Researching Sexual Violence against Older People: Reflecting on the use of Freedom of Information Requests in a Feminist Study

Background

Sexual violence is recognised as a serious violation of human rights affecting millions of people worldwide, the majority of whom are women. According to the World Health Organisation (2013), around 35 per cent of women will experience at least one form of sexual violence in their lifetime.

The focus of the majority of studies, campaigns, policy and practice developments over the last four decades have been on young women, aged 16 to 30 years, who are considered to be the most statistically at risk of becoming a victim of rape or sexual assault (Breiding et al., 2014; Office of National Statistics, 2015). Older women (and men) have been excluded from the majority of feminist and criminological research (Jones and Powell, 2006). Most notably, in England and Wales, the Crime Survey for England and Wales (CSEW) has imposed an upper age limit of 59 years on the intimate violence module that collects data on domestic and sexual victimisation (Walby and Allen, 2004). The exclusion of older women from sexual violence research and discourses has been noted by several commentators (Ball, 2005; Jeary, 2005; Jones and Powell, 2006; Lea et al., 2011; Mann et al., 2014); however, there remains a lack of research and policy attention to those aged 60 years and over and there have been few research studies specifically examining sexual violence against older populations.

In the UK, the only study to examine the extent of sexual violence was conducted by Ball and Fowler (2008). The study was based on all recorded sexual offences within a semi-rural English county with a population of approximately 800,000. They analysed all recorded sexual offences over a 5-year period (n = 1061) and found 3.1 per cent involved victims aged 55 years or more. Internationally, a small number of studies attempting to measure the prevalence of sexual violence against older women have been conducted, mainly in the USA, and report rates of between 0.9 per cent (Cannell et al., 2014) and 7 per cent (Ramin et al., 1992); however, methodologies, size of the samples, and definitions of ‘older’ range across these studies, making cross-comparisons difficult.

The purpose of this paper is to outline the methods used to conduct the first national study examining the extent of recorded sexual violence against people aged 60 years and over in the UK, which formed the first phase of a larger mixed-methods study. Freedom of Information (FOI) requests were used to gather police-recorded data with positive results; forty-five out of the forty-six UK police forces provided the requested information. This article outlines the positives and challenges of using this method with the central aim of encouraging other feminist researchers to adopt this in future research.

Feminist Approaches to Understanding Sexual Violence

Feminist approaches to sexual violence have developed over the last four decades since the second-wave women's movement in the late 1960s drew international attention to the widespread incidence of sexual violence against women and girls (Kelly, 1988; Brownmiller, 2013 [1975]). The feminist analysis of rape and other forms of sexual violence emerged from early criticisms of a range of academic disciplines that were accused of being oblivious to gender (Kelly, 2013).

Feminist scholarship has focussed on two key interrelated issues in explaining sexual violence against women. Gender (rather than sex) and power are seen as fundamental in defining, explaining and understanding sexual violence. It is argued that, through the societal subordination of women and simultaneous superiority and advantages that men receive in patriarchal society, widespread sexual violence against women is enabled. Within this construction, sexual violence is understood as a form of gendered control, whereby men exert their power and control over women as a way of oppressing women and maintaining the patriarchal system which advantages men (Baron and Straus, 1987; Daly and Chesney-Lind, 1988; Scully, 1990; Avakame, 1999), focussing on societal and individual causes of rape. One of the primary ways of challenging this is to make it visible, and much of feminist efforts has focussed on raising awareness of the extent of sexual violence against women.

Whilst this construction of sexual violence is not limited to women in certain age groups, the vast majority of feminist scholarship and activism in relation to sexual violence has focussed on younger women (Whittaker, 1995; Jones and Powell, 2006). Moreover, the majority of research into violence and abuse of older people (usually termed ‘elder abuse’) has instead been approached from a gerontological perspective where age, rather than gender, is the primary issue. In the context of understanding domestic and sexual violence, prioritising age as the principle factor has been criticised for being oblivious to gender and not acknowledging that the majority of victims of sexual violence are female, whilst the perpetrators are overwhelming male (Whittaker, 1995). Furthermore, the dynamics which characterise much of the sexual violence in younger populations (namely power and control) are argued to be the same in older populations and therefore a distinct analysis of sexual violence based on the age model is argued to be unnecessary and depicts abuse to be, at least partly, the victim's fault for being ‘old’ (ibid.). Despite commentators criticising the application of an age model to understanding sexual violence and calling for a feminist perspective, there remains a lack of feminist research exploring sexual violence against older people.

Feminist Methodological Approaches to Researching Sexual Violence

Traditionally, feminist research was associated with qualitative methodology; the early emerging works of Brownmiller (2013 [1975]), Oakley (1974) and Kelly (1988) established qualitative methods, particularly qualitative interviews, as a method particularly suited to research with women exploring domestic and sexual violence. Despite differing feminist theories and approaches, a common aim across the strands is commitment to conducting research for women and, where possible, with women (Ramazanoglu and Holland, 2002). Qualitative interviews have been considered a particularly suitable method of achieving these principles by empowering women through the research, using their own experiences and stories and allowing for reciprocal dialogue between the researcher and participant, which removes the traditional hierarchy observed in quantitative methods. However, a range of methods are now seen as suitable for achieving these aims (Oakley, 1999; Westmorland, 2001). Whilst quantitative methods were once largely seen as unsuitable for feminist research because the fundamental principles of objectivity and validity did not allow for the lived experiences of women to be at the core of research, over the last two decades the quantitative/qualitative divide in feminist research has been challenged (Oakley, 1998; Westmorland, 2001), and feminist research increasingly utilises quantitative research either as a standalone methodology or as part of a mixed-methods approach. As Kelly (2013) points out, feminist research involves both a theoretical premise and practical commitment to understanding women's oppression in order to change it. Such a definition allows for more than one theoretical approach and a variety of methods.

There are four key primary advantages to using quantitative methods in feminist research (Miner-Rubino and Jayaratne, 2007). First, quantitative methods such as surveys can introduce social justice issues, such as racism, sexism or ageism, into the mainstream discussion in public policy arenas; second, statistics and numbers can be more concise and therefore can be easier to remember and communicate to others; third, quantitative methods can be useful in identifying patterns in women's oppression, which can then be used to inform policy and social change; and, fourth, methods such as surveys can access large numbers of people and can therefore potentially represent a wider population from multiple perspectives (ibid.).

However, whilst quantitative methods can provide the ‘bigger picture’ of social problems, qualitative methods are more suited to telling the ‘personal story’, and increasingly social researchers, including feminists, have been realising the benefits of mixing quantitative with qualitative methods (Hodgkin, 2008). Combining methods, such as questionnaires with interviews, allows feminist criminological researchers to examine the extent of violence as well as the lived experiences of those who have been victimised. Freedom of Information requests are a method of obtaining quantitative data but to date have not been utilised in feminist research, either as a standalone method or in combination with other methods.

Freedom of Information Requests in Social Research

Internationally, legislation and policy governing access to information is not a new development; Walby and Larsen (2012) point out that Sweden's Freedom of the Press Act of 1766 is the oldest existing form of access law, and the ‘global trend’ of openness has seen a significant increase in the adoption of FOI laws across the world (Byrne, 2003), with more than eighty countries adopting some provisions providing citizens with the right to access information in 2006 (Banisar, 2006).

In the UK, the two pieces of legislation governing access to information are the Freedom of Information Act 2000 (2000) (England, Wales and Northern Ireland) and the Freedom of Information Act (Scotland) 2002 (2002). Under these acts, individuals and organisations are able to make requests for data held by public authorities. Importantly, applicants do not need to be citizens or residents in the UK to make a request (Lee, 2005). Public authorities include a range of organisations and institutions, including government departments, local councils, health service bodies (including the National Health Service), schools, colleges, universities, police forces and associated authorities, and the armed forces. Requests must be made in writing (email is acceptable) and authorities have a twenty working-day timeframe within which to respond, either with the requested information or with a refusal, based on one of the exemptions contained in the Acts.

Authorities must provide the requested data in full (Section 10, Freedom of Information Act 2000, 2000), and, wherever possible, in the format requested, unless it falls within one of the exemptions contained in the Act. As such, there is a positive obligation on the authority to release the data requested, and that is the starting point of the legislation. However, there are twenty-three exemptions in total, broadly categorised as ‘absolute’ and ‘non-absolute’, and these can create something of a tangled web that researchers must navigate in order to use the legislation to gain access. There are six absolute exemptions which relate primarily to information that is already available or that is prohibited by another piece of legislation, information relating to security matters and information held by courts. For these absolute exemptions, there is no requirement to apply the public benefit test; however, for the remaining seventeen non-absolute exemptions, a public benefit test must be applied, which seeks to assess the balance of the public interest for and against disclosure. If an authority refuses to release the requested information because it believes it to be exempt, it must provide the applicant with the reasons for this and inform them of the right to complain to the Information Commissioner who regulates the acts. As such, despite the positive obligation on public authorities to release information, which is by virtue considered public information, those working in FOI teams or roles act as gatekeepers to the information and the broad range of exemptions can provide a basis for refusal of the information in a multitude of situations. These issues are discussed in detail further on in this paper, with particular reference to negotiating with gatekeepers.

These Acts have proven to be a useful tool for individual citizens and journalists. According to the Ministry of Justice (2008) there were over 30,000 requests per year to central government bodies alone and the majority came from members of the general public (Brown, 2009, p. 89). An analysis of news stories in 2006 and 2007 revealed that over 1,000 stories were based on results from FOI disclosures (Campaign for Freedom, 2008 in Brown, 2009). Interestingly, many of these news stories were based on criminal justice issues, yet criminologists have been slow to utilise the act for criminal justice data or information relating to their own research. Brown (2009, p. 89) points out that many universities ‘offer training from a compliance perspective on how university staff should respond to FOI requests, but fail to provide any training on how academic researchers can use the tool to obtain information’. FOI is thus seen as something to protect against rather than as a resource.

It has been argued that social researchers use a relatively narrow range of data collection techniques (Lee, 2005) and the need to find innovative methods has been of increasing importance in social science and related disciplines. Despite the comparably widespread use of FOI in other jurisdictions, social scientists, and criminologists in particular, have not utilised FOI requests. In respect of sexual violence, no studies were found that employed FOI requests, and researchers in other fields using this approach rarely discuss it in their methodological literature, with a few exceptions mainly emerging from medical and health disciplines (Murray, 2012; Farrukh and Mayberry, 2015; for a review see Fowler et al., 2013).

The Study

Given the lack of research on sexual violence amongst older populations—in particular a lack of data on the extent or prevalence of sexual violence victimisation among people aged 60 years or over because of the CSEW age cap—the present research set out to develop a detailed study of sexual violence against people aged 60 years and over in the UK. In order to examine the incidence of sexual violence against older people, Freedom of Information (FOI) requests were used as a way of gaining information from police forces across the UK on the number of recorded serious sexual offences, specifically rape/assault by penetration, from 1 January 2009 to 31 December 2013 involving a victim aged 60 years or over at the time of the offence. These offences were selected because they are considered the most serious sexual offences, and because asking for data on all sexual offences may have made the study too broad and thus more likely to be refused under the exemptions within the FOI Act in the event that it exceeded the time/cost allowed within the Act. However, this does limit the study to these specific offences only.

As well as the number of recorded offences, demographic data relating to the victim, perpetrator and incident was requested. All of the data was pseudonymised and approval for the research was given by the School of Applied Social Sciences (Durham University) Ethics Committee. However, because of the limitations of police-recorded data (Thomson, 2010) and the high levels of under-reporting of sexual offences (Office of National Statistics, 2015), which may be higher for older groups (Blood, 2004; Burgess and Clements, 2006), interviews were also conducted with practitioners who were asked about the number of older victims they had supported over the same five-year period and their experiences of supporting older victims. Qualitative interviews were also conducted with female victim-survivors to document their experiences in their own words, to highlight not only the incidence of sexual violence but also the impacts on and support needs of older people. Thus, this study utilised mixed methods (i.e. both quantitative and qualitative) to examine a range of factors relating to sexual violence against older people.

Following a pilot with five forces to test the data collection tools, FOI requests were sent out to all forty-six forces in the UK. The requests were based on two questionnaires, which were designed with a pilot force. The first questionnaire asked about the number of recorded Section 1 and Section 2 forces between 1 January 2009 and 31 December 2013, broken down by year (i.e. 1 January 2009 to 31 December 2009) and offence type, and the proportion of those cases involving a victim aged 60 years or over at the time the offence was committed, broken down by year and offence type. The second questionnaire asked for demographic data for each case involving a victim aged 60 years or over at the time. The questionnaire asked for the following specific data: the type of offence and year it was reported; the age of the victim and perpetrator; the ethnicity of the victim and perpetrator; the relationship between the victim and perpetrator; the location of the offence; whether the crime was linked to any other crime; and whether the perpetrator was a known serial offender. For the requested data on the ages, relationship and location variables, pre-defined categories were provided; for example in the relationship variable, broad categories such as ‘partner or husband/wife’ and ‘friend’ were used. The data findings have been published in the Freedom of Information Act 2000 (2000) and the Freedom of Information Act (Scotland) 2002 (2002).

Theoretical and Methodological Approach

A mixed-method approach was particularly appropriate for this study, which sought to examine a number of issues (e.g. extent, nature, impacts, support needs and gaps in services) drawing on both feminist and gerontology theory to inform the research. It has already been highlighted that feminist research on sexual violence has tended to focus on younger women, typically aged 16 to 30 years old. Furthermore, gerontology research has examined other forms of abuse (e.g. physical, emotional, financial) under the ‘elder abuse’ paradigm, but much of this research has failed to examine sexual violence. Feminist research has therefore been criticised for not taking into account ‘age’ (Jones and Powell, 2006); it has been argued that ‘historically, academic feminism has paid scant attention to ageism, to age relationships or to old age itself (Freixas et al., 2012, p. 45), whilst gerontology has been criticised for being oblivious to gender (Whittaker, 1995). A feminist gerontology approach to this research was thus particularly appropriate, combining the central tenets of both feminist and gerontology theories. Feminist gerontology is grounded in feminist theory and critical gerontology, focussing on power relations and intersecting oppressions across the life course (Hooyman et al., 2002). Analyses that incorporate a feminist perspective broaden our understanding of ageing, moving beyond a singular focus on gender to a broader focus on multiple aspects of diversity, including privilege, inequality and interdependence (ibid., p. 3). Gender is inextricably bound up with other social inequalities such as age, race, ethnicity, sexuality and class. As Calasanti (2004, p. S306) points out, ‘each of these hierarchies comprises power relations in which the privilege of one group is tied, intentionally or not, to the oppression of another. These power relations privilege men—give them an unearned advantage—while they disadvantage women, even as people resist and reformulate seemingly “natural” gender differences and gender meanings’.

As this was a feminist study that sought to examine sexual violence against both men and women, feminist gerontology was particularly useful as a theoretical framework. As Calastani (ibid., p. S305) argues, feminist gerontology ‘challenges mainstream scholarship for not naming men as men and thus enables us to theorize gender relations: the power relations that construct the interdependent categories man and woman’. Moreover, this perspective ‘sensitizes us to other power relations such as age relations and is appropriate for studying both female and male victims of sexual violence, even though women are the majority of victims’ (ibid.).

FOI requests were particularly appropriate for researching sexual violence against older people. As well as the lack of research attention in feminist and gerontology fields, and the exclusion of older respondents in the CSEW intimate violence module, one of the major limitations of existing surveys examining abuse of older people is that they exclude respondents who are living in care/nursing homes or hospitals. This is particularly problematic as those living in such environments may be more vulnerable to experiencing sexual or other forms of violence. By using FOI requests to police forces, all reported incidences, regardless of the location of the assault, could be captured. Moreover, using FOI requests may be useful when researching issues related to old age where the population may be too ill, or deceased, to participate in interviews, as FOI allows access to data retrospectively. Similarly, where the number of interviews is likely to be small, FOI requests can assist in providing further detail or background information to support interview findings. For the present study, the data collected on the gender and age of victims and perpetrators, and relationship between victims and perpetrators were particularly useful as practitioners had seen so few cases that such variables would not have been possible to analyse.

Managing Responses

Forty-three forces provided some or all of the requested data, a response rate of 93 per cent. This is slightly higher than has been reported in other studies utilising FOI requests (Murray, 2012). One force (Scotland) did not comply on the basis of their recently integrated size and their previous recording methods, discussed below. Two forces did not comply with the FOI request, but follow-up emails via personal contacts with senior officers resulted in the data being provided directly rather than via the FOI route. Therefore, the total response rate was 45 out of 46 forces (98 per cent).

The majority of forces (36 out of 46 forces) responded fully within twenty-eight to forty days of acknowledging receipt of the request. A small number of forces emailed asking for clarification on the terms used in the request. For example, two forces asked what was meant by the question ‘was the rape (Section 1, Sexual Offences Act, 2003) or sexual assault by penetration (Section 2, Sexual Offences Act, 2003) linked to any other offence’. The forces stated that they were unsure what was meant by ‘any other offence’ and asked for examples. On reflection, this question was quite vague, but was intended to examine whether other offences, such as burglary, theft or assault offences, occurred at the same time as the sexual offence. This highlights the need to consider the terms used when drafting FOI requests and providing detail on what is meant by the question being asked. A pilot of the request was initially carried out with five forces who helped to identify potential issues with the wording and format of the request; however, issues such as the lack of clarity in this question still arose.

A number of forces initially exercised the right to refuse the request because it exceeded the time/cost allowed under Section 12 of the Freedom of Information Act. However, in these cases forces gave most of the data and left one or two variables out; for example, one force provided all of the data other than the ‘relationship between victim and perpetrator’ variable. In some cases, further discussions with the FOI team led to the information being released; as this exemption is ‘non-absolute’, the force is required to apply the public interest test. Through discussions with the individual forces about the need for this research and, on some occasions, promising to provide them with an overview of the findings, access was granted. However, a very small number of forces initially refused to provide the data altogether, or provided only the total number of recorded offences and the proportion involving a victim aged 60 years or over. For those refusing, the reasons were usually based on the Section 12 (time/cost) exemption or Section 40 (identifying an individual). For example, one force had only received one report of rape involving a victim aged 60 years or over and thus providing information on the victim and perpetrator age, relationship, ethnicity and so on may have compromised the anonymity of the victim. Refusals on Section 12 grounds were challenged successfully by highlighting that other, often larger, forces had provided the data within the time/cost allowance or by negotiating on how the data were presented. For example, some forces agreed to send the data of all the cases involving an older victim, which meant the researcher had to analyse the variables. Others asked that they be allowed to provide the requested data in more aggregated ways, for example, giving the total number of victims in each age group, the total number under each relationship category and so on. This was particularly the case where a victim or perpetrator's identity might be compromised by providing individual-level data for each case. This is considered in more detail below.

One force (Scotland) could not provide the requested data, because the previously independent Scottish forces had recently merged into a single, central force and the process of transferring all of the data from the various databases to a single database had not taken place. Thus, the information requested was not retrievable in the time allowed under the Act.

One force completely refused to provide the individual data based on the time/cost exemption, and instead provided the overall number of recorded offences but no data on the variables requested. Due to the size of the force, which was considerably larger than the other forces who had refused, they stated they could not provide the data within eighteen hours. Another force kept promising the data, but after six months had still not provided it despite numerous telephone calls and email exchanges with the FOI team. In both of these cases, the data were obtained by using existing connections and networks in the forces. This highlights the limitation of FOI in social research where the organisation can rely on the Section 12 exemption, either because of the size of the data requested or the databases/methods of retrieving the information utilised by the organisation. This can lead to inconsistency, particularly with the police where there is no single database used by all forces and therefore a lack of consistency in relation to how feasible it is to obtain the information within the eighteen-hour timeframe. Moreover, this may be a particular challenge for early-career researchers who may not have existing contacts or networks on which they can rely to get the data outside of the FOI request.

Public authorities, in this case police forces, act as ‘gatekeepers’ to what is prima facie public information but which has not previously been published to the general public. Issues associated with negotiating gatekeepers in research have been well-documented. Gatekeepers are defined as ‘individuals in an organisation who have the power to withhold access to people or situations for the purposes of research’ (Sanghera and Thapar-Björkert, 2008, p. 549). Most of the existing research discusses issues relating to gatekeepers in qualitative research, specifically in negotiating access to research participants who are often, particularly in feminist research, from marginalised groups (Heath et al., 2007; Sanghera and Thapar-Björkert, 2008). There is significantly less literature dealing with gatekeepers in quantitative research. In the present study, FOI teams acted as gatekeepers to the data being requested. Whilst in the majority of cases the data were provided within the time allocated within the Act and in full, a number of gatekeepers had to be negotiated with to provide access, particularly as the majority of refusals were based on the Section 12 time/cost exemption, which is non-absolute, thus requiring the gatekeeper to assess whether the release of information is in the public interest. In this sense, the exemptions provided both a barrier and opportunity for the researcher. Whilst initially it was a barrier that some forces relied on to refuse the disclosure of the request, the built-in public interest test allowed the researcher to discuss the need for the research and potential benefits, which in the majority of cases was enough to persuade the gatekeeper to release part or all of the data. However, this highlights the inconsistency in the way information is released under FOI requests and the limitations of the FOI Acts, which intended to remove the need to jump through hoops to gain access to public information.

FOI Requests and Feminist Ethical Considerations

Given the relatively small number of cases recorded by each police force, it was critical that the anonymity of victims was protected. One of the guiding feminist ethical principles is to ensure research does no harm to women and actively seeks to protect women's anonymity. Anonymous data were requested from the forces in the FOI requests and categories were used where possible to reduce the risk of victims or perpetrators being identifiable; for example, rather than asking for a specific age of the victim and perpetrator, forces were asked to indicate in which age group (e.g. 60 to 69 years, 70 to 79 years, and so on) victims fell and in which age group (e.g. under 16, 17 to 19, 20 to 29 years and so on) perpetrators fell. FOI requests proved useful in achieving the core feminist and social research ethical aims for a number of reasons.

First, the legislation governing freedom of information in the UK provides a number of exemptions where requested information falls outside the scope of the general duty requiring public bodies to disclose requested data. The first exemption relevant to anonymity is found under Section 40 of the Freedom of Information Act 2000 (2000) (England, Wales and Northern Ireland) and Section 38 of the Freedom of Information Act (Scotland) 2002 (2002). Under these sections, any requests for data relating to ‘personal information’ are exempt from disclosure. Thus, any personal information requested about individuals that would contravene existing data protection principles or would be likely to cause damage or distress is not considered appropriate for disclosure in FOI requests. Furthermore, under Section 30 and Section 32 of the Freedom of Information Act 2000 (2000), any information held by the authority is exempt if it has at any time been held for the purposes of an investigation that the public authority has to conduct in order to establish whether a person should be charged with an offence or whether a person charged with an offence is guilty of it. These sections therefore cover information held by criminal justice agencies. Finally, under the Freedom of Information Act 2000 (2000) and (Scotland) 2002 (2002), Section 41 and Section 36 also make exempt any information that would constitute a breach of confidence.

These exceptions and the subsequent pubic benefit test applied to non-absolute exemptions act as an internal ethical evaluation by which the public authority assesses whether the information requested is likely to harm any individual involved in a sexual violence case (which includes both the victim and the perpetrator), whether there is a risk to personal information being identified, whether this would constitute a breach of confidentiality, and finally whether the requested information is likely to damage any ongoing investigations. With these risks in mind, the public authority is then required to consider the wider potential benefits of the requested information being made public.

This process is in addition to the ethical procedures to which the majority of social researchers have to adhere. Most institutions have their own ethical boards which assess proposed research before it is conducted; similarly, most academic disciplines have their own boards which have their own codes of ethics. For example, the British Society of Criminology (2015) specifically states that researchers should strive to protect the privacy of those they research and must work within the confines of current legislation relating to privacy and confidentiality, data protection and human rights. Furthermore, their code of ethics states that researchers should not breach the duty of confidentiality and should not pass identifiable data to third parties without the participant's consent (ibid.).

Given that the fundamental underlying commitments in feminist research are to conduct research for the benefit of women that seeks to do no harm and to protect the anonymity and confidentiality of women, FOI requests can be a useful method in achieving these principles whilst also conducting research that gains access to previously unknown information, which helps to develop understandings of violence against women.

However, whilst the internal checks and exemptions contained within the Act provide additional safety measures, which can be beneficial in terms of protecting individuals and information that could cause harm to the people to which the data relates, there is also the potential for these to act as barriers for researchers seeking sensitive information held by authorities, as discussed earlier in this article. This highlights the need for researchers to be flexible when requesting information from public authorities and to be prepared to work with forces in order to find ways of disclosing data that do not compromise an individual's information.

Informing Qualitative Interviews

The use of FOI requests in this study was particularly beneficial as it helped inform the qualitative interviews which followed. The approach of combining quantitative and qualitative methods in feminist research has been encouraged by feminists (Oakley, 1999; Shapiro, Setterlund and Cragg, 2003) as it has the potential to give a more powerful voice to women's experiences (Hodgkin, 2008) and it increases the likelihood of researchers understanding what they are studying and their ability to persuade others of the credibility of the research (Reinharz, 1992). Whilst there have been some concerns about the incompatibility of qualitative and quantitative paradigms (Kidder and Fine, 1987; Bryman, 2006), a sequential approach can overcome these issues because the data collection methods involve ‘collecting data in an iterative process whereby the data collected in one phase contribute to data collected in the next’ (Driscoll et al., 2007, p. 21). This is particularly useful where there are different methods used for different elements of the study and there is a multi-research rather than mixed-methods approach.

The data gained through the FOI requests was used to inform the interview schedule with practitioners and older women victims of sexual violence in a number of ways. First, the total number of recorded offences involving an older victim was stated to practitioners at the beginning of the interview when they were asked ‘how many women/men aged 60 or over have you seen in the last five years who have come to see you about a recent (not historic) rape or serious sexual assault?’. Most of the practitioners’ organisations had seen very few people in this age group disclosing recent sexual violence, and this figure was useful in making comparisons between the police recorded number and the number held by support organisations. This figure was also useful for asking practitioners ‘do you think older people are more or less likely than younger people to access rape support services?’, as it allowed practitioners to use their own experiences of working with older people in comparison to the number reporting abuse over the same period to their local police force (figures from the force area in the same geographical area as the support organisation were cited to help practitioners contextualise the number of older victims they had supported compared to the number reporting to the police).

The demographic information gained from the FOI requests was also useful in informing the interviews and drawing comparisons between the recorded offences and the older victims that practitioners had seen. For example, sixty-seven (12 per cent) of cases recorded by the police involved rapes or assault by penetration offences perpetrated by carers, whereas only a couple of practitioners had experience of supporting a victim who had been raped by a carer. Similarly, whilst the second more common relationship in reported cases involved a husband or partner perpetrator (20 per cent), practitioners had very rarely supported women who had experienced this. Instead, the majority of women coming forward for support had been assaulted by acquaintances or strangers. It would therefore appear that the older people reporting sexual violence to the police are not necessarily the same people accessing support services. The data gained through the FOI requests has helped to identify a gap; whilst the majority of practitioners stated in the interviews that they felt older people would be less likely to access rape support services than younger victims, all were surprised by the number of recorded crimes compared to the number of victims they had seen over the same period. As a result, several practitioners said that they would explore how they could increase engagement in their services by older people. This links to another commitment of feminist research: conducting research on women, for women, which has a practical impact on the issue being explored by influencing policy and/or practice developments (Sarantakos, 2012).

FOI requests can therefore serve as a useful tool to uncover initial information about an unknown phenomenon which can then be used to inform interview schedules. Without understanding the demographics and contexts of victims who had reported to the police, it would have been impossible to contextualise the experiences of victims who had accessed support agencies. Furthermore, the data collected through FOI requests highlighted some of the methodological weaknesses in interview data, particularly when researching marginalised groups. The small number of older people accessing the sexual violence organisations (typically less than five over the previous five-year period) would have made it challenging to draw out any themes; however, the quantitative data allowed comparisons to be drawn between the 655 cases reported to the police and the one-hundred or so described by practitioners. Practitioners admitted having little experience of supporting older people and several asked for the information gathered from the FOI requests to use in future campaigns or funding bids. However, the depth of the data on the short and long-term impacts of sexual violence on older people, and the challenges they face in accessing support and reporting to the police, would not have been possible through an FOI request; much of this information came from the open-ended questions asked during interviews and the subsequent probing to uncover further detail. Thus, both the data gained from FOI requests and those gained from the interviews provided useful insights into different factors relating to sexual violence against older people, which highlights the benefits of using mixed-method approaches to provide the ‘bigger picture’ and ‘personal story’.

Moving beyond Research: FOI Requests and Feminist Activism

The potential for FOI requests stretches beyond academic research to feminist support organisations and activists, yet both academics and feminist organisations have been slow to take advantage. However, the potential has already been realised by other campaigners; for example, activists working in the area of animal rights have used requests to force universities to reveal details about animal experiments (Adam, 2015). This has then been used to campaign for better rights for animals and to lobby for policy change. At the same time, there are limitations on the organisations defined as ‘public’ for the purposes of FOI legislation; charities, third sector organisations and private organisations are not usually subject to the duty to provide data. Thus, information held by private organisations who have been contracted out by public authorities may not fall within the scope of the act and researchers, campaigners and citizens wishing to access data should be mindful of these restrictions. Furthermore, citizens who do obtain information from a public authority in any context should be aware that the information becomes ‘public’ once it is released to the applicant and often the response from the authority is published online. Thus, any information provided to the applicant is usually made available to general members of the public who can use it for their own purposes.

Nonetheless, this research demonstrates how FOI requests can help develop policy and practice. The data collected in this research have already been shared with sexual violence and age-related organisations and have helped to initiate relationships between these two organisations at a local level. Moreover, the data are being used by organisations in current funding bids for local projects.

Conclusion

The use of FOI requests in academic research has been under-recognised. Despite recent publications highlighting the potential usefulness of FOI requests (Brown, 2009; Savage and Hyde, 2014) and commenting on the lack of academic utilisation, research adopting FOI requests as a methodological tool remains scarce. Furthermore, to date there is no published feminist research that has adopted FOI requests as a method.

This reflective paper has highlighted some of the strengths of using FOI requests as part of a mixed-method feminist study. In particular, the access to data held by the forces would have been a much more difficult process without the use of FOI. Furthermore, as the group being researched may be considered hard to reach—due to a lack of existing engagement with sexual violence services and an increased chance that they will be living in institutions such as care homes which are not routinely included in research—FOI requests are a tool for gathering data on such populations. Accessing this data enabled the research to get a national picture of recorded serious sexual offences involving an older victim, and this was particularly useful for informing the qualitative interviews which followed in the second and third phase of the study. The benefits of this sequential approach to mixing methods has been highlighted elsewhere (e.g. Driscoll et al., 2007) and it proved to be a very useful strategy in this present study.

Design of requests is critical. A lack of clarity around the data being requested or terms used can cause delays in the data being provided or, even worse, can result in refusals to provide data or the wrong data being provided. In particular, it is important that researchers are aware of the language and terminology used by the authority from whom they are requesting data. For feminists, this may go against the terminology they would usually use; however, in this context it is important to be consistent with the language used by the authority. It is therefore recommended that those seeking to use FOI requests gain feedback on the requests at the design stage, ideally by a willing public authority who falls within the class of those who will receive the final request and can provide comments and suggestions and mitigate potential issues. Similarly, those working in university FOI departments may be able to provide guidance to academics and researchers on the design.

The additional scrutiny required by law before any data by public authorities can be released under an FOI request, provides further ethical protection and is beneficial for researchers who seek to adhere to basic ethical principles of doing no harm and ensuring the confidentiality of those who are being researched. For feminist researchers who are often researching sensitive topics such as domestic and sexual violence, this may be particularly helpful in supporting ethics applications made to their institution's board and for ensuring the research protects the anonymity of those being researched.

Whilst the data held by public authorities are prima facie public information, FOI does not provide unrestricted access to such information and researchers may find they have to negotiate with gatekeepers, as in other research projects, for the information to be released. The exemptions within the Act can create something of a tangled web that researchers must navigate. However, the public interest test that public authorities must apply to the majority of exemptions when considering whether to refuse a request, provides the researcher with an opportunity to work with the authority to emphasise the importance of the research and to communicate why it may be of public interest. If requests are refused, there is an appeals process, but this can be lengthy and there is no guarantee an appeal will be found in the researcher's favour. Therefore, as noted elsewhere (Lee, 2005; Savage and Hyde, 2014), it may be unfeasible or undesirable for researchers to appeal, particularly if they are on tight timescales for research projects. Being aware that some requests may be refused, in part or full, and building time into a project to allow for this at the initial development stages may be useful for researchers seeking high response rates.