Building a Research Partnership in a Prison Context: From Collaboration to Co-Construction

Introduction

1.1 This paper is about the use of a participatory research design to investigate prisoners' participation in prison programmes. The positive associations related to participation in prison programmes are increasingly recognised as sources for improving mental and physical health (Nelson et al. 2006), reducing recidivism (Kim & Clark 2013), and contributing to self-esteem (Coyle 2009). In Flanders (Belgium), there is a 'Decree on the organisation of care and services for prisoners' that states that high-quality programmes must be offered in each prison regarding culture, education, health, sports, vocational training, and wellbeing (Flemish Government 2013). Despite the attention given to the positive associations related to participation, there is less academic research on factors that might lead to greater participation (Hall & Killacky 2008; Rose 2004). In addition, the activity providers at a remand prison in Flanders reported that they lacked knowledge about the participation profile and needs of their target group. The activity providers asked for help from our research department to examine the profile of participants and non-participants in prison programmes and prisoners' motivations and/or the barriers involved.

1.2 Prison research is often a privilege granted to very few academics (Quraishi 2008) and is accompanied by different practical, methodological, ethical (Martin 2000) and emotional challenges (Jewkes 2012; Liebling 2014). Therefore, it is essential that researchers have the possibility of reflecting their concerns, struggles, frustrations and challenges during their research process (Beyens 2013). Some academics have focused on the practice and challenges of conducting ethnographic research (e.g. difficulties in getting access – Waldram 2009; emotional responses to conducting research in a prison setting – Jewkes 2012; competing perspectives – Liebling 2001; establishing trust among prisoners – Waldram 2009). Conversely, there are seldom descriptions and reflections of research settings that rely on quantitative methods (Sutton 2011) or on a combination of quantitative and qualitative methods. In response to this gap, we aim to describe and analyse our experiences in setting up a participatory qualitative and quantitative research in a prison context.

1.3 Recently, the added value of collaborative projects (e.g., Apa et al. 2012; Cislo & Trestman 2013) and participatory (action) research in prison has been acknowledged (Burdon et al. 2011; Fine & Torre 2006; Torre & Fine 2005; Ward & Bailey 2013; Walsh et al. 2014). Jagosh et al. (2012: 312) consider participatory research as "the co-construction of research through partnerships between researchers and people affected by and/ or responsible for action on the issues under study". A researcher can act as a facilitator in bringing together multiple stakeholders (Mackenzie et al. 2012), stimulate them to develop their own understanding of the research process and the subjects investigated, and consequently take actions to improve their practices (Clark et al. 2009). This means that the aim and research questions are developed out of the convergence of the perspectives of practitioners and academics (Bergold & Thomas 2012). They are both involved in the process of decision-making and conduction of the research (Bourke 2009). Through a collaborative research partnership, the input of institutional stakeholders, such as prisoners, correctional officers and activity providers can be included in all aspects of the research. Nevertheless, for many academics and correctional institutions, this is a new model that will require commitment to implement (Gostin et al. 2007).

1.4 In order to get insight into the specific characteristics of this research method, the challenges associated with conducting participatory research are discussed. According to Cargo and Mercer (2008), all research projects progress through three phases: (1) shaping the purpose and scope of the research, (2) research implementation and context, and (3) interpretation and application of the research outcomes. They acknowledge that a participatory research methodology may have an added value for all the research phases (Cargo & Mercer 2008). Our purpose in what follows is to elucidate the use of a partnership in these different research phases. The article concludes by reflecting on the factors that contributed to the success of the research, and the ethical, emotional and policy challenges raised by this study.

Challenges associated with participatory research

2.1 The literature on participatory research reveals critique on the concept of participation (Braye & McDonnell 2012), and relates this to the power dynamics between researchers and participants (e.g., Braye & McDonnell 2012; Lyon et al. 2010). The levels of participation of participants and co-researchers may vary significantly throughout 'participatory studies' (Kindon et al. 2007). According to Cornwall and Jewkes (1995), the degree of participation – and thus also the amount of power members of the partnership have – is dependent on the degree of control over the process held by the researchers. Various academics have tried to deal with the issue of power relations by identifying different levels of participation.

2.2 The first typology is the 'ladder of citizen participation', developed by Arnstein (1969). He states 'there is a critical difference between going through the empty ritual of participation and having the real power needed to affect the outcome of the process' (1969: 216). The involvement of participants ranges over eight levels, each level indicating the degree of power citizens have in determining the end product. The two bottom rungs are manipulation and therapy, indicating levels of 'non-participation'. The next three rungs are called 'levels of tokenism': informing, consultation and placation. Citizens are informed, consulted or asked to provide advice, but the power holders retain the right to take the decisions. The last three rungs are degrees of citizen power: partnership, delegated power and citizen control. At the level of partnership, citizens negotiate and engage with power holders, while at the other two levels the citizens have full control over the process of decision-making (Arnstein 1969).

2.3 Participatory research can be situated within Arnstein's ladder of participation (Krieger et al. 2002; Washington 2004). Washington (2004) argues that taking part in a participatory research project falls on rung six of Arnstein's ladder: partnership. Conversely, Krieger and colleagues (2002) do not support the 'hierarchical' ladder because it assumes that more participation is always better. Instead, they advocate a horizontal spectrum of participation. This admits the possibility for members of a research partnership to choose the degree of participation that is best suited to the project. Moreover, some academics state that participation and control are never static; research projects can start with a university-driven agenda, but move towards a community-driven agenda (Wallerstein & Duran 2006). The participatory mechanisms present in the spectrum of participation range from providing information and input to facilitating shared influence over the project. The challenge is to equalize power imbalances between the more powerful (in our case the researchers and the activity providers that are part of the steering committee) and the less influential people (in our project prisoners). Information gathered by, for instance, focus group interviews can be very influential in shaping the research project, but this depends on the commitment and goodwill of the researchers to take their results into account (Krieger et al. 2002).

2.4 We adopted the point of view of Krieger et al. (2002). Different groups of stakeholders can all have their own influence on the research project and take a different position in a wide spectrum of participation. The different groups of stakeholders in our research project were activity providers, prisoners, prison management, prison guards and researchers. We have to make a distinction between those having influence on or power over the content or the practical aspects of the research. The first two groups had in particular an influence on the content of the project; the prison management and prison guards on the practical side.

2.5 The members of the steering committee had the greatest power in determining the process and end product and for them, taking part in the research project falls on rung six of the ladder of citizen participation: partnership (Arnstein 1969), which means that all the members had equal power to make decisions (Washington 2004). Through negotiation, the members of the steering committee agreed to share planning and decision-making. First, the ground rules had to be established through some form of give-and-take, but afterwards no unilateral changes could be made (Arnstein 1969). The activity providers approached the researchers with the request to set up a research project, which entailed that the research agenda was initially community-driven. Their goal was to get insight into the profile of prisoners that (did not) take part in prison activities, and their related motivations or barriers. It was the task of the researchers to translate the activity providers' needs and wishes into research questions and to propose research methodologies by which the questions could be answered. The ultimate decisions about the methodology, but also data analysis, discussion of the results and formulation of recommendations were taken during the meetings of the steering committee. The members of this committee played thus an active and crucial role in the knowledge construction.

2.6 A second group of stakeholders were prisoners. This group can be situated on the left side of the spectrum of participation, indicating that they were central figures in providing information, giving feedback, etc. during the whole research process. For instance, as is acknowledged by Krieger et al. (2002), the information that is gathered through the focus groups with prisoners was very influential in shaping the standardised questionnaire. If we make a comparison between the contribution of members of the steering committee and prisoners, we can state that the steering committee had more power over the development of the research project. We have tried to involve prisoners as much as possible and to give them power, but their agency was impeded by the structures of the correctional institution. Prisoners were not officially allowed to be part of the steering committee by prison management, due to the fact that the research was conducted in a remand prison (i.e. a fast changing population due to transfers to other prisons and releases from prison) with enormous overcrowding (an average prison population of 164.4% in 2012 – FOD Justice 2013), both leading to serious safety concerns.

2.7 The last two groups of stakeholders, namely prison management and prison guards, had in particular a practical influence on the project. More information about their power can be found in the paragraph 'communication with stakeholders'.

Shaping the purpose and scope

The research request

3.1 The request to set up a research project came from the activity providers of a remand prison and was of interest for several reasons. First, our research department has a tradition of doing research on the social, educational and cultural participation of vulnerable and disadvantaged groups. Although prisoners are recognised as a socially excluded group (Murray 2007; Social Exclusion Unit 2002), our research group so far had no experience with doing research inside a prison. Because we lacked this experience, we needed to be briefed about security awareness and the need for institutional rules and regulations (Peternelj-Taylor 2005). Second, as we have a tradition of setting up participatory research projects, it was essential that the research idea originated from the practical field. This not only created the opportunity to establish collaboration between the university and activity providers, but also to evolve to co-construction of knowledge. Co-construction goes a step further than collaboration. Previous research has shown that collaboration with employees of the setting wherein the research takes place enhances the recruitment of research participants (e.g., Ross et al. 2010; Sullivan-Bolyia et al. 2007), while co-creation and co-construction is more about the participation of academics and non-academics together in the different phases of the research project (Jagosh et al. 2012). Although some participatory (action) researchers assume that community members (i.e. people affected by the study) are willing and able to be engaged in a research project, not everyone is qualified or sufficiently interested to become involved (Arieli et al. 2009). This was no barrier in our study, as the proposal to conduct the research originated from the practical field. This opened the opportunity for building a research partnership and co-constructing the project.

Forming a partnership

3.2 Academics and non-academics can enrich each other while setting up the research project (Cargo & Mercer 2008). To co-construct the research, we started a continuous cooperation in 2011 between activity providers and a scientific team by forming a steering committee. The steering committee met bimonthly and was responsible for the coordination and decisions of the whole research process. Members of the steering committee were representatives of each sector offering activities in the respective prison (i.e. culture, education, health, sports, vocational training, and wellbeing), the coordinator of the activities and two members of the scientific team. Although other research projects demonstrated the added value of involving (ex-)prisoners (e.g. Fine & Torre 2006; Walsh et al. 2014), we made the decision not to include them in the steering group. This was a conscious choice as the reason for this research is to provide activity providers with insight into the participation of prisoners in prison activities. Based on the findings, they can (if necessary and possible) adapt their offer and anticipate the reasons that motivate prisoners to take part, or the barriers that hinder prisoners' participation in prison activities. Nevertheless, prisoners were central figures in providing information and giving feedback during all the research phases.

3.3 Another group that was not involved in the steering committee were prison guards. The rationale behind this was that prison guards have different philosophical perspectives, such as security and safety (Walsh et al. 2014). Nevertheless, prisoners and prison guards were important groups of stakeholders and interaction with them was of utmost importance (see section communication with stakeholders), but not as a member of the steering group. Also, the prison manager was not included in the steering group, but nevertheless was briefed about the entire research design on several occasions. Notwithstanding, some academics possibly state that 'involving' the prison management gives them influence over the research design, delivery and dissemination of the results, and that this contradicts with the autonomy and independence of academic intellectual production, we found this consultation useful. As time passed, a sense of co-ownership was created. The prison manager was especially concerned about the practical aspects of the research project: giving permission to conduct the research, providing access to the institution, allowing for recording material to be brought in, the influence of the research on the daily running of the prison and security issues. The prison manager was an important stakeholder, without having an influence on the academic development of the research. Besides, to harmonise the research and the prison routine, good agreements had to be made with the prison guards (Martin 2000) (see section communication with stakeholders).

Obtaining insight into the various expectations and needs

3.4 Participation, or forming a partnership, is something that needs to be developed. It requires a time investment to discuss the whole research project while taking into account the needs of all partners to create a supportive context (Lister et al. 2003). It was important to get everyone to adopt the same line about the research aims and methodology. Also, having insight into the different (and possibly competing) practical and policy agendas was necessary. To achieve a comprehensive understanding of the research goals that each of the different partners wanted to obtain, we conducted six in-depth interviews with each of the six sectors represented in the steering committee. The aim of these interviews was to obtain insight into the various perspectives of the partners, to understand what the main research needs were, to discover how we could match those different concerns and questions, and to achieve a joint, supported research goal entailing benefits for each. Based on a joint analysis of these in-depth interviews, the steering committee decided to set up a quantitative research because the activity providers needed figures to apply for projects and influence policy. However, the quantitative research phase was preceded by focus groups interviews. The choice to conduct a quantitative research might be surprising as participatory approaches outside prison mostly rely on qualitative data collection methods (Kidd & Kral 2005). Nevertheless, there are some examples of prison research that use both qualitative and qualitative research methods (e.g., Martin et al. 2009; Torre & Fine 2005). We also made a combination of these two. Participatory research can rely on various methods, the most important is that the knowledge is created in collaboration (Fine et al. 2003), e.g., joint development of questionnaire, joint data-analysis, formulation of conclusions, etc.

3.5 The overall aim of the research project was to derive knowledge about participation of prisoners in correctional activities^[1]. The steering group formulated four research questions:

• What is the profile of the prisoners in Antwerp prison (e.g., socio-demographic characteristics, socio-economic characteristics, command of the Dutch language, factors related to imprisonment)?
• How frequently do prisoners participate in correctional programmes (e.g., culture, education, mental health care, sports, and vocational training)? What motivates or impede prisoners to participate?
• What is the profile of the (non-)participants?
• What needs do prisoners have concerning their reintegration into society?

3.6 It was clear that the success of the research would depend on the response rate. A response rate of more than 50%% among prisoners is high (Dyb 2009). As this was also a clear concern of the steering group, we focused on maximising the response rate and reaching as many prisoners as possible. Most of the prisoners must be given the opportunity to participate in order to make it possible to generalise the findings.

Research context and data collection

4.1 The second phase concerns the implementation of the research. The added value of a collaborative research partnership during this research phase is acknowledged, for instance in increasing recruitment (Cargo & Mercer 2008; Jagosh et al. 2012) and enhancing the research to suit the context (Cargo & Mercer 2008). The data collection was spread over one year in three steps. The first step consisted of the development of a standardised questionnaire. Step two concerned the testing of this questionnaire, and step three involved conducting the standardised survey.

Development of the questionnaire

4.2 The first step consisted of the development of a questionnaire. It took six months of literature review, exploration, qualitative preliminary research, discussion and alterations to finalise the questionnaire.

4.3 The qualitative preliminary research included focus group interviews aimed to gather information that could be used in the development of a standardised questionnaire. In total, we conducted 13 focus groups that represented a wide range of views and profiles (N=86). Table 1 provides an overview of who took part and the number of participants of each focus group.

Table 1. Overview respondents of focus groups

4.4 A similar topic list was used for all respondent groups, including issues such as the reasons why prisoners participate in activities, positive and negative experiences related to participation, and barriers why prisoners do not participate. All focus groups lasted about two hours and were audiotaped and transcribed verbatim. They were coded and analysed by thematic content analysis (Lapan et al. 2011) using MAX.QDA 2007, a qualitative analysis programme. The analysis formed the basis of the standardised questionnaire. After the analysis, the steering committee selected and delineated the themes, and determined the order of the questions.

Testing the questionnaire

4.5 Since the average educational level of inmates is lower than the national average (de Maeyer 2005), it was necessary to develop an accessible and user-friendly questionnaire. To do so, first specialists in clear linguistic usage checked the questionnaire. Second, as previous research indicated that it is worthwhile to conduct a pilot study (Martin 2000), the questionnaire was piloted in Dutch, English and French among 34 male prisoners, male internees, and female prisoners and internees. During and after the prisoners had completed the questionnaire, they were asked to reflect on the user-friendliness and the content of the questionnaire. This piloting necessitated some amendments to the questionnaire lay out, the structure and the wording of the questions. The piloting also provided information about the feasibility of the study, time needed, and about the workload of the data collection for the prison guards.

4.6 Because of the large number of foreign inmates in Belgian prisons (Snacken 2007), it was important to anticipate the possible language barriers (Slotboom et al. 2011). Consequently, after the pilot testing official translators translated the questionnaire into 12 languages: Albanian, Arabic, English, Farsi, French, German, Italian, Polish, Romanian, Russian, Spanish, and Turkish. Together with the original language (Dutch), the questionnaire was available in 13 languages.

The survey phase

4.7 To conduct the survey, the steering committee developed a detailed step-by-step plan. This plan included how to interact with a variety of stakeholders who could influence the research process and a detailed scenario for the data collection.

Communication with stakeholders

4.8 To implement the research, the steering committee needed to interact with a variety of stakeholders who all had their own influences on the research project: management, prison staff and prisoners. Although support of all of these stakeholders was of vital importance for the success of the research (Freudenberg 2007; Gostin et al. 2007), identifying and gathering stakeholders is often a step that is overlooked (Cislo & Trestman 2013).

4.9 One of the main difficulties of doing research in prison is getting access to the prison (Dolan et al. 2007; James 2013; Martin 2000). Consequently, the prison manager was an important stakeholder because he needed to give permission to conduct the research in his institution. During a meeting the researchers, the coordinator of the activities and the prison manager discussed the whole research procedure. Maintaining security was the main concern of this last actor. Conducting research in the institution would increase the workload of the prison guards because they had to transport the prisoners, escort the researchers, etc. (Cislo & Trestman 2013; Wakai et al. 2009). Based on that concern, the steering group took decisions about the practical side of the research – while keeping the workload of the guards to a minimum – and ultimately gained permission to carry out the research.

4.10 Second, cooperation with prison guards was crucial because they were the only people who could bring the respondents to the researchers (Freudenberg 2007). They can be considered as 'gatekeepers' (James 2013). The steering committee thought that it would be a challenge to persuade the prison guards of the general interest of the research. They retrieved the timetable from the prison manager who had to work during the days of the data collection. The activity providers started to inform the prison guards two weeks in advance. Every prison guard was approached personally and received an explanation about the research. This personal approach was appreciated by most of the guards. In addition, the fact that the data collection on the male sections took place on a weekend was appreciated, because of the fact that we took their workload into account. Still, staff varied in their reactions towards our presence. Some were very helpful and interested in the project, others (a minority) were troubled by our presence because they had to manage the research (e.g. a lot of additional movements, opening and closing cells, etc.) in addition to their regular tasks.

4.11 Third, informing the target group of the research – prisoners – was essential to increase their willingness to participate and to increase the response rate. An initial method to spread information about the research was through the distribution of flyers and posters. Two weeks before the research took place, all prisoners received a personal flyer in their cell. Research has demonstrated that sending letters of invitation and subsequently a personal request to partake in the research is effective (De Donder et al. 2014). In addition, information was distributed through posters at locations where prisoners passed through: the library, the doctor's waiting room, the fitness room, classrooms, etc. These flyers and posters were circulated in the 13 languages in which the questionnaire was available. In addition, some activity providers developed a promotional video that was shown on the prison TV channel repeatedly during the weeks before the data collection. The activity providers also gave more information about the research during activities.

Conducting the survey

4.12 The next step in the project was administering the standardised questionnaires. In this phase, the prison population was divided into three subgroups: male prisoners (N=432), male internees (N=21) and females (N=54). Although the study aimed to question the whole prison population, not all prisoners were able to participate. For instance, prisoners who were locked up in an isolation cell, staying in a hospital, under a special security regime, had semi-liberty status, or were suffering depression could not take part. All of the other prisoners were personally invited to be involved in the study. In total, 507 respondents agreed to participate in the survey (response rate = 73.37%). Table 2 provides more information about the characteristics of the respondents.

Table 2. Overview of characteristics of respondents

4.13 The data collection on the male sections took place during a weekend. On both days, ten members of the steering committee who were responsible for the activities, six members of the university and approximately six volunteers (i.e. activity providers who were not members of the steering committee) helped to collect data. All collaborators of the study received the detailed scenario and were informed and trained about the practical aspects of the data collection during the prior week. The male internees section and the female section were questioned during a quiet weekday (i.e. few planned activities). Two members of the steering committee and one researcher conducted the study on these days.

4.14 Academics emphasise the need to inform prisoners about the research and to underline its voluntary nature (Martin 2000). Likewise, the focus group interviews revealed the importance of a personal approach. Prisoners indicated that they were more likely to participate if they were asked personally compared to when they received a flyer in their cell. Consequently, a prison guard, accompanied by an activity provider, opened the cell doors. Every prisoner was given general information about the research and personally invited to participate. The activity provider emphasised that participation was completely voluntary and encouraged prisoners by stressing that it was important to know their opinions to improve the activities offered according to their needs. By doing so, the activity providers presented an open attitude and had respect for their life circumstances (e.g., overcrowding and limited visiting opportunities). During the explanation of the research, the activity providers also emphasised that the questionnaire was available in 13 languages. To reach prisoners who did not understand Dutch, French or English (these are the languages that most of the activity providers spoke), they used a paper with a short description of the research in all 13 languages.

4.15 In the morning, some prisoners refused to take part because they still wanted to sleep. As Sutton (2011) states, waking up inmates can be seen as a negative effect of our research. To increase the response rate, the activity providers approached these prisoners again during the afternoon, and many prisoners decided to participate at that time. Prisoners were never coerced to become involved in the study, and if a prisoner refused participation after the second request, we did not ask again. Prisoners who had already participated were encouraged to introduce the research to their cellmates and to stimulate them to fill in the questionnaire. As was demonstrated by Martin (2000), prisoners who were recruited by 'friends' were more likely to participate. However, we acknowledge that coercion is an ethical debate and that some academics will probably state that asking prisoners for a second time or by their cellmates is a form of coercion. When a prisoner completed the questionnaire, they received a ball pen as a form of recognition of their willingness to take part. The majority of the respondents appreciated this gesture.

4.16 Prisoners were brought together in a classroom in small groups with a maximum of 16 persons. The questionnaire was meant to be self-administered but the respondents had the opportunity to ask additional questions regarding unclear questions. Furthermore, prisoners with reading and writing difficulties could receive assistance from a researcher or a volunteer to complete the questionnaire, which also allowed less literate persons to participate. The assistance given to each participant was different depending on the difficulties experienced, but overall we provided assistance to many prisoners. The questionnaire was read aloud verbatim to some prisoners, while for others only a few words were explained. We acknowledge that some academics may pose questions about the power relationship between the prisoner and the helper. Nevertheless, in previous prison research the method of reading aloud the questionnaire has also been used to tackle literacy problems (e.g., Buchanan et al. 2008; Unver et al. 2013).

4.17 The volunteers were trained not to influence respondents' answers in any way. They encouraged prisoners to give the real answers, and not to portray them in a generally favourable fashion. While completing the survey, prisoners were also allowed to help each other and clarify the meaning of the questions. In particular, prisoners who spoke a language that the helpers did not could refer questions to fellow-prisoners that spoke the same language and Dutch/French or English. Observation of their body language made clear that the respondents were seriously discussing the survey. If the fellow-prisoner was not able to answer the question(s), they could ask a researcher or volunteer for help. If prisoners preferred to fill in the questionnaire in their cell this was also allowed. Nevertheless, most of the respondents went to the classroom. For the data collection on the male internees section, we worked together with the specialised care team because this team had closer contact with the internees than the activity providers.

Interpretation and application of research outcomes

5.1 The last research phase was the interpretation and application of the research outcomes. In this phase, the use of participatory research had advantages for several reasons. First, both academic and non-academic partners enriched each other in interpreting the research results (Cargo & Mercer 2008). This increased the quality of the outputs and outcomes (Jagosh et al. 2012). These interpretations were reflected in a research report. The steering group discussed the content, structure, layout, etc. To give an example, in order to demonstrate the role of the steering group in shaping interpretations of the results: a member asked if it was possible to develop a typology of (non-) participants based on the barriers. Thereupon, we developed this typology of (non-) participants and discussed this in the steering group. Was this what they intended? Was it useful for practice? What factors should be added to the typology to increase the value for practice?

5.2 Another advantage was the potential for wider dissemination and translation of the research findings (Cargo & Mercer 2008) to make the project sustainable over time (Jagosh et al. 2012). The steering committee organised a colloquium to present the results to a wider public, including staff of other prisons, policy makers, and academics. During this colloquium, the research report was distributed. Currently, interested people can download the report. Furthermore, all of the members of the steering committee can disseminate the results, but academic investigators reserve the right of scientific dissemination.

5.3 The steering committee was also challenged to formulate recommendations based on the analyses and conclusions. This was also done in a participatory manner. The members of the steering committee were asked to respond to the following questions independently: (1) What are the three most important results/ conclusions of the research related to your work in prison? (2) What recommendations would you give for the organisations you are working for? (3) What recommendations would you give for the prison of Antwerp? (4) What policy recommendations would you give for the national government? Afterwards, their answers were discussed during a steering committee and formed the basis for the formulation of the ultimate recommendations. In addition, policy makers and activity providers were provided with options for making possible changes, adjustments or improvements.

Challenges associated with prison research

6.1 Different academics state that doing research in prison is often accompanied by practical, methodological, ethical (Martin 2000) and emotional challenges (Jewkes 2012; Liebling 2014). In the previous paragraphs we have discussed the practical and methodological challenges of setting up a participatory research project in a correctional institution. In this paragraph, we focus on the ethical and emotional challenges that emerged during our research.

Ethical challenges

6.2 It is important to pay attention to the ethical considerations of doing research in a prison setting (Gostin et al. 2007). The research protocol, which included information about the research goals, data collection methods, the sample, period wherein the research took place and the methods of data analysis, was approved by the Ethical Committee of the university. An important ethical consideration in the protocol was confidentiality. We made an agreement with the respondents (both during the qualitative and the quantitative phases of the research) that anonymity was guaranteed, and consequently, all of the personal details, such as names and places were changed. Furthermore, we reassured the prisoners that the information provided would not influence their legal file nor their trial (if they still had to appear in court).

6.3 A second ethical consideration was that prisoners needed to be treated as autonomous people who may not be coerced or unduly encouraged to participate in a research project (Gostin et al. 2007). Despite the fact that some prisoners were invited to participate for a second time (because the first time they were still sleeping), we feel we never coerced them. We respected their choice if they did not want to participate. As the role of the researcher is to ensure that potential participants understand the risks and benefits associated with their participation in the research project (Peternelj-Taylor 2005), we also gave the prisoners the opportunity to ask for additional information about the research from the activity provider who came to their cell before they decided whether to participate.

6.4 A third point of concern was the ownership of the data. The data collection was a result of a close cooperation between different partners. We made an agreement that in communication about the research the activity providers, the prison manager, and the university would emphasise that the research is a result of a collaborative research project. The research method, the questionnaire and the results are the intellectual property of the research partnership, and every member can disseminate the results separately. For instance, a service provider can use the evidence derived from the research to improve practice and to apply for funding to ameliorate or implement activities. The university can use the data for academic purposes and to write research papers about, for instance, the profiles of those who did (not) participate, and what motivators and barriers prisoners face.

6.5 It was also important to inform the participants (prisoners) about the results of the research (Lister et al. 2003). Because the study took place in a remand prison, it was difficult to disseminate the results to the participants because the majority of the respondents were transferred to other prisons or released. However, the steering committee had given some prisoners the possibility to reflect on the research, and these reflections were recorded. In two movies, prisoners reflected on the reasons why prisoners take part in prison programmes. The other movies were about the barriers that impede participation and the needs they have concerning re-integration. The short movies were shown during the colloquium. Furthermore, it was possible to make the research report and the newspaper articles available for them in the prison library.

Emotional challenges

6.6 Entering the field of prison research not only poses practical, methodological and ethical challenges, but also raises different emotions (Jewkes 2012; Liebling 2014), as it is almost impossible to conduct research in a human environment, and more particularly in prison, without subjective feelings (Liebling 1999). The first emotion was unlocked by getting the request to set up a research project in prison. None of the researchers had been into a prison before and we did not know what to expect. After our first meeting inside the prison we agreed that it was an overwhelming experience. We had to hand in our passports, leave our possessions in a locker, pass the metal detector, and wear a visible badge at all times. There was no door that we could walk through without the permission of a prison guard, there was a specific smell and it was noisy. After we had been in the prison several times we were used to the customary procedures. Nevertheless, during the rest of the project we were frequently confronted with our emotions. Throughout the survey phase we helped a lot of prisoners to fill in the questionnaire and some of them talked about their life and disadvantaged situation (being unemployed for a long time and not able to find a job, do not see their children anymore, having an addiction, being in prison several times, etc.). Some of them asked if they could write something about their personal life at the back of the questionnaire. Afterwards, we scanned all these comments into a document and sent it to the activity providers.

Concluding remarks

7.1 The importance of partnerships in conducting research in a prison setting is increasingly acknowledged; however, for many academics and correctional institutions, it remains a new and innovative field. We consider that the strength of our research project is to describe to the collaborative research methodology and the strong partnership between a scientific team, the activity providers and prison management. In the partnership, we obtained a broad range of viewpoints, knowledge and interpretations, and together, we were able to establish the research agenda, handle access problems, implement and conduct the research project and achieve a high response rate. One of the lessons learnt from other academics is that anticipating low response rates, which are common for research in correctional facilities, is important (Fox et al. 2011).

7.2 We think that some of our attempts to increase the response rate are transferable across institutional and national contexts. First, the availability of the questionnaire in 13 languages possibly increased the response rate because we anticipated and could overcome language barriers. Furthermore, we detected three other factors that contributed to this high response rate and thus indirectly affected the success of the research. First, prisoners were personally approached by an activity provider and invited to be involved in the research project. When prisoners decided to participate, they could obtain assistance from one of the more than 20 volunteers while filling in the questionnaire. Through this assistance, prisoners with reading and writing problems could also participate.

7.3 Another factor that was essential for the success of the research was the awareness of different groups of stakeholders who could all have an influence on the research process, such as the prison manager, prison guards and prisoners. The steering group tried to approach every stakeholder in person to inform them about the research, which resulted in a greater appreciation of the project by most of them.

7.4 A third factor that contributed to the success of the research was the detailed scenario. Everyone who provided help during the days of the data collection (the volunteers, members of the steering committee, prison guards, etc.) received a scenario. Consequently, everyone was informed about the timetable and the division of tasks.

7.5 Besides the high response rate, there are other academic benefits arising from this research. The process of shaping the research questions together with the steering committee resulted in the development of questions other than those initially thought of by the researchers. To achieve a comprehensive understanding of the research goals, in-depth interviews with each sector presented in the steering committee were conducted. During these interviews, it became clear that they wanted to obtain insight into the factors that motivate prisoners to participate in prison activities, but foremost also into the barriers that impede prisoners' participation. The focus on barriers was novel in research, as a systematic review revealed that only 10 out of 22 studies addressed this topic. We can conclude that the literature concerning participation in correctional programmes is scarce (Brosens 2013) and mostly focuses on the benefits of participation (e.g. Digennaro 2010; MacKenzie 2006; Ward 2009). Academics emphasize that not only participants, but also non-participants should more often be included and given a voice (Brosens 2013; Johnsen 2001). This project responded to this research gap by exploring the situation of both participants – focusing on the motivations, and non-participants – focusing on the barriers.

7.6 Literature was used to search for motivators and barriers to participation in prison activities that could be included in the standardised questionnaire. These items were supplemented with motivators and barriers revealed during the focus group interviews. For instance, during the focus groups it became clear that some prisoners did not participate because they did not know how to subscribe. This was not found in the literature, but the results of the standardised questionnaire showed that this had an important influence (e.g. sport activities: 11.8%; education: 10.0%; vocational education: 15.8%; socio-cultural training courses: 10.9%). Formulating the research and survey questions in a participatory manner has resulted in different, deeper questions than if solely based on a literature review, and thus also led to other results.

7.7 A final benefit is related to the challenging research setting: a remand prison with many foreign nationals. Having a large number of foreign nationals within a prison poses different challenges, including language barriers (Atabay 2009; Banks 2011). The literature review of Yildiz and Bartlett (2011) has shown that foreign national prisoners are frequently excluded from research. Despite the difficulty of involving them in the project, we took up the challenge, and with success. Our research provides insights into the participation of foreign nationals in prison programmes, which was until now an under-researched topic.

7.8 Despite the success of the research project, there are some aspects that we need to take into account in the future. First, it was difficult to get prisoners involved in the morning because they wanted to sleep. In addition, we have to acknowledge that setting up a collaborative research project was a time-consuming process. However, it not only generated academic research material, but it also empowered the different partners of the steering committee. We experienced that the impact of the research is more widespread than we initially anticipated. The members of the steering group, especially the activity providers, have learned what they can achieve through collaboration. During the research project, they started to experiment with changes in the way of informing prisoners about the activities offered. In the future, they will work together more and come out of their own niche to set up cross-sector cooperation. This cross-cooperation not only concerns the activity providers at the prison where we performed the research project, but also activity providers of other prisons. The members of the steering group informed their colleagues at other correctional institutions about the research process and results, and some of them have already decided to conduct research in their own institution.

7.9 The research model developed by the steering group can be considered as a template for this future research. However, the model will only succeed if we take into account the specific circumstances in each prison. For instance, a remand prison has a different working method than an institution with convicted prisoners. In addition, low-, medium- and maximum-security prisons also differ. Nevertheless, we hypothesize that several lessons learned during our research project are applicable to studies concerning prisoners in a range of correctional settings (e.g., other remand prisons, prisons housing convicted felons, youth facilities). For instance, being flexible in scheduling the data collection is necessary (Fox et al. 2011). The prison manager proposed organising the data collection on the male section during the weekend. In addition, the prison guards called a strike on the day the questionnaire would be administered on the internees section, which led to the necessity of being flexible in the scheduling. At all times, we could rely on support from the prison management. In order to achieve mutual trust, researchers need to discuss the whole research process with the prison management representative(s). This can also be considered as one of the lessons that can be used in future prison research.

7.10 Lastly, one of the challenges of our research project was to equalize the power imbalances between the more powerful (researchers and activity providers) and the less influential (prisoners). The evaluation of the steering group indicates that we only partly succeeded by including prisoners as central figures by means of providing information, giving feedback, etc. during the whole research process. By involving them in various research stages, we tried to increase their power over the content of the project as much as possible. Nevertheless, their agency was still impeded by the structures of the correctional institution (i.e. an overcrowded remand prison). If the prison management had given permission to involve prisoners in the steering committee, they might have been a real part of the partnership. In other research settings, for instance prisons where sentences are served, it is possibly more feasible to equalize the power imbalances. Mostly, these prisons do not have high levels of overcrowding and are less frequently confronted with transfers to other prisons.

Notes

¹The activities that were available in the prison are sport activities (e.g., football, volleyball, fitness training, zumba), educational courses (e.g., Dutch for foreign speakers, computer courses), socio-cultural training courses (e.g., social and communicative skills, aggression management, surviving prison), vocational training (e.g., a consultant who guides prisoners in their search to work, group courses about how to build up a CV or job interview training), and mental health care (i.e. individual conversations with an employee of the centre for mental health care).

References