Journals Library

The TARA Project: A longitudinal study of the service needs of homeless women

Project title

The TARA Project: A longitudinal study of the service needs of homeless women

Project reference


Final report date

01 June 2013

Project start date

01 April 2011

Project end date

31 March 2013

Project duration

24 months

Project keywords

Homeless; Women; Complex Needs; Abuse; Neglect; Children; Social Worker; Care; Support

Lead investigator(s)
  • Dr Emma Williamson, School of Policy Studies, University of Bristol
NIHR School Collaborators
  • Dr Hilary A Abrahams, School of Policy Studies, University of Bristol
  • Dr Karen J Morgan, School of Policy Studies, University of Bristol
  • Ms Ailsa M Cameron, School of Policy Studies, University of Bristol
  • Dr Lorna Henry, School of Policy Studies, University of Bristol



This project was collaboration between Gender and Violence Research Centre and the Centre for Research in Health and Social Care at the University of Bristol.



Project objectives


Women who are homeless often have complex social care and health needs, but all too often these needs are not fully recognised nor adequately met. Better understanding is needed of how services can best support women to move out of homelessness. This project aimed to examine the service use and needs of homeless women and consider how these can be met efficiently in both the short and long term. Its objectives were to:

  • develop a method for ‘tracking’ homeless women across different housing and support services, taking into account the ethical aspects of doing so;
  • gain a fuller understanding of the complex needs of homeless women, including their social care and health needs;
  • identify the barriers that service providers face in working with homeless women over a sustained period of time;
  • understand how best to support homeless women to access and maintain engagement with support services;
  • devise appropriate targets that can be used to monitor the effectiveness of services for homeless women.

The research also sought to develop outcomes measures that capture the complexity of work in this field.

Brief summary

The TARA project was a longitudinal study of the service needs of homeless women. The aim of the study was to identify homeless women’s experience of the wider support system over time.


The research design involved interviewing women using homeless services three times over a period of 18 months. Participating women ranged in age from 19-59 years old. Difficulties were experienced identifying younger women to participate in the study primarily due to a lack of engagement in the research by generic young people’s services. Those who provided services relating to the issues raised in the study, often did not recognise that their service users might be at risk of ‘homelessness’. The research team also interviewed 15 practitioners, from a range of services, asking both general questions about their roles and experience, as well as about their experiences of working specifically with individual clients.

An essential part of the research process was to ‘track’ women who might be moving and changing phone numbers several times within the duration of the study. In terms of ‘tracking’ women, the majority of the participating women gave the research team permission to speak to the service where they recruited them which enabled them to locate a number of women whom they might not otherwise have found. It is particularly interesting that in terms of providing consent for the research team to speak to other ‘third party’ agencies, all but one of the women gave the team this permission. In terms of the type of permission given, women either gave full or partial consent for all the services they were in contact with. Full consent allowed the research team to ask services how a client was progressing, whereas partial consent limited them to whether services had been offered. Those women who gave partial consents were more likely to be in contact with probation and other criminal justice agencies.

There were some key lessons learned from the experience of tracking which included:

  • The need for careful management of  tracking through third parties including allowing participants choice in the information-sharing/consent process, explaining the benefits of sharing personal information and the potential risks;
  • Building meaningful relationships between the research team and participants;
  • Ensuring that relationships exist between the research team and service providers;
  • Including the expertise of service user advisors across the lifetime of the research;
  • Recognising the importance in all of these relationships of maintaining values.

Practical lessons learnt in improving the ability of the research team to maintain contact with participants included: not telephoning in the morning, as those on any form of medication take time to surface, as do sex workers; recognising that women may not respond because they have no credit on the phone; and that some women were very wary of calls where they did not know who was calling. In this final case, a short text message, indicating who it was from and that a call would be made from the same number in five minutes, proved surprisingly effective.


Multiple and complex needs

The prevalence of complex and gendered issues impacting on the participants’ lives was high. This included experience of a range of issues identified in table 1.

Table 1: Presence of difficult experiences in the lives of homeless women (disclosed at interview).


Interview 1


Interview 2


Interview 3


Total and apparent

N (%)

Sex work




12 (32%)

Drug abuse issues




23 (61%)

Alcohol abuse issues




23 (61%)

Mental health issues




37 (97%)

Criminal Justice involvement




27 (71%)

Child abuse/neglect/sexual abuse




24 (63%)

Domestic violence past




30 (79%)

Domestic violence current/recent




9 (24%)

Sexual violence past




21 (55%)

Sexual violence current/recent




5 (13%)

The final column in table 1 shows the total number, and percentage, of the original 38 participating women who either disclosed experience of the issues within any of the interviews or where it was apparent, to both the interviewer and key worker, that this was an issue. As table 1 illustrates it was important to have repeated contact with the women as some women were reluctant, understandably, to disclose longstanding problems at a first interview, and new problems could arise, or be recognised by participants, at any time.

The prevalence of complex and gendered issues impacting on the participants’ lives was high. This included experience of childhood abuse, mental health problems, domestic or sexual violence, drug or alcohol dependencies, sex work and involvement with the criminal justice system. 13% of the participating women reported having 1 or 2 of these ‘issues’ to deal with, 34% had between 3 and 5 issues to address, and 47% mentioned 6-8 problem areas. 2 women (5%) identified dealing with 9-10 of the listed difficulties.

21 of the 38 participating women (55%), as far as the research team were aware, had living children and of these, 29% had children who were under 18 at the time of the first interview. Of these, 4 women had no contact at all with any of the children; 6 had occasional contact with some, but not all, of their children; 1 was in regular contact; and 4 had some young children with them, but had no contact with others. This was the issue which participants found most difficult to deal with and therefore discuss. Even during later interviews participants found this aspect of their experiences difficult to talk about and in many cases the research team were only made aware that women had had children at later interviews. This area of homeless women’s lives was also an issue for adult and children’s practitioners who often struggled with the contradictions inherent with balancing the needs of women and their children.

Practitioners’ perspectives

In comparing the interview data from service users and practitioners the research team were able to identify some key areas of interest.

Workers identified concerns about the lack of follow-up information they received about clients who disengaged or were evicted from a service. In these cases workers often did not know what had happened to a client and therefore were not able to see the positive, or negative, outcomes for clients they had worked with. Practitioners expressed serious concerns about a lack of coordinated provision through joint working, particularly around complex needs and mental health support. Where some clients engage with multiple agencies, workers across the services sometimes abdicated responsibility because other service providers were involved with a client.

Practitioners discussed difficulties in setting clear boundaries in relation to the expectations of clients and service providers when working with service users. For example, if a client reduced their drug or alcohol use but had not stopped entirely. For some clients this was deemed a positive move, while practitioners might deem it a failure. In these cases practitioners sometimes drew on a concept of ‘rock bottom’ to justify a breakdown in the relationship between workers and clients rather than considering whether expectations were realistic.

The majority of practitioners who were interviewed raised concerns about the reduction in time allocated to working with service users with complex needs. As a result of funding cuts and strains on services due to increased targets, agencies were often reducing the amount of time a client could access a service which many practitioners felt was counter-intuitive given the complexity of the needs being faced by this client group. This concern about the reduction of resources was not limited to the research team’s interviews with practitioners, but was also a theme which emerged within their interviews with the participating women. Many were aware of the cuts to services and the strain under which practitioners were operating. For many, uncertainty about the services they were receiving, whether support would be ending, or whether service funding would be cut, made them feel guilty about receiving services and further added to the complex issues they were attempting to address.


Homeless women are used to making themselves invisible in order to survive. Services need to work hard to build meaningful and trusting relationships to enable women to access appropriate services and to remove the barriers which currently prevent women from accessing the services they need.

Summary of key findings and recommendations

The homeless women in this study were struggling to survive the impact of a large number of traumatic, and often gendered, life events. These experiences contributed to the multiple service needs they identified. All services should recognise the gendered nature of homeless women’s experiences, and the impact these gendered life events have in terms of complex trauma.

Both service users and practitioners recognised the need for a pivotal key worker role to facilitate access across services. This includes sharing information with workers about clients’ progress through the system as a whole, as well as facilitating better joint working across agencies with often competing aims and objectives.

When asked about the services and support they were receiving the participants did not voluntarily refer to involvement with adult social care services. When explicitly prompted participants immediately associated social care with children's social workers and talked about either their own experience of social services as a child, or in relation to their involvement with regards their own children. None of the participants referred to having a social worker for themselves, as adults. In relation to wider support, they did refer to mental health teams, workers within housing services, and a wide range of third sector support workers. This finding suggests that other workers, often housing support workers, are fulfilling functions that in the past social workers would have carried out, for example by providing and co-ordinating services.

From the women’s accounts the research team found that re-telling their life story to a large number of practitioners, over and over again, in order to access services, was felt to be humiliating and disempowering and contributed to the impact of complex trauma. Women are willing to tell their stories if they feel valued and listened to. Where they share personal information they want to be involved in information-sharing decisions and for disclosures to result in a greater understanding of their needs across different services.

Service users preferred practitioners who treated them as human beings, genuinely listened to them, and took the time to build meaningful and trusting relationships. Where these relationships existed, clients felt supported to begin to take control of their own lives.  Practitioners often felt that their services lacked the resources to adequately build meaningful relationships with clients.

Stable and emotionally and physically safe environments are essential to addressing the needs of homeless women. Women are unlikely to be able to access services related to their emotional well-being and traumatic histories whilst they feel vulnerable and insecure. In order to recognise that homeless women may not feel safe or ready to disclose traumatic personal information, questions about the potential issues affecting clients should be asked at different points in time and not just at an initial assessment.

Service users recognised the constraints on workers in terms of funding cuts, however, this often resulted in them feeling worse, guilty about having a service, and less supported.  Organisations need to ensure safeguards are in place to prevent low staff morale, where this exists, from compounding the barriers which prevent women being able to access services.

Many women, including those who had made progress over the course of the research, still felt vulnerable at times to relapse and felt that some form of on-going, low level support (such as on-demand, infrequent telephone contact with a pivotal key worker) would  reduce the risk of them, and other homeless women, re-entering the system. Commissioners need to recognise that limiting provision to only short term interventions is an inefficient use of resources, and a failure of care, if it results in women becoming homeless again.

Plain English summary

Homeless women often have complex social care and health needs. This project explored how homeless women could benefit from support by social services and from mental health, housing and other public and charity services. It also looked at how services could operate better for these women. Researchers interviewed 38 women initially, 28 women six months later and 22 at a third and final interview six months later. Interviews were also conducted with social workers. Part of the research process involved tracking the women through their contacts with social services and, for example, housing agencies. Nevertheless, it was often difficult to maintain contact with participants.

The homeless women in this study were struggling to survive the impact of a large number of traumatic life events, often resulting from abusive and/or violent behaviour of men. These experiences contributed to them having multiple needs. Issues impacting on significant numbers of participants included involvement in prostitution, crime, and drug and/or alcohol abuse. Most participants reported experience of abuse (including sexual abuse) and/or neglect during childhood, and experience of domestic and sexual violence in adulthood. Almost all (97%) reported mental health issues.

Researchers found that repeated contact with the participants led to them being more open about their problems. More than half the women had children and they had most difficulty dealing with issues involving their children. Women reported that re-telling their life story over and over again to social workers, in order to access services, was humiliating and degrading. Workers often felt they did not have enough time to help homeless women. Both reported that reductions in services made the situation worse.

Practitioners had concerns about a lack of follow-up information about women who ended contact or were denied service, and about failures to coordinate different services to meet the women’s complex needs. Both service users and workers believed a key worker system could improve the situation.

Many women, including those who had made progress over the course of the project, still felt vulnerable to relapse at times and would have liked the option to telephone a key worker at times of crisis.

Homeless women are used to making themselves invisible in order to survive but they responded to workers who treated them as human beings, listened actively and took the time to build relationships. Such stable, supportive relationships could only develop when the women felt safe.


Published articles

  1. Abrahams HA, James J, Powell L, Williamson E, Morgan KJ, Cameron AM, Henry LW (2015) Service user involvement in longitudinal research, International Journal of Qualitative Research in Services, 2, 1, 13-27.
  2. Cameron AM, Abrahams HA, Morgan KJ, Williamson E, Henry L (2015) From pillar to post: homeless women's experiences of social care, Health and Social Care in the Community, 23, 5. Published online: 26 Feb 2015.
  3. Williamson E, Abrahams HA, Morgan KJ, Cameron AM (2014) The TARA Project: Tracking homeless women in longitudinal research, European Journal of Homelessness, 8, 2, 69-91.
  4. NIHR School for Social Care Research (2014) The TARA Project: A longitudinal study of the service needs of homeless women, Research Findings 13, NIHR School for Social Care Research, London.


  1. Karen Morgan and Hilary Abrahams, A longitudinal study of the service use and need of homeless women, SSCR Annual Conference 2013, London, March 2013.

Public involvement

Service users and professionals were members of the project Advisory Group.

Research advisors

This study asked previously homeless women who were in stable accommodation and who had a wide experience of statutory and voluntary agencies to join the research team as service user advisors, to assist in the design and execution of the research. Four participants from a previous study by the core research team were approached and asked if they would join the team. All of them agreed. These advisors supported both the design and recruitment. The work that they did on the interview schedules, the explanatory leaflets and the consent forms had a major impact on the content and appearance of the material and, in fact, made them easier to work with. They asked direct questions and challenged the team’s assumptions about how women would react to the research questions. They suggested areas to explore that had not been considered, because they were beyond the research team’s experience, but which added depth and understanding to the research. The advisors also suggested that there were questions that should be asked when the research team were reluctant to include them.

A number of key issues were considered by the research team in involving service users as advisors, including for example ethical considerations (for both the study and the impact on the advisors themselves of being involved in such a study). Although it was a steep learning process for all concerned, the advisors became a valued part of the research team and valued the experience themselves.

The research team’s and advisors experiences are discussed in Voices of experience: service users as advisors in a research project on the needs of homeless women(


The TARA project has provided important findings on the method of “tracking”. This is particularly important where service users or participants are from vulnerable groups without a fixed contact or address (such as homeless women). The research team’s experiences are discussed in Tracking homeless women in qualitative longitudinal research(

The study also provides additional evidence on the involvement of service users as research advisors in research studies (see above).

The TARA project held an event for study participants and advisors at the Wild Goose café on the 19 April 2013. The Wild Goose Café is run by Crisis Ministries who currently run a night shelter for women three nights a week. Research participants and advisors were invited to attend, to hear about the research and to celebrate their involvement in the research. Over 90 delegates attended including service providers, front line workers, commissioners and researchers from Bristol and beyond. At the event, women were asked to send messages to the research team, other homeless women, and service providers. In relation to their involvement in the project, one participant told us, “All the support and people being there has changed my life, being able to be honest was the start of my recovery. Your time and unconditional support means so much to me. Thank you.”

The research team also held an end-of-project conference on 24 April 2013, which was equally well received with over 90 delegates attending and engaging throughout the day. The research team presented three papers from the research findings focusing on the experiences of homeless women, practitioner perspectives, and issues related to safety, home and empowerment. Additional papers were presented from external speakers: Sandra Ferguson, a Consultant Clinical Psychologist from Edinburgh Hospital, talking about complex trauma, and Jennifer Holly and Miranda Horvath, talking about their findings from the Stella Project – a project looking at the intersection between services addressing domestic violence, alcohol and drug use, and mental health service provision.

This project was funded by the National Institute for Health Research School for Social Care Research (project number T976/T10-032/UBEW).

Department of Health Disclaimer

The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the NIHR School for Social Care Research, NIHR, NHS or the Department of Health.