Intersectional impact of multiple identities on social work education in the UK

Summary The study reviews the records of 671 social work students and graduates including the seven intakes from the first cohort in 2003/2004 to the intake in 2010/2011 to examine the interacting effect of learning difficulties, ethnicity and gender on the completion of social work training at a university in the South East of England. Findings Among the students, 79.9% of them were female, 50.1% were black, 27.9% white, 10.7% Asian and 11.3% other ethnicities. A majority of students did not report any disability. Among those who did (n = 84), 52.3% (n = 44) reported a learning difficulty. The percentage of students who have successfully completed the training is 76.4%, a completion rate that is comparable to the UK's national figure. Having controlled the confounding variables, hierarchical logistic regression identified the risk factor for dropout from undergraduate social work programme as black female students with learning difficulties (odds ratio = 0.100, 95% confidence interval = 0.012–0.862, p < 0.05). Findings suggested that students with multiplicity of identities, i.e. being black and female and with a learning difficulty, have a lower probability to complete the programme successfully. Applications Strategies for tackling the intersecting disadvantages of race, gender and disabilities in social work training should embrace three principles: providing continuous support, focusing on how the support is provided and addressing contextual and structural barriers.


Introduction
A UK national report on the retention and completion rate of students studying at higher education institutions (HEIs) showed that 78.1% of local UK students were qualified with a first degree (National Audit Office (NAO), 2007). According to this result, UK ranked as the fifth highest in estimated graduation (survival rates) in the Organisation for Economic Co-operation andDevelopment in 2004 (NAO, 2007). However, nearly one in seven social work students in England failed to complete their training in 2008/2009, with some colleges showing an attrition rate of up to 39% (McGregor, 2010). Stevens et al. (2010) reported that the dominant reason for being social workers among UK students' is altruistic motivations. The high dropout rate may affect the number of newly qualified social workers for helping vulnerable people in the future. Therefore, social work educators and policy-makers should develop effective strategies to enhance the retention and completion rate in social work training in the UK.
The literature indicates that there are many factors contributing to the failure to complete higher education, for instance, 'lower socio-economic status, mature entry, being from certain ethnic groups, late starting, living at home and being male' (Hassell, Seston, Eden, & Willis, 2007, p. 250). Students with learning difficulties may encounter various emotional consequences and challenges including 'poor self-concept, isolation, inadequate peer relationships and academic failure' (Roer-Strier, 2002, p. 914). Holland (2011) and Higher Education Statistics Agency(HESA) (2011a) referred learning difficulties to specific challenges and problems in learning but not including those who have a significant general impairment in intelligence. Data from HESA (2011a) indicated that among all students who self-declared to have a disability, 44.8% claimed to have a specific learning difficulty, rendering it a commonly reported disability among first-year university students. Some students may experience transitional difficulties if their prior experience of assessment does not enable them to deal with the assessment methods used in social work education (Worslet et al., 2009). Therefore, those students with learning difficulties may find these demands particularly challenging.
While learning difficulties may contribute to non-completion of training, having a disability does not necessarily result in poor attainment among students (Richardson, 2010). The interaction between student and institutional factors may affect the retention rate in social work training programmes (Davis, 2010). The cultural capital of students may be devalued through the organisational and management progress of HEIs . This echoes NAO's (2007) suggestion that non-completion of a programme may not be solely due to one single reason but rather a mixture of personal, institutional, course related and financial factors. Analysis on the UK data regarding students registering for a Diploma in Social Work (DipSW) programme from 1995 to 1998, it is found that men, students with any form of disability and black ethnic minorities were the three significant risk factors for less likely to pass at first attempt (Hussein, Moriarty, Manthorpe, & Huxley, 2008;). This shows that an increasing number and diversity of social work student cohorts may affect their retention and completion rate in social work training . The findings are similar to the report of Mir, Nocon, Ahmad, and Jones (2001) that 'people with learning difficulties from minority ethnic countries experience simultaneous disadvantage in relation to race, impairment and, for women, gender' (p. 2). Racism and disablism are barriers affecting the improvement of services for black disabled people (Nasa, 2002).
Social work is a predominantly female profession (Hyde & Deal, 2003). However, some women in the profession may face discrimination because of their race (Pease, 2011). Unlike the US where social work profession is increasingly diversified in terms of race and ethnicity (Schilling, Morrish, & Liu, 2008), a majority of registered social workers in the UK are white (70%), followed by black (10%), Asian (4.3%) and other ethnic minority groups(General Social Care Council (GSCC), 2010). Whites are proportionally over-represented in the social work profession in the UK, which matches the general impression that social work is a white profession (Perry & Cree, 2003). Drawing from the findings of a qualitative study on diversity and progression among social work students in England, Bernard, Fairtlough, Fletcher and Ahmet (2011) concluded that there are cumulative effects on certain vulnerable students, which may result in delay progression. Ghazala (2004) argues, 'people with learning difficulties who are also from ethnic minorities face double discrimination and exclusion in all areas of their lives' (p. 38).
Cumulative advantage theory originated from studies on social stratification and was expanded to examine cumulative disadvantage, which is the inequality on socioeconomic outcomes within a group or within an entire population because of the direct and indirect effects of accumulation of life-course capital (DiPrete & Eirich, 2006). Cumulative disadvantage is a vicious cycle that exposure to earlier disadvantages will continue to have an effect across the life course. There are three types of cycles of disadvantages: intrinsic and systemic cycles, acquired and cumulative cycles and period and episodic cycles (Australian Government, 2011). However, Mullaly (2010) argues that 'multiple oppressions are not simply cumulative or additive '(p. 199). From the intersectional model, the multiplicity of oppression may be a consequence of the intersection of individual's multiple identities. In other words, people may experience oppression and social inequalities because of the interaction effects between a combination of their social identities that are contextualised within systems of power, domination and oppression (Koehn, Neysmith, Kobayashi, & Khamisa, 2013, p. 446). To effectively eliminate the multidimensional disadvantages in social work training, the focus of interventions should be directed at those people who are highly vulnerable even though they are not currently at risk (Layte & Whelan, 2002).
While good social and academic integration can prevent students from social or academic failure (Hassell et al., 2007), cultural sensitivity seems to be not fully integrated into social work education despite the fact that the proportion of students of minority groups in schools of social work is increasing in western countries (Schiff & Katz, 2007). If social work students of minority groups cannot develop good interrelationships while at university, they may encounter difficulties arising from cultural differences, which may affect their progression of study. The interaction effects of lacking awareness, knowledge and sensitivity of different culture in higher institutions may result in a lack of recognition of the particular needs of students of minority ethnic groups, which may subsequently affect these students' ability of completing their course (Hoodless, 2004). Furthermore, higher education and especially social work education in the UK will encounter serious hardship because of financial cuts (Lymbery, 2011). The increasing staff-student ratio and limited resources become a huge challenge to HEIs in widening participation and diversity of learning (Hafford-Letchfield, 2007).
Although disability, gender and ethnicity are determinants of social work students' non-completion of their study, those students who are more emotionally resilient are more resistant to stress and psychologically healthy (Grant & Kinman, 2011). This indicates that students with good self-competency, persistence and resilience can handle their challenges effectively despite multiple disadvantages. Yet, it is not widely reported in the literature whether there is an interacting effect of these three determinants on social work students' academic performance, and whether the interacting effect has significant impact on social work students in the UK. This article may provide an international practitioner audience with an evidence-based understanding of the challenges social work students with multiple identities encountered in their social work training in the UK.

Research design
This is a single case study design using the documentation review method to analyse the profiles of the graduates of a university in the South East of England. The chosen university was one of the Health and Care Professions Council and The College of Social Work approved HEIs providing social work programmes from undergraduate to postgraduate level in the UK, since the introduction of an undergraduate degree qualification for replacing the DipSW in 2003. From the first cohort in 2003/2004 to the intake in 2010/2011, there were a total of 671 students in the Bachelor of Arts (BASW) (both full-time and part-time employment route) and Master of Arts (MASW) degrees in Social Work. The archival data of these seven cohorts were analysed. Instead of examining the progression of these social work students, the focus of this study was to identify the risk factors attributed to their dropout, because of either failure or withdrawal, from the programmes. In this article, the dropout rate is defined as the percentage of students who have withdrawn from the qualifying social work training within or after the maximum period of registration, i.e. four years of their first entry into the MASW programme, five years for BASW (full-time) programme and seven years for BASW (part-time) programme.

Background
The MASW is a two-year full-time programme open to all non-social work degree holders. Students in the BASW (full-time) programme normally take three years to complete their study, while part-time students may need to take longer as they are working full-time in social care sectors simultaneously with their study. In terms of assessment, students are assessed by means of placement portfolios, written assignments, presentations and exams. There are two placements in the MASW and the BASW programmes, respectively. Students' annual academic performance are discussed at the Examination Board in June every year, and students may be required to leave the programme if they cannot fulfil all the academic and professional requirements.
According to the figures of the university, the total number of students who declared having disabilities (7.88%, 1217 out of 15,446 on 31 March 2011) is similar to the HESA figure. However, the prevalence of learning difficulties at the university (57.4%, n ¼ 698) is higher than the UK national figure (48.9%) (HESA, 2011a). A majority of the 100,882 registered social workers (87%) reported to be disability-free in 2009 (GSCC, 2010). Among the students studying at the College of Health and Life Sciences at the University such as social work students, the prevalence of learning difficulties was 24.9% (n ¼ 174) in March 2011. Compared with the profile of professional social workers in the country, the university has a higher number of students reporting learning difficulties. About fourfifths of those students in the qualifying social work programmes were female, a proportion that is similar to the UK's national figures in (HESA, 2011b. On the other hand, half of the social work students were black indicating the proportion of black students in the university was significantly higher than that of other universities in the UK (HESA, 2011c).

Data collection
To ensure confidentiality, no students or graduates were interviewed or followed up. When students apply for the social work programmes in the UK, they are required to complete a self-administered University and College Admission Service (UCAS) form to declare their academic results and socio-demographic data such as gender, ethnicity and their perceived disabilities. In addition to data from students' UCAS forms, information including students' final award and confirmed diagnosis of disabilities were also collected from the university. All the students' names were anonymised and partial identification such as students' ethnicity and gender were managed in a sensitive manner to ensure confidentiality. The research ethics committee of the university, therefore, granted ethics approval for the present study.

Statistical analysis
Using SPSS 20.0, the relationships between dependent variables (dropout from the programme) and independent variables including self-reported disability, ethnicity and gender were examined by Chi-square tests. Hierarchical logistic regression was also used to identify attributive factors for successful completion of social work training at the University. Hierarchical logistic regression models measure the association between attributive factors and an event after adjusting confounding variables. If the odds ratio is equal to 1, there is no association between the dependent and independent variables. The larger the value of odds ratio, the stronger the association. Value of odds ratio less than 1 indicates a negative association between the attributive factors and the event.

Results
A total of 671 students have enrolled in the university's social work programmes over the past seven intakes;34.4% (n ¼ 231) were in the MASW programme, 54.2% (n ¼ 364) in the full-time BASW programme and 11.3% (n ¼ 76) in the part-time BASW programme (Table 1). A majority of students (79.9%, n ¼ 536) were female, and only about one-fifth (n ¼ 135) were male. One half of the students were black (50.1%, n ¼ 336), 27.9% (n ¼ 187) white, 10.7% (n ¼ 72) Asian and 11.3% (n ¼ 76) other ethnicities (11.3%). A majority of students did not report any disability. Similar to what Hussein et al. (2009) reported, the most frequently report disability in this study is learning difficulty. Among those who did (12.5%, n ¼ 84), just over half (52.3%, n ¼ 44) reported a learning difficulty, which was higher than the UK's national figure.
About three-fourths (n ¼ 356) of the students have successfully completed the professional social work training, a completion rate which is comparable to the UK's national figure (NAO, 2007). The dropout rate was 23.6%, which is significantly higher than that reported in Hussein et al.'s study (2008). Among those students who have dropped out from the training (n ¼ 110), 74.5% of them (n ¼ 82) were in the BASW programme and 58.2% of them were black ( 2 ¼ 19.03, df ¼ 6, p ¼ 0.005). Table 2 shows that 54.5% of those who have withdrawn from the programme were due to personal reasons such as health problems, family issues, financial difficulty and other personal factors. However, 51.9% of them in the BASW programme left because of poor academic performance (p ¼ 0.004).
Chi-square tests showed that among the students who could not complete the programme successfully, 34.4% of them reported learning difficulties (p ¼ 0.029), 27.6% of them were black (p ¼ 0.044), 42.3% of them were female with learning disability (p ¼ 0.021), 43.8% of them black with disability (p ¼ 0.005), 46.2% of them were black female with disability (p ¼ 0.005) and 61.5% of them were black female with a learning difficulty (p ¼ 0.001) ( Table 3).
There are four hierarchical logistic regression models in Table 4, but the statistics of Hosmer-Lemeshow goodness-of-fit test shows that only the BASW model is good fit for the data. Model 1 of the BASW indicates that only gender (i.e. female) is the significant factor attributes to the success of completing the BASW programme (odds ratio (OR) ¼ 2.072, 95% confidence interval (CI) ¼ 1.142-3.760,  CI ¼ 0.012-0.862, p < 0.05) ( Table 4). In other words, those black female students with learning difficulties have the odds of 90% decrease in completing their training comparing to those students who were not black female with learning difficulties in the same programme.

Discussion
Hierarchical logistic regression models show that personal characteristics did not have any negative impact on the academic performance of those postgraduate students in the MASW programme. However, analysis on the BASW programme indicates that though female students were more likely to complete the programme successfully, the intersectional effect of black, female and learning difficulties contributes to non-completion in the programme. Similar to other disciplines such as pharmacy (Hassell et al., 2007), female social work students are more likely to complete their training successfully than male are. However, after controlling confounding variables, the intersection between race, gender and learning difficulty has moderated the academic performance of some students. In other words, black female students with learning difficulties have a relatively higher probability in failing the BASW programme. The result is similar to the analysis of 10,891 fulltime social work students during the period 1995 and 1998 in England (GSCC, 2008), and that of Bernard et al.'s study (2011). GSCC explained that one of its contributing factors for unsuccessful completion of social work training was due to the failure of current support systems in meeting the complex and diversified needs of social work students (GSCC, 2008).
Students' persistence and resilience in dealing with pressures Baron, Philips, and Stalker (1996) argued that social work students with disabilities 'felt under added scrutiny' (p. 373) from their placement. The practice learning is in fact an essential factor affecting the progression in social work training (GSCC, 2008). Social work training is relatively demanding compared to non-professional training programmes -for example, 26% to 54% of social work students encounter client violence such as verbal abuse and physical assault in their placement (Criss, 2010). However, the present study found that mature postgraduate social work students, despite having learning difficulties, were less likely to withdraw from the training programme than undergraduate students did. Bernard et al. (2011) suggest that those students who could overcome the cycle of disadvantage and barriers to progression might be due to their persistence and resilience. In addition, those students who were more emotionally intelligent were more resilient to stress in social work training (Grant & Kinman, 2011). All masters' students are survivors in the competitive higher education environment and have already developed successful strategies to cope with the challenges of learning in higher education from their previous study. This is in line with what Ozga and Sukhnandan describe (1998), 'mature students were more focused than conventional students on the academic demands of university life' (p. 327).

Intersectional effect of multiple identities on academic performance
From the hierarchical logistic regression models, black female students with learning difficulties have a higher risk of failing in the University's BASW programmes than their counterparts. This is similar to the findings of the UK's Equality Challenge Unit (2010) that black students are less likely to achieve good degrees than white students. Our data suggest that there is an intersecting nature of disadvantages in vulnerable students studying in the BASW programme. In other words, those BASW students with multiple identities, i.e. being black, female and having a learning difficulty, may experience multiple disadvantages within a social context that assigns very different value to such identities. This may explain the reason why 'the effects of learning disabilities on academic, vocational and emotional wellbeing not only continue but often increase over time' (Roer-Strier, 2002, p. 915).
In essence, when a BASW student possesses any of the three identities -female gender, black ethnic group and report of learning difficulties, she or he may have a higher probability of failing a qualifying social work programme and being denied access to the profession. Yet, because of the limitations of this small-scale cross-sectional study, it is impossible to identify the causal link among variables. It is, therefore, difficult to determine whether the high dropout rate of these students with multiple disadvantages is attributed to multiplicity of disadvantages or individual factors. This may require further investigation in future by large-scale longitudinal studies.

Implications for social work training
In response to the findings of the present study, interventions for tackling the multiplicity of disadvantages should embrace three principles: providing continuous support, focusing on how the support is provided and addressing contextual and structural barriers (Australian Government, 2011).
Regarding providing continuous support to students with disabilities, Parker (1998) observed that since the first Disability Discrimination Act in 1995 individuals with disabilities has more equal access to higher education in the UK. The rights of disabled people to receive education in HEIs are further protected under the Equality Act 2010. Yet, having more equal access to higher education does not necessarily mean disabled students can receive adequate support from HEIs in their studies. The Widening Participation policy of the Higher Education Funding Council for England (HEFCE) clearly requests HEIs to improve provision for disabled students. In view of how the support is provided, the provision of support to students with severe learning difficulties in further education is poor and unfocused (Wright, 2006). In reply to the criticism of HEFCE regarding the inconsistent practice for disabled students across the HEIs (HEFCE, 2009), social work educators should, therefore, pay attention to their students' needs, pace and style of learning (Worslet et al., 2009).
Regarding focusing on how the support is provided, teachers should receive proper training in designing effective interventions to develop sound knowledge of learning principles and instructional strategies for helping students with learning difficulties (Woolfson & Brady, 2009). This can enhance teachers' self-efficacy, beliefs and competence in interacting with disabled students. It may also enable them to have a more positive perception of disabled students' learning and coping abilities. In fact, students' perceptions towards teachers' practical and emotional support contributed significantly to their successful completion of social work training (Simpson, Mathews, Croft, McKinna, & Lee, 2010). Involvement of individuals with learning difficulties in work has great impact on challenging oppression in the workplace (Skelton & Moore, 1999). Involvement of service user is also highly valued and recommended in social work education (Branfield, Bereford, & Levin, 2007). While disabled students can have more opportunities to be involved in consultation when universities improve their provision of facilities for these students now (Bhatia, 2005), students who are users of social work training programmes should also be actively involved in the design and delivery of the programme.
Higher education in the UK has an ambivalent role in ethnic equality (Shiner & Modood, 2002) as minority ethnic students may suffer from different forms of racism (Jesspo & Williams, 2009). Social work educators should, therefore, tackle the contextual, structural, environmental and personal issues, such as poor social integration, help-seeking behaviours and negative interaction with university personnel, that affect the academic performance of students of minority groups (Manalo, Ede, & Wong-toi, 2010). It is also essential to increase staff awareness and skills for helping those who are doubly discriminated (Ghazala, 2004). For example, university administrators and specialist disability officers should provide disability awareness training to academics to enhance their understanding and acceptance of disability (Stanley, Ridley, Manthorpe, Harris, & Hurst, 2007). This is to ensure that university staff can develop a clear understanding of antiracist and anti-discriminatory practice in higher education setting.
Academics should use different strategies to advance disabled students' studies rather than placing responsibilities for predicament on disabled students (Richardson, 2010). As many students with learning difficulties have experienced lifelong marginalisation, social work educators should ensure continuous commitment of resources to maximise their potentials (Dowse, 2001). Nind and Seale (2009) suggested a multidimensional model of access, which is related to both the process and product of access for enhancing the quality of life of people with learning difficulties. These recommendations can guarantee accessibility for people with interactional disadvantages to higher education.

Limitations
There are some limitations in this small-scale cross-sectional study. As the number and classifications of disabilities collected from the UCAS application form were mainly based on students' self-reported information, the actual prevalence of disability in British HEIs may be under-reported (Miller, Ross, & Cleland, 2009) or over-reported. Accordingly, students should be given a Disability Discrimination Act definition of disability when they complete the UCAS form or formal assessment if they are uncertain of what kind of disabilities they have.
Moreover, the present study examined the intersectionality of multiple identities by using secondary quantitative data collected from students' UCAS form and their academic profiles. Wineman argues that oppression is a qualitative and not a quantitative experience, and therefore, it is not possible to quantify the impact of effect of oppression (cited in Mullaly, 2010).However, Covarrubias (2011) suggests that a quantitative intersectional analysis can provide a complete account of the impact of multiple identities on the population individually and collectively. And the patterns of impact may not be captured by qualitative approach. Therefore, a mix of methodologies and research designs should be used for future research regarding intersectionality (Koehn et al., 2013).
Despite these limitations, the present study has provided an update of what particular contextual and structural barriers may have possibly hindered social work students at this university from qualifying as a registered social worker in the UK. It has also provided a new direction for future research in exploring the intersectional impact of multiple identities on the retention and completion rate in higher education.

Conclusion
The present study identifies an intersectional effect of multiple identities in social work education in the UK. Black female students with learning difficulties were found to be more likely to fail qualifying social work programme. Unlike Pacheco and Plutzer's (2008) cycle of cumulative disadvantage, the present study provides quantitative findings to support the idea of intersectionality, which believes that 'different systems of inequality are interconnected' (Mullaly, 2010, p. 195).

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethics
The research project was approved by the Research Ethics Committee of Brunel University London.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.