Do stress and coping influence resilience in social work students? A longitudinal and comparative study from India

Stress among social work students is an issue of concern and has adverse outcomes. This study used a longitudinal design to assess stress, resilience and coping in undergraduate social work students in India and a comparative cohort of non-social work undergraduates. Standardised instruments were administered at three time points: the first on course entry, then in the second year of study and finally on the verge of course completion at the end of three years. Moderate levels of stress and resilience were seen in the social work cohort, but no incremental progression in their manifestation was observed as students progressed through their degree. Stress, problem-focused and emotion-focused coping strategies emerged as significant predictors of resilience. The findings indicate a need to develop resilience in social work students to enable them deal more effectively with various stressors.

and non-academic factors such as high academic burden, financial strains, peer competition, the need to excel, homesickness, social problems and disturbed sleep patterns (Lund et al., 2010;Tosevski et al., 2010). Adverse psychological problems that can result from high stress include anxiety, depression and engagement in high-risk behaviour (Kumaraswamy, 2013;Saravanan and Wilks, 2014).
Social work students face a number of unique stressors as compared to the general college student population. Field placement, which is an integral component of most social work degrees, has been acknowledged as an important stressor (Maidment, 2003). A study of UK-based social work students relating to perceived academic demands reported moderate or high stress with regard to attendance, academic essays, the course structure, financial responsibilities, and issues on placement relating to practice teachers and service users (Coffey et al., 2010). Other sources of stress identified by social work students were in relation to academic stress, financial stress and family problems (Bulanda et al., 2018). Maintaining a work-life balance is also a key issue for many social work students as they cope with the demands of work, school and personal responsibilities, resulting in high levels of stress (Radey and Figley, 2007;Ying, 2011).
Besides stressors experienced by their counterparts in the West, social work students in India also face stressors in the form of subject-oriented written exams that are the predominant forms of assessment used in higher education settings. Placement stressors often stem from a lack of role clarity on the part of accommodating agencies, a dearth of qualified agency supervisors, and participation in field activities beyond regular work hours. Not many studies have specifically addressed the issue of stress in social work students in India, although a comparative study of undergraduate students in India and the United Kingdom revealed higher demands experienced by the UK students than their Indian counterparts (Coffey et al., 2014). Another study from India indicated higher levels of stress in newly admitted Bachelor of Social Work (BSW) students and those in the final year of their degree (Stanley and Mettilda, 2016).
The issues of coping and resilience gain significance in the context of understanding how individuals deal with stress. Lazarus and Folkman (2013) consider coping to be the constantly changing cognitive and behavioural efforts made by individuals to manage specific external and/or internal demands that are appraised as being taxing or exceeding the resources of the person. Appraisal of stressors and the coping response used can predict perceived levels of stress and play a crucial role in shaping the meaning and impact of stressful life events (Cherkil et al., 2013;Crary, 2013). These strategies have been broadly classified as being emotion-focused and problem-focused (Carver and Scheier, 1994). While Carver and Scheier identified 14 coping strategies for their two-way classification of coping, Wong et al. (2006) delineated dysfunctional coping within this classification. Problem-focused coping involves engaging in specific behaviours to alleviate, minimise or eliminate the stressor altogether (Schnider et al., 2007). Problemfocused coping and behaviours that increase acceptance, seeking social support and positive reappraisal of stressors have been associated with lower levels of perceived stress (Reeve et al., 2013). Emotion-focused coping refers to the strategies used to deal with the emotions experienced as a consequence of a stressful situation (Lazarus and Folkman, 1984). Maladaptive or dysfunctional strategies, on the other hand, such as blaming, substance use, denial, self-distraction and behavioural disengagement, are associated with higher levels of perceived stress (Cherkil et al., 2013). Studies on coping among university students have found that problem-focused coping is positively associated with academic achievement (Kariv and Heiman, 2005) and the use of emotion-focused strategies to heighten anxiety levels (Liverant et al., 2004). Gender differences in the use of coping strategies have been reported by Vungkhanching et al. (2017), with female students using both emotion-focused and problem-focused strategies more often as compared to male students. Furthermore, they found that avoidant coping was a significant predictor of psychological distress.
Resilience has been conceptualised in various ways, and Luthar et al. (2000) consider it a dynamic process encompassing positive adaptation within the context of significant adversity. The ability to tolerate stressful conditions and gain strength from adversity is another way that resilience has been conceptualised (Fletcher and Sarkar, 2013;Secades et al., 2016). The literature on resilience indicates that some view it as a personality trait and others as a contextual or circumstantial disposition . Characteristics such as hardiness, sense of control, emotional intelligence, optimism, hope, self-efficacy, persistence and an ability to appraise 'problems' as challenges have been attributed to resilient people (Collins, 2015;Grant and Kinman, 2014).
There is burgeoning literature on resilience within social work. Given the stressful nature of social work practice, resilience is seen as an important attribute for social workers that may enable them to cope better with adverse circumstances. Resilience plays a key role in exerting a moderating influence between stress and burnout experienced by social workers in practice . However, very few studies have investigated resilience in social work students. Wilks (2008) reports high levels of resilience in a sample of social work students from the United States. The influence and predictive ability of personality traits on the resilience of students and social workers has been established in a study from Spain (De las Olas Palma-García and Hombrados-Mendieta, 2017). In the Indian context, hardly any studies have examined resilience in social work students. A cross-sectional study of social work students from India reports higher levels of resilience in students in the final year of their degree when compared to those in their first year (Stanley and Mettilda, 2016).
The relationship between the three variables of interest in this study can be conceptualised in terms of the definition of stress that considers it to be a physiological and psychological response to perceived threat (Wright, 2014), while resilience provides protection from its impact (Fletcher and Sarkar, 2013) and coping mechanisms enable dealing with or managing the stressors faced.
Most studies reviewed in this section have been cross-sectional in nature (except De las Olas Palma-García and Hombrados-Mendieta, 2017). We used a longitudinal design to identify whether our variables of interest (stress, resilience and coping) change in any way as students progress through their degree, owing to different academic and life experiences. Furthermore, previous studies involving social work students have not involved a reference group for comparison. This precluded drawing inferences about whether these variables manifested differently in social work students when compared to similar students from other disciplines. It has been observed that professional education and training for a career in the 'helping' professions such as teaching, nursing or social work combine academic and professional requirements that can induce more stress in students when compared to other academic programmes (Dziegielewski et al., 2004). We were thus interested in comparing social work students with those from a non-professional degree programme. Against this background, this study seeks to answer the following questions: 1. What is the extent and nature of stress, resilience and coping in social work students when compared to students of another academic discipline? 2. Do these attributes change over time in social work students? 3. What is the nature of the relationship (if any) among these key variables? 4. Which of these variables predict the manifestation of resilience in social work students?

Research design
This study collected non-experimental longitudinal data from two cohorts of undergraduate students over a 3-year period, as they progressed from the first year of their degree to their final year.
Data were collected at three points: on admission to the course in their first year (T1), then in their second year (T2), and finally at the point of graduation, in the final year of their degree (T3). Students belonged to the social work degree (BSW) programme and to a reference (control) group of students doing their Bachelor of Arts (BA) degree in Tamil (vernacular language of South India). Both groups were followed concurrently from T1 to T3 and data were collected by administering the same measures to both at all three points. Survey methodology was used for data collection and the analysis done is comparative and predominantly correlational in nature.

Measures
A socio-demographic data sheet was prepared to collect background information.
Seven items from the stress sub-scale of the Depression, Anxiety and Stress Scale (DASS 21) by Lovibond and Lovibond (1995) were used to assess stress levels. Respondents were required to rate each item on a scale from 0 to 3 based on whether each statement applied to them over the previous week or not. Higher scores reflect higher levels of stress. The Cronbach's alpha in this study was calculated as 0.71, which indicates an 'acceptable' level of scale reliability (George and Mallery, 2003).
The Connor-Davidson Resilience Scale (2003) (CD-RISC) is a popular instrument used to assess resilience and was administered in this study. It measures five factors of resilience, namely 'personal competence, high standards, and tenacity' (factor 1); 'trust in one's instincts, tolerance of negative affect, and strengthening effects of stress' (factor 2); 'positive acceptance of change and secure relationships' (factor 3); 'control' (factor 4); and 'spiritual influences' (factor 5) (Connor and Davidson, 2003). It comprises 25 items, each rated on a 5-point scale (0-4), with higher scores indicating greater resilience. The Cronbach's alpha of the CD-RISC in this study was calculated as 0.90, which indicates 'excellent' scale reliability (George and Mallery, 2003).
The Brief Cope Scale (Carver, 1997), which has 26 items, was used to identify predominant coping styles used by the respondents. The instrument consists of three composite subscales measuring emotion-focused, problem-focused and dysfunctional coping styles that assess 14 coping styles such as self-distraction, venting, active coping, positive reframing, denial, planning, selfblame, use of emotional support, humour, use of instrumental support, acceptance, behavioural disengagement and religion. The Cronbach's alpha of this scale in this study was calculated as 0.80, which indicates 'good' scale reliability (George and Mallery, 2003).

Data collection
Thirty-four BSW students consented to be a part of the study and were administered the instruments of data collection at T1. An equal number of women students who were in the BA (Tamil) degree constituted the reference group and were used for comparative analysis. At T2, 33 students remained in the BSW cohort and 34 continued in the reference group. At T3, there was one drop out in both the groups and data were collected from 32 study group and 33 reference group respondents. We finally had 99 data sets for study group respondents and 101 in the reference group, and these data have been used for analysis.

Setting for the study
Cauvery College for Women is an educational institution exclusively for women students in Tiruchirappalli, the southernmost state of Tamil Nadu in India. The college is affiliated with the Bharathidasan University and is an Arts and Science college. It offers 14 undergraduate 3-year degree programmes and nine 2-year postgraduate courses including social work at both levels.
The college caters to about 4000 students at all levels. The undergraduate social work programme confers a BSW degree on students and is of 3-year duration.

Ethical issues
The protocol of the study received clearance from the research ethics panel of the college. Informed consent was sought from the respondents and participation was voluntary. Participants were told that they could drop out of the study at any point and the questionnaires were anonymised. No personal identification data were sought from the respondents and no follow-up contact was made.

Respondents' profile
Study group. Social work students ranged in age from 16 to 23 years with a mean of 17.88 years. They were predominantly from Hindu (94.1%) nuclear (85.3%) families, residing in the urban areas of Tiruchirappalli city (76.5%). The total monthly family income ranged from Rs.1000 per month to Rs.79,000, with a mean income of Rs.15,876.47 (approx. US$225). The majority of these respondents were the eldest child in their family (55.9%) and had one sibling (50%). The medium of instruction at school for the majority had been in Tamil (70.6%), and it is possible that transition to their current degree course taught in English would have posed difficulties for these respondents. A high majority of them were day scholars, residing with their families (73.5%) and the remainder lived in student hostels. When asked about the reasons for choosing a social work degree, they stated self-interest (11.8%), to serve society (20.6%), suggested by others they knew (38.2%) and their perception that it was a useful course (11.8%). When asked how they felt about the course, responses indicated they felt happy (20.6%), and found the course to be interesting (29.4%). When asked about their career aspirations on course completion, the majority wished to take up a job relating to social work (64.7%); other options included studying law (8.8%) and the human resources sector (14.7%).
Reference group. The age of the BA (Tamil) students ranged from 17 to 21 years with a mean of 18.12 years. They were predominantly from Hindu (85.3%) nuclear (84.8%) families, residing in the urban areas of Tiruchirappalli (58.8%) and the remaining from rural areas. The total monthly family income ranged from Rs.1000 per month to 75,000, with a mean income of Rs.10,338.24 (approx. US$145). Most of these respondents were the eldest in the family (41.2%) and had one or two siblings (52.9%). The medium of instruction at school for the majority had been in Tamil (97.1%) and it is possible that transition to their current degree course in Tamil would have been relatively easy for these respondents. A high majority of them were day scholars, residing with their family (76.5%), and the remainder lived in student hostels. When asked about reasons for choosing to do a degree in Tamil, they stated that it was due to their interest in it (73.5%). When asked how they felt about the course, responses indicated they felt happy (76.5%) and found the course to be useful (23.5%).
The two groups were comparable across several domains. Student's t-tests reveal that in terms of age (t = 0.91; p > .05) and family income (t = 1.26; p > .05) there were no significant statistical differences between both groups. Both respondent groups hailed predominantly from Hindu nuclear families residing in urban areas. They were all women students from the same college, and this ensured further similarities in terms of key academic experiences. Furthermore, the majority in both groups were day scholars who were living with their families of origin.

Distribution of respondents by key study variables
In order to generate a profile of the respondents and categorise them according to the extent to which they manifested the key variables of the study, we used quartile deviation to categorise them into 'low', 'moderate' and 'high' levels. The scores obtained for each variable by both groups were pooled together and the cut-off for the three levels identified by the 25th, 50th and 75th percentile score. The three levels for each group were then classified based on these cut-off scores for each variable. This respondent profile is depicted in Table 1.
The majority of respondents in both groups were classified as experiencing moderate levels.
In the first analysis, we used a t-test to compare the mean scores for the key variables for both groups of respondents for all years of study pooled together. Results (Table 2) revealed significant differences between both groups on all the key variables except for the use of problem-focused and dysfunctional coping strategies and the F5 component of resilience. Mean score comparison reveals lower stress scores and higher resilience scores for the study group respondents. In terms  of coping strategies, study group respondents obtained higher mean scores for both problemfocused and emotion-focused scores, and lower scores for dysfunctional coping, than the reference group. Figure 1 portrays the mean scores of respondents of both groups at various points of time corresponding to their year of study. Paired t-tests were then used to assess any significant change in the key variables for the study group respondents at T3 compared to T1 (Table 3). Results show a statistically significant difference only for problem-focused coping, and the mean difference (2.38) indicates a decline at T3. Similar paired t-tests conducted for the reference group respondents did not elicit any statistically significant change at T3 for any of the key variables.
We then compared mean scores across the 3 years (T1, T2 and T3) for both respondent groups using one-way analysis of variance (ANOVA). The F values tabulated in Table 3 showed statistically significant variance between the six groups on all variables except for dysfunctional coping. Mean scores reveal that stress scores were highest for social work students in year 1 and for Tamil students in year 3. Resilience was highest in year 2 social work students and for Tamil students in their year 1. In terms of coping, emotion-focused coping scores were highest for both student groups in year 2, problem-focused coping scores for social work students in year 1 and Tamil students in year 2, whereas dysfunctional coping was higher in year 1 social work students and for Tamil students in year 2.

Correlations among key variables
Significant correlations were obtained between the total resilience score and the three coping strategies at both T1 and T3. While stress scores correlated positively with emotion-focused coping at T3, it did not at T1. Stress scores did not correlate significantly with problem-focused and dysfunctional coping at either of the two time points. As expected, the five sub-dimensions of resilience correlated significantly among themselves and with the overall resilience score ( Table 4).

Predictors of resilience
SPSS Amos was used for regression analysis by treating the total resilience score as the dependent variable (DV) and the stress and three coping styles as independents (IV). The model generated along with standardised estimates is depicted in Figure 2. The chi-square value for the model was not significant and indicated model saturation and adequate data fit.
The squared multiple correlation value (R 2 ) showed that together the IVs accounted for 40% variance in the total resilience score. The arrows leading from the four IVs to the DV show the standardised regression weights (β values) indicating their direct effects on the DV. It was seen that along with stress, both emotion-focused and problem-focused coping strategies were statistically significant. Dysfunctional coping, on the other hand, while influencing the DV did not exert a statistically significant influence. Stress levels negatively influence the manifestation of resilience. The curved arrows in the model show the correlation among the IVs and the values indicate statistically significant relationships among them.

Discussion
Our respondent profile shows that the majority of social work students experienced 'moderate' stress levels. This is in agreement with studies done with social work undergraduate students reported from Australia (Moran and Hughes, 2006) and the United States (Wilks and Spivey, 2010). We also found that social work students showed higher stress levels in their first year, perhaps associated with the move into higher education. Transitioning to college requires students to navigate rigorous academic coursework and assume greater independence with increased responsibilities (Kreniske, 2017). We also feel that high levels of academic stress for the social work students in our sample could possibly be attributed to the change in the medium of instruction for the majority, from Tamil in school to English in college. It is important to note that a sizeable percentage of social work students in this study have been placed in the 'high' stress category, and it is this group that needs to be identified and targeted for stress-alleviation strategies.
The comparative profile, however, shows that social work students experienced lower levels of stress than the reference group. This suggests that in spite of the different nature of the curriculum and training demands of the social work degree, social work students are in no way different from Correlation is significant at the ***.001, **0.01 and *0.05 levels (two tailed). Table 4. Inter-correlation matrix for key variables (social work only -T1 and T3). other undergraduate students in their experience of stress. Furthermore, no significant statistical difference has been observed for social work students from T1 to T3, indicating a certain consistency in their experience of stress. Our findings are inconsistent with other cross-sectional studies that have investigated stress in social work students. For example, Vungkhanching et al. (2017) reported that junior-year social work students in the United States perceived higher levels of psychological distress than those in their final year. An earlier cross-sectional study from India also reports higher stress levels in social work undergraduates in their final year of study than those in their first year (Stanley and Mettilda, 2016). Our results pertaining to student stress do not indicate a pattern as students move through their degree. We see that while stress scores have been decreasing in social work students from the first to the third year of their degree, mean scores for the reference group have, on the other hand, registered an incremental progression.
Resilience is an important positive resource, which has a positive effect on the psychological well-being of students (Ríos-Risquez et al., 2018). In relation to the manifestation of resilience, the high majority of both groups of respondents in our study have been classified as being 'high'. Wilks and Spivey (2010) have reported moderate levels of resilience in their sample of social work students. Furthermore, social work students in our study showed significantly higher levels of resilience than their counterparts in the reference group. However, within the social work cohort, there was no significant difference in its manifestation at T1 and T3, again suggesting a certain stability of this dimension over time. These findings are consistent with a longitudinal study done with undergraduate nursing students that also did not obtain any statistical difference in resilience scores when they were compared in their first year and then prior to graduation (Pitt et al., 2014). However, our findings are not consistent with a recent longitudinal study of social work students conducted in Spain (De las Olas Palma-García et al., 2018), which reports a gradual increase in the manifestation of resilience as students progressed through their degree.
Stress appraisal and coping can predict perceived levels of stress (Crary, 2013). The important role of coping strategies in managing stress among students has been established by previous investigations (e.g. Mahmoud et al., 2012;Wang and Miao, 2009). Coping strategies are hence important in understanding how perceived stress is dealt with and managed. Social work students in our sample scored higher in terms of both emotion-focused and problem-focused coping strategies and scored lower in the use of dysfunctional strategies than the reference group. These findings support those of De la Fuente et al. (2018), who reported that social work students used more functional forms of coping (e.g. problem-solving and cognitive restructuring) compared to dysfunctional forms of coping (e.g. problem avoidance and wishful thinking). This is important as avoidant coping is associated with poorer outcomes (Compas et al., 2001; Connor-Smith and Flachsbart, 2007) and is a significant predictor of psychological distress (Vungkhanching et al., 2017) and adverse well-being (Gibbons et al., 2011). Furthermore, task-oriented coping has been related positively to resilience, and emotion-focused coping to low resilience (Campbell-Sills et al., 2006). It is, however, important to note that both problem-and emotion-focused coping are complementary strategies rather than being two fully distinct and independent dimensions (Lazarus, 2006). The importance of emotion-focused coping and well-being is borne out by a recent study, which reported that young people with better emotion-focused coping were .80 times less likely to report suicidal ideation (Yoon et al., 2018).
The regression analysis indicates that resilience in social work students can be enhanced by reducing stress levels and increasing problem and emotion coping strategies, all three of which have been elicited as significant predictors of resilience. This is consistent with the results of Tomás et al. (2012), who obtained positive correlations between emotion-and problem-focused coping and resilience. Our findings are also in line with those of Wilks and Spivey (2010) who also reported that academic stress significantly exerted a negative influence on resilience. Studies with social work practitioners in India show that resilience moderates the effect of stress on burnout . It is hence important that the social work curriculum incorporate strategies that would enhance problem-and emotion-focused coping strategies in students, enhance resilience and prepare them to cope better with the various stressors that they face. The extant literature on resilience suggests that it can be enhanced later in life (e.g. Edward et al., 2009). Farchi et al. (2014) provide evidence of enhancing self-efficacy and resilience in social work students. We conclude our discussion by endorsing the observation of Grant and Kinman (2012) that social work training needs to focus more explicitly on developing skills that underpin resilience, and that building resilience in the future workforce should be a key ingredient of social work education.
The findings of the study need to be viewed in context and are not without limitations. Social work education and training is significantly different in India from that in the West and so the nature of stressors that students face is potentially different. Furthermore, our study has included only women students and so gender-based comparisons have not been possible. The study also has not considered the role of personality variables and social support in influencing resilience, and the literature indicates that these do influence the perception of stress, resilience and coping in students (e.g. De la Fuente et al., 2018; De las Olas Palma-García and Hombrados-Mendieta, 2017). While we have been able to portray levels of stress experienced by our respondents, the use of structured instruments limits our understanding of specific sources of stress experienced by students in our sample. Despite these limitations, this study by virtue of its longitudinal nature provides valuable insights that could be of use to academics and researchers in understanding stress, resilience and coping in social work students.

Conclusion
This longitudinal and comparative study found moderate levels of stress and resilience in social work students. Stress levels tended to be higher in the first year of their degree. Both problemfocused and emotion-focused strategies are predominantly used in dealing with perceived stress. While fluctuations in stress and resilience years were seen at different points in time, there was relative stability and no incremental progress or decline in their manifestation. Stress as well as problem-and emotion-focused coping were elicited as significant predictors of resilience. Our findings indicate the need for social work education programmes to devise strategies to enable students to deal more effectively with various stressors in their environment. Development of resilience through appropriate strategies needs to be an essential element of academic programmes to ensure that practitioners of the future are equipped with better coping and stress management skills.