Engagement represents the opposite to burnout and includes strengthening factors and motivational processes (Alarcon & Edwards, 2010).
Unlike the victims of burnout, engaged persons feel energised and positively connected to their work.They do not experience work as stressful and daunting, but perceive it as challenging and fulfilling, while also strongly identifying with their work-related tasks (Bakker, Schaufeli, Leiter & Taris, 2008).Like burnout, engagement also has three dimensions, namely vigour, dedication and absorption, as discussed below.
 Vigour relates to high levels of mental resilience and the ability to persist with work despite experiencing some difficulties in the process.Vigour is characterised by an abundance of energy and investment in the accomplishment of tasks.It includes the mobilisation and building of resources (Schaufelli et al., 2002).
 Absorption refers to focused attention, a clear mind, enjoyment, a loss of sense of time, and immersion in one's work (Schaufeli et al., 2002;Mostert et al., 2007).Vigour and dedication are the direct opposites of exhaustion and cynicism, which are the core symptoms of burnout.
Engaged persons display several cross-cutting characteristics related to their work, which extend to their personal lives.These characteristics include initiative and having direction in life; generating one's own positive feedback as encouragement; extending engagement to elements outside the work environment; aligning personal values and norms to those of work; experiencing fatigue as positive fatigue; addressing burnout when it occurs; and acknowledging the need to occasionally do something other than work, thus not becoming enslaved to work (Schaufeli, Taris, Le Blanc, Peeters, Bakker & De Jonge, 2001).Some of these characteristics should not be confused with those of people who are considered workaholics.Workaholics spend much time on work activities, since they choose to do so and are reluctant to detach themselves from work, even when they are not at work.Engaged persons work hard (vigour), are involved (dedicated) and enjoy being immersed in their work (absorption).
Although they might appear similar to workaholics, engaged persons lack the typical compulsive behaviour with which workaholics present (Bakker et al., 2008).
In their review of the literature Bakker et al. (2008) identified the following factors that contribute to engagement:  Resources, including social support from peers and supervisors;  Feedback on performance, coaching, and involvement in decision-making;  Stimulation of personal development and learning opportunities;  Active and functional involvement in achieving collective goals;  Environments where innovation and information are valued;  Effective coping mechanisms to remove or rearrange stressors;  Rewards for and recognition of achievements.
Engaged persons experience numerous work-related benefits.Thoughts of resigning or dropping out are minimised, which contributes to positive emotions and overall wellbeing.Work satisfaction is facilitated by using available resources optimally and creating one's own (often personal) resources.Moreover, engaged persons transfer their engagement to others.Personal achievement, high selfefficacy and positive self-feedback ensure dedication to the task, which for organisations means decreased turnover intentions (Alarcon & Edwards, 2010;Bakker et al., 2008).

BURNOUT IN STUDENTS
The evidence available on burnout in students, although substantial, largely reports on the experiences of students in general and in particular those in the health care professions.It is worthwhile to observe trends from these study populations before reflecting on the more limited evidence on burnout in social work students.Dyrbye, Power, Massie, Eacker, Harper, Thomas, Szydlo, Sloan and Shanafelt ( 2010) explored resilience and protective factors against burnout in a sample of 792 medical students attending five institutions across the USA.They found that resilient students were less likely to suffer depression, had an increased quality of life, experienced fewer stressful life events, and had a positive perception of their learning climate and increased levels of social support compared to students vulnerable to burnout.In a similar cohort of participants, Enoch, Chibnall, Schindler and Slavin (2013) found that among 145 fourth-year medical students in the USA, 29% experienced emotional exhaustion, 18% experienced depersonalisation and 21% experienced diminished personal accomplishment.In total, 37% of the students met the criteria for burnout.Fang, Young, Golshan, Moutier and Zisook (2012) found that female and Hispanic premedical students presented higher levels of emotional exhaustion than their male counterparts and those from other population groups in American tertiary training institutions.Morgan and De Bruyn (2010) investigated the relationship between personality and burnout among 297 students from different faculties attending a South African university.The results showed that neuroticism was associated with high levels of emotional exhaustion and cynicism; extroversion and openness to experience were related to professional efficacy; and emotionally stable, outgoing, hardworking individuals with good interpersonal relationships exhibited lower levels of burnout.
As mentioned, evidence has shown high levels of emotional exhaustion and other symptoms related to burnout among various student populations, such as those studying in dentistry (Brazil, Sudan), medicine (United States, Spain) and nursing (Australia, Brazil) (Bonafé et al., 2014;Enoch et al., 2013;Galán, Sanmartín, Polo & Giner, 2011;Ghali & Awooda, 2013;Rella, Winwood & Lushington, 2008;Tomaschewski-Barlem et al., 2014).Among young adult populations in the United Kingdom, Taiwan and Finland, students have been found to experience high levels of burnout, a finding that could be expected given the pressures experienced during the completion of studies (Collins et al., 2010;Huang & Lin, 2010;Parker & Salmena-Aro, 2011).Professionals in the helping professions such as psychotherapy (Puig et al., 2014) and nursing (Iglesias et al., 2010) seem more likely to experience burnout.Importantly, studies from abroad identified high levels of burnout in social work students (Harr & Moore, 2011;Ngai & Cheung, 2009;Piatkowska, 2014).Therefore, the very nature of the helping professions might result in service providers experiencing burnout.
The discussion thus far demonstrates that burnout and engagement are as relevant to student populations as to the general work population.The conservation of resources model has been selected to frame the study, as it relates burnout and engagement to the availability of resources and provides a theoretical model that assists in explaining burnout and engagement in students.

THEORETICAL MODEL
The conservation of resources (COR) model effectively explains an important link between prolonged stress and the development of burnout.The COR model indicates that four types of resources are imperative to one's survival and wellbeing, and predict strain and/or optimal functioning in the individual.The four types of resources are: objects (e.g. a house), conditions (e.g.social support), personal characteristics (e.g.conscientiousness) and energies (e.g.money).Burnout will most likely occur as an outcome of stress resulting from a perceived or actual loss of key resources, or a failure to gain resources in comparison to demands that are placed on an individual (Alarcon et al., 2011;Gorgievski & Hobfoll, 2008;Hobfoll 1989in Wright & Hobfoll, 2004).
In the COR model burnout and engagement are described in terms of resource loss and gain.Stress and optimal functioning are likened to resources and demands respectively, hence people with high demands and low resources are more likely to experience burnout (Alarcon & Edwards, 2010).As such, the prolonged exhaustion and lack of interest characteristic of burnout is symptomatic of a prolonged period of resource loss (González-Romá et al., 2006;Maslach et al., 2001).When resources are diminished, an individual may develop maladaptive coping strategies, further weakening resources and leading to a spiral of resource loss, resulting in burnout (Alarcon et al., 2011).In contrast, an engaged individual creates more resources in the environment and may invest in resources such as physical fitness and social support, leading to an increased positive emotional experience.The availability of sufficient resources results in productive coping (Alarcon et al., 2011).Therefore, individuals who are exposed to high demands but have sufficient resources will be likely to experience engagement, resulting in further effective coping with stressors (Alarcon & Edwards, 2010).
Ultimately, the level of resources that are available to an individual will determine the level of adaptive or maladaptive coping with stressors (Alarcon et al., 2011).In South Africa the infrastructural resources that students can tap into are severely constrained, which could exacerbate the incidence of burnout.In addition, students from disadvantaged backgrounds may be confronted disproportionately with poverty and limited access to facilities and services (Ross, 2010).
The arguments thus far confirm the reality of burnout in the helping professions.Burnout is also recognised in student populations and perhaps even more so in disciplines such as social work that entail simultaneous exposure to theory and practice training.International studies demonstrate that burnout indeed exists among undergraduate social work students.Although some local evidence is available on burnout in general student populations (cf.Morgan & De Bruin, 2010;Mostert et al., 2007), this article sets out to investigate burnout and engagement in fourth-year social work students.In addition, levels of burnout and engagement are described in relation to the background characteristics of students.Therefore, the research question that guided the investigation was: What are the levels of burnout and engagement in fourth-year social work students?

RESEARCH METHODS
This study followed a quantitative approach with a descriptive research purpose (Neuman, 2011).The research design entailed a survey among fourth-year social work students attending a South African university.By means of availability sampling, 43 of a potential 58 students participated in the survey (sample coverage 74.1%).
The research instrument consisted of four parts:  Firstly, a biographic component was developed to obtain data regarding respondents' gender, age, population group, type of area they mainly grew up in, and the economic status of their households;  Secondly, the Maslach Burnout Inventory -Student Survey (MBI-SS) was used to measure burnout in students.The inventory represents the three subscales of burnout, namely exhaustion, cynicism and professional efficacy.Items were scored on a seven-point frequency rating scale ranging from 1 (never) to 7 (always);  Thirdly, the Utrecht Work Engagement Scale -Student Survey (UWES-S) was used to measure the levels of engagement among students.The scale represents the three subscales of engagement, namely vigour, dedication and absorption.Items were scored on a seven-point scale similar to the MBI-SS, which allows for a total score for each category to be calculated;  Both the MBI-SS and UWES-S (publicly available) have been validated internationally (Schaufeli et al., 2002) and locally (Mostert et al., 2007;Pienaar & Sieberhagen, 2005).In the present survey the Cronbach's alpha coefficients of each subscale were determined to test the internal consistency of the questionnaire.The alpha coefficient for the categories of burnout was 0.919 for exhaustion, 0.870 for cynicism and 0.803 for professional efficacy.The alpha coefficient for the categories of engagement was 0.769 for vigour, 0.879 for dedication and 0.705 for absorption.Therefore, the reliability of both the MBI-SS and the UWES-S was at an acceptable to excellent level;  Fourthly, the work of Ngai and Cheung (2009) drew the researchers' attention to factors which contribute to or reduce burnout in social work students.Toward this end, two open-ended questions were included for students to indicate the three most important factors which positively and negatively affected their studies.
The survey was conducted toward the end of the first semester of the academic year in order to allow sufficient time for students to be immersed in their academic work.Arrangements were made with lecturing staff well in advance to cause minimal disruption during lecturing activities.The questionnaire was handed out in class, after which students were asked to complete the instrument individually.The questionnaire took about fifteen minutes to complete.
Data were coded and captured in the Statistical Package for the Social Sciences version 23 (IBM, 2016).The MBI-SS and UWES-S contain scale-type questions which allow for the calculation of a total score for each respondent.The dataset was subjected to the Kolmogorov-Smirnov Z test to determine normality.Across the 29 sub-scales of the BMI and UWES-S it was found that p<0.05, thus indicating that data distributions were not normal (Field, 2009).Subsequently, significant differences were determined using non-parametric tests, in particular the Mann-Whitney U and Kruskall-Wallis H tests.The results of the Mann-Whitney U tests are accompanied by effect sizes, where -0.1 indicates a small, -0.3 represents a medium and -0.5 denotes a large effect size (Field, 2009).Results of the Kruskall-Wallis H test include the chi-square value (X 2 ).Where appropriate, the standard deviation (SD), median (M), median rank (MR) and total score (S) of responses are provided.
Respondents participated voluntarily in the survey after providing their informed consent (Maxfield & Babbie, 2009).The informed consent letter indicated the purpose of the study, contact particulars of the researchers, and that the completed questionnaires would be stored for archiving purposes.The standard ethical considerations of anonymity, confidentiality and no harm were honoured throughout the study (Leedy & Omrod, 2010).The questionnaire did not gather any personal information that could identify individual respondents.Authorisation to conduct the survey was obtained from the Ethics Committee of the Faculty of Humanities and university management.

Background characteristics of respondents
The mean age of respondents was 22.7 years (SD=1.82years).The vast majority of respondents were female (n=41; 95.3%), with only two males (4.7%) participating in the survey.Approximately three quarters of respondents were Black (n=33; 76.7%), followed by nine White respondents (20.9%) and one Coloured (2.3%) respondent.While only one respondent (2.3%) indicated having grown up in a high-income household, 69.8% (n=30) indicated that they grew up in middle-income and 27.9% (n=12) in low-income households.The greater proportion of respondents originated from rural areas (n=18; 41.9%), followed by one in five from township (n=9; 20.9%) and one in twelve (n=8; 18.6%) each from small town and city areas respectively.Nearly a third of respondents (n=14; 32.6%) lived with extended family, followed by one in four (n=10; 23.3%) who lived with their parents, 16.3% (n=7) who lived on their own, 14.0% (n=6) living in a student commune and only one (2.3%)living in a university residence.Five respondents (11.6%) indicated that they had "other" living arrangements.
It is important to take note of the characteristics of the sample as it has implications for the interpretation of the results.Scrutiny of the dataset revealed a significant association between household economic status and the area in which respondents mainly grew up.While the number of respondents from middle-income households was equally spread across the types of areas that respondents grew up in (i.e.city, small town, township and rural environment), respondents from lowincome households were significantly more likely to originate from rural areas (X 2 =11.026; p=0.004).Furthermore, a significant association was found between Black and White respondents (excluding the one Coloured respondent) and the area they mainly grew up in.All White respondents (n=9) were from city and small town areas, while half of Black respondents (n=18; 54.5%) were from rural areas and a quarter (n=9; 27.3%) were from townships (Z=-4.216; p<0.001; r=-0.64).Furthermore, White respondents lived exclusively with their parents or family members, while Black respondents' living arrangements were varied, including parents and family members (n=15; 45.5%), a university residence (n=1; 3.0%), student commune (n=6; 18.2%), living on their own (n=6; 18.2%), or other living arrangements (n=5; 15.2%) (Z=-2.147; p=0.035; r=-0.32).

Descriptive results for burnout and engagement
In terms of burnout, the key indicators of respondents' burnout were exhaustion, cynicism and professional efficacy (Table 1).Scale: 1 = never and 7 = always The respondents' experiences of exhaustion appeared relatively high, given that the mean for all results in this dimension was above 3.5, in other words tending more towards 'always' than 'never'.
Respondents generally presented with low levels of cynicism and high levels of professional efficacy.
In terms of engagement, the key indicators were vigour, dedication and absorption (Table 2).Scale: 1 = never and 7 = always Respondents presented average scores in the vigour dimension of engagement and high scores in the dedication dimension, with scores ranging from 5.49 and 5.88 in favour of 'always'.All scores for elements in the absorption dimension were slightly above the 3.5 average.Table 3 represents the mean and total scores for the dimensions of burnout and engagement (the range of scores depends on the number of items that make up each burnout or engagement category).It generally confirms the results from the two tables above, where respondents experience relatively high levels of exhaustion, low levels of cynicism and high levels of dedication to studying social work.The background characteristics of the sample reflect the intersection between low income, growing up in a rural area, and historically marginalised communities.Despite high levels of exhaustion and average levels of vigour, respondents presented above average levels of professional efficacy, dedication and absorption.

Bivariate results for burnout and engagement
Analysis showed that Black respondents experienced higher levels of exhaustion compared to White respondents.The effect size (r) of these differences was medium to high (Table 4).* The one respondent from the Coloured population group was omitted from the calculations because the Mann-Whitney U test entails two independent groups.
Respondents from middle-income households experienced lower levels of exhaustion compared to those from low-income households.The effect size (r) of these differences was medium to mediumhigh ( Table 5).* The respondent from the high-income background was omitted from the calculations because the Mann-Whitney U test entails two independent groups.
Two elements of burnout showed a significant difference in terms of the area in which respondents mainly grew up.
 Respondents who grew up in city (M=3.50) and small town (M=4.38)areas were less likely to feel emotionally drained compared to those who grew up in township (M=5.33) and rural (M=5.72)areas (X 2 =12.715; p=0.005).
 Respondents who grew up in township (M=4.33) and rural (M=5.28)areas showed higher levels of burnout compared to those who grew up in city (M=3.50) and small town (M=4.00)areas (X 2 =8.878; p=0.031).
In terms of professional efficacy, respondents from city (M=5.13) and small town (M=5.00)areas were more likely to contribute in class compared to those from township (M=4.22) and rural (M=3.50)areas (X 2 =10.688; p=0.014).
Exhaustion, which is a key indicator of burnout, showed significant differences across biographic variables in terms of total score (the total score for exhaustion can vary between 7 and 35, where a lower score indicates lower levels of burnout).
 Respondents from city (S=17.75)and small town (S=20.13)areas were less likely to suffer from exhaustion compared to those from township (S=24.22)and rural (S=25.89)areas (X 2 =9.598; p=0.002).

Qualitative results for burnout and engagement
The respondents indicated a variety of factors that contributed positively to their studies, including being dedicated to one's prospective career (14%); having support from family and friends (12%); receiving adequate support (11%) and supervision (8%) from lecturers; and working hard towards achieving one's dreams (7%).Factors that they perceived to affect their studies negatively included high workload (17%); time constraints (11%); financial problems (7%); time management (7%); and participation in social activities (4%).

DISCUSSION
Studies on burnout in students indicate that university students, in general, face challenges related to their academic work.In addition to the challenges faced by all university students, social work students experience demands relating to their exposure to stressors and vicarious trauma during their field practicum (Napoli & Bonifas, 2011;Ngai & Cheung, 2009).The reality of burnout among social work students has been acknowledged in a number of studies (cf.Harr & Moore, 2011;Piatkowska, 2014).
While the present study found above average levels of emotional exhaustion among a sample of social work students, it is noteworthy that the sample presented low levels of cynicism and high levels of professional efficacy.Based on the interrelatedness of the three dimensions of burnout, these results suggest that the high levels that were found in terms of positive attitude and inspiration experienced by the respondents might have a mediating effect on the negative influence of exhaustion.Nevertheless, emotional exhaustion is a central feature and the most obvious manifestation of burnout (Maslach et al., 2001) and it is of concern that many respondents felt emotionally drained, tired and burnt out as a result of their studies.As burnout leads to depleted physical and emotional resources, it could negatively affect students' emotional and cognitive capacities that are part of the characteristic professional competencies required by the social work profession (Adamson, Beddoe & Davys, 2012).
The findings suggest that a particular profile of social work students in the sample appears to be at higher risk of emotional exhaustion, namely Black students, those from lower-income households, and students from rural and township areas.It is important to note that the significant differences across population group, income status and area of origin were medium to strong in nature.Similar to the findings in a study by Ross (2010) on the profile of applicants for admission to the social work degree at a South African university, many respondents in the present survey came from a historically marginalised context associated with poverty, low incomes and a lack of resourcesa profile that seems to correspond with that of the majority of higher education students in South Africa (Dykes & Green, 2015).In addition, more than half of the Black respondents did not live with their parents or close family, thereby not benefitting from the support of members of the immediate family who are physically available on a daily basis.From the responses to the open-ended questions, support from family and friends was indeed indicated as an important factor in contributing positively to their studies.These realities may well contribute to Black social work students' vulnerability to emotional exhaustion and concomitant academic performance.The latter situation could potentially explain why the Black respondents and those from low income backgrounds were significantly less likely to contribute effectively in class and more likely to experience their studies as more stressful.
While all three domains of engagement indicated above average scores for respondents in this study, the domain of vigourwhich constitutes items on physical and mental energyshowed notably lower scores than dedication and absorption.Vigour as a dimension of engagement is the opposite of the dimension of emotional exhaustion in burnout, and the fairly average score of vigour might be mediated by the relatively high levels of emotional exhaustion among the respondents.
Of all the dimensions of burnout and engagement, dedication presented with the highest score among respondents in the present study.Combined with the results related to respondents' high levels of professional efficacy, the findings demonstrate very high levels of commitment to their training to become social workers.It is noteworthy that, in the open questions, the respondents identified dedication to the social work profession as a key positive contributor to their studies.These observations are in agreement with the findings of Harr and Moore (2011) that the social work students in their study were motivated and committed to the profession, and showed high levels of compassion satisfaction.It is also noteworthy that respondents from the present survey found meaning, purpose, inspiration and pride in their studies, and this could possibly explain their below average score on cynicism.However, it should be noted that Ngai and Cheung (2009) warn that positive sentiments such as idealism and over-commitment to the social work role could in fact lead to emotional exhaustion and burnout when there is a discrepancy between the idealistic expectations of the students and the observed challenges faced by social workers in practice.
The conservation of resources model provides a relevant framework to explain the interplay between the availability of resources (or lack thereof) and experiencing burnout or engagement in social work students.Student populations with limited resources in the form of objects (accommodation), conditions (family support) and energies (financial means) (Alarcon et al., 2011), combined with the stressors associated with the social work profession (Adamson et al., 2012), might well be at higher risk of burnout.It is, however, equally important to acknowledge that the historical disadvantages some students experience could exacerbate their risk of stress and burnout.As social and emotional competencies can enhance resilience to stress, social work educators can play a significant role in mobilising resources and serve as a resource themselves towards the promotion of students' wellbeing and prevention of burnout (Kinman & Grant, 2010).The importance of support by lecturers and supervisors was highlighted in the responses to the open questions, in which a relatively high number of responses indicated that support and supervision from lecturers (in combination, 20% of the responses) contribute positively to their studies.

RECOMMENDATIONS FOR SOCIAL WORK EDUCATION AND TRAINING
It is important for social work departments at training institutions to adopt a preventative (instead of reactive) model regarding burnout in students.A first step could be for departments to use validated instruments to determine the profiles of students who are at high risk of burnout during their theory and practice training.Based on context-specific profiling, i.e. the availability of support structures, and the mental health and financial situation of students, academic departments should develop intervention strategies that meet the unique needs of students who are either at risk of, or suffering from, burnout.These might include student support in the form of peer mentoring and motivational activities (Salanova, Schaufeli, Martinez & Breso, 2010.)Students should be made aware of such support initiatives.In addition, continued monitoring is needed of students who fit high-risk profiles and those who receive support services to track the effectiveness of interventions.Monitoring of high-risk students could take place by means of tracking the uptake of support services and following up on student progress.
Furthermore, social work departments could include a teaching module on enhancing the selfawareness of social work students, as well as related curricular activities focused on coping skills such as interpersonal and time management skills.These modules should already be introduced in the early stages of undergraduate social work training.
As the present research indicates, quality support from lecturers and supervisors is imperative to counteract potential burnout in students.In addition to the support provided by staff, social work departments can consider mentoring programmes where senior students can mentor junior-level students, either in a one-on-one or group format.To strengthen notions of engagement, it may benefit students if departments plan social activities in the academic calendar where students can participate in enjoyable activities with minimal academic work.Lastly, as the present findings indicate, personal meaning and hope (dedication) which drive the ethos of the social work profession should constantly be reaffirmed in educational activities in order to strengthen engagement and counteract burnout in students.

LIMITATIONS
One should be cautious about generalising the results of this study to other settings, given differences in the curricula and practical training across institutions that offer social work training in South Africa.The present study relied on a limited number of respondents, which further limits generalisation of the findings across settings.In addition, results may vary in terms of the time of measurement; despite the potential continuity of academic stress that students experience over the span of an academic year, it is likely that burnout peaks during times of assessment, for example, tests, exams and deadlines for the submission of assignments and practice reports.This study was conducted during the first semester of the academic year, but the increasing workload over the course of the year might have resulted in an increase in the level of stress among the respondents later in the academic year.

FUTURE RESEARCH
The present study demonstrates the value of research into the wellbeing of social work students and should be followed by measuring burnout and engagement across all year levels, including postgraduate training.In addition, longitudinal research is needed to determine trends in burnout and engagement in social work students, and whether fluctuations can be linked to particular curricular activities and study demands.Research of this nature will further be enhanced by linking burnout and engagement results to students' academic performance.Furthermore, burnout and engagement should be investigated in social work students attending different institutions of higher education, while comparative work is needed regarding social work students' experiences of burnout and engagement relative to students who study in other professional fields.