Exploring the challenges of implementing Participatory Action Research in the context of HIV and poverty

Curationis 33 (2): 69-78 HIV/AIDS is having a devastating impact on South Africa and particularly on poor communities. Empowerment of communities has been identified as an important step towards mitigating the consequences and helping communities to overcome the challenges presented. Participatory Action Research (PAR) has been identified as a useful methodology for the purpose of facilitating empowerment. This study ex­ plores the challenges involved in implementing PAR in the context of HIV/AIDS and poverty. In this article, the author describes a PAR project that took place in 2003/ 2004 with a group o f five Xhosa speaking people living with HIV/AIDS in Masiphumelele, Cape Town. The aims of the study were to: 1. Create an opportunity for the participants to engage in a participatory process aimed at self-awareness and empowerment. 2. To record and analyse this process with the intention of producing insight into the use of PAR in the context of poverty and HIV/AIDS and to identify the challenges involved. The findings of this study highlight some important insights into the process of engaging people in the PAR process and the experiences of HIV positive people living in the context of poverty. The study explores the challenges involved in the process of empowerment and examines the process of “transferring” power and control from the researcher to the participants. Challenges were uncovered both from the point of view of the researcher who had to “let go of control” and participants who had to take on control. Participants struggled with issues of low self-efficacy and learned helplessness. Fluctuations in health also contributed towards alternat­ ing periods of hope and despair and these problems had an impact on their motiva­ tion to participate in the study. Lack of motivation to participate is a challenge high­ lighted in the literature and explored in this study. Participation is necessary for a study of this nature to be of benefit to the community, but unfortunately those most in need were found to be least likely to participate. The study also critically examines the research process that was conducted and highlights the positive and negative contribution of the process towards empower­ ment. Certain aspects of the research process, including the contracting process, were identified as being problematic as they emphasize the power and control of the researcher rather than the participants. Recommendations for future research in­ clude: Promoting participation among the disempowered; the Contracting process and Power relations in PAR Correspondence address Mrs. Wendy Rosenthal Address: Postnet 114 Private Bag X I8 Rondebosch 7701 Tel: (021)959-3482 Email: wrosenthal@uwc.ac.za Kev words Participatory Action Research, HIV/


Introduction
This article describes a Participatory Action Research (PAR) project that took place in South Africa in 2003/2004.The participants in the research were a group o f five black, Xhosa speaking, HIV p o sitiv e p eo p le living in Masiphumelele, an informal settlement near Cape Town.PAR is an unusual research methodology, the purpose of which is not only to produce knowl edge, but also to facilitate the empow erment o f the participants and enable them to use the knowledge produced to improve their situation (Babbie and Mouton, 2001).PAR is not, however, a simple methodology and there are many challenges involved in the implemen tation (Cornwall and Jewkes, 1995;Gray, Fitch, Davis and Phillips, 2000).This article explores the challenges that took place in the PAR process and pro duces recom m endations for future studies.

Participatory Action Research
PAR has proved to be an important tool for addressing issues o f powerless ness.In PAR, the participants are seen as partners in the research process and not merely subjects o f the study.This means that they are involved in shap ing and directing the research and take part in the decision-making process (Cornwall and Jewkes, 1995).PAR in volves "collective, self-reflective en quiry that is undertaken to improve a situation" (Koch, Selim and Kralik, 2002, p 110).The process begins with exami nation and collaborative enquiry into the lived experiences o f the partici pants.The participants are assumed to have esse n tial and valid understandings and experiences, which are vital to comprehending their world and to bringing about change.New insights into the situation and the par ticipants' lives are gained through the action reflection cycle, a key compo nent o f PAR.This cycle involves a cir cular process o f action followed by re flection, followed by further action which is informed by the reflection phase.The goal o f this cycle is for the researcher and participants to con sciously engage in a process o f learn ing from their actions (Zuber-Skerritt, 2001).From the new insights, which the participants gain into their experiences, the researcher helps to facilitate the planning o f action to bring about change (Babbie and Mouton, 2001).The goal o f PAR is to bring about mean ingful change in the lives of the partici pants and the community.Change may come in the form of individual or group empowerment, increased capacity to problem solve, and/or the transforma tion o f o rg an isatio n al stru ctu res (Cockbum and Trentham, 2002).

Background and motivation for the study
The HIV/AIDS pandemic has spread rapidly across the world to affect all nations.Not all groups, however, have been equally affected.HIV/AIDS has been particularly devastating to those sectors o f society, which were already marginalised, powerless and vulner able.It appears that a critical factor in the pandemic is the issue o f power re lationships.Those most affected are those who have little control over their lives.They are the most vulnerable to infection and the least able to deal with the consequences o f the disease (Nel son and Wright, 1995;Van der Vliet, 1996).Just as the virus depletes the human body o f its natural defenses, it also depletes the individual and family of their resources and ability to cope (Lyons, 1998).There is growing evi dence that sustainable solutions for these kinds o f problem s cannot be found by experts alone.It is necessary for the people to be involved in finding th e ir ow n so lu tio n s (B abbie and Mouton, 2001).Thus the motivation for this study is based on the need to en gage people living with HIV/AIDS in an empowerment process aimed at find ing their own solutions.

The use of PAR in the study of HIV/AIDS and empowerment
PAR has been employed extensively in the study o f HIV/AIDS.For example, Kesby (2000) and Mabala and Allen (2002) studied how the community could be engaged in the prevention of the spread o f HIV/AIDS.Lindsey, Staj duhar & McGuinness (2001) report on a PAR intervention, which engaged the community in developing community-based organisations for caring for people living with HTV/AIDS.However, there appears to be a gap in the litera ture in the area o f individual empower ment in the context o f HIV/AIDS and poverty.Some studies were found that focused on empowering individuals to address their personal needs in con texts other than HIV/AIDS (e.g.chronic illnesses (Koch and Kralik, 2001) and disability (Stewart and Bhagwanjee, 1999).These studies bear relevance to the current study as they also deal with chronic conditions and relatively marginalised groups and thus involve similar issues to those raised by the current study.

Challenges in implementing PAR
It is important to note that PAR is not a simple method to implement or an easy alternative to other research methods (Cornwall and Jewkes, 1995).Gray et al. (2000) point out that the nature of PAR and its emphasis on partnership and participant control, creates a number of challenges for the process of implementation.These include: prob lems in transferring control to the par ticipants, conflicting expectations and agendas, problems in motivating peo ple to engage in the research process and the time required for building part nerships.

Contribution of this study
The literature review ed shows the great potential o f PAR in helping peo ple to gain control over their lives in disabling environments.However, a number o f challenges have also been highlighted.In the context of the great need for sustainable solutions to the many problems caused by the HIV/ AIDS pandemic, it is believed that this study can make a worthwhile contribu tion by analysing the PAR process that was implemented, exploring its impact on empowerment and highlighting the challenges encountered in the process.

Research problem I question
What are the challenges involved in implementing PAR in the context of HIV and poverty?

•
To create an opportunity for a group of people living with HIV/ AIDS to engage in a participa tory process aimed at selfawareness and empowerment.

•
To record and analyse the PAR  Mouton, 2001).This study recorded the PAR process that the researcher facili tated with the participants and de scribed the challenges involved.

Sampling methods
The Project and this research also led to an unanticipated problem.A misunder standing arose between the researcher and the participants, where the partici pants assumed that they would be paid for their participation in the research as they had been paid for the Memory Box Project work.

Gaining Entry
The researcher asked the permission o f the Memory Box Project facilitators to address the p articipants in the Memory Box Project and to invite them to take part in the research.At the first meeting the researcher explained the purpose o f the research and invited the participants to take part in the research.
It was explained to the participants that they were under no obligation to take part.All five members o f the memory box project agreed to participate in the project.

The Participants
The participants included four women and one man.For the purpose of pro tecting their identities, the participants' names have been changed.Details about the participants are summarized in table 1.

Data Collection
Data collection was a process within a process -that is, while the participants took part in the PAR process and gath ered information on themselves, the researcher recorded how the PAR proc ess unfolded.
D ata collection was conducted by means o f focus groups, which were tape-recorded and transcribed.A trans lator was employed for the sake o f the group members who were not fluent in English.Focus groups were chosen as the most appropriate data collection method for this research as they allow the participants to interact with one another, share ideas, solve problems and make decisions together (Flick, 2002;Chiu, 2003).This study was de signed to involve the participants in a PAR process aimed at empowerment and taking action against the disabling circumstances in their lives.The proc ess was based on the Action Research steps described by Stringer (1999, p 18): "look", "think" and "act".Stringer's steps are a cyclical process involving a number o f repetitions o f the steps (Stringer, 1999).

Analysis
The approach used for data analysis was qualitative content analysis.This approach was used as it is an effective method for analyzing transcript data and for organizing and simplifying data into meaningful themes (Patton, 1987).Analysis was done using the follow ing techniques: 1. Immersion in the data This was done by listening to the tapes, reading the journal entries and reading through the transcrip tions several times while taking notes (Creswell, 1998).

2.
Coding and categorizing: In this process, the researcher identi fied themes in the transcripts and codes and categories were assigned in order to organize the data and make it more ac cessible for further analysis (Pope, Ziebland, & Mays, 1999;Mason, 2002).

3.
Developing pictorial represen tations: According to Pollock (1991, p 300) "the task at hand...is not to describe every piece o f the jigsaw puzzle, but rather to stand back and paint a picture o f the whole".The re searcher developed a number of pictorial representations and tables to organise the informa tion and to help the researcher gain an overall perspective of the data.

Reliability of the Data and Data Capturing Methods
A number o f issues are important in evaluating the reliability and accuracy o f tape-recorded data and transcrip tions.Firstly, there is the sound qual ity o f the recordings and the adequacy o f the transcriptions and secondly, the inclusiveness o f the data (Perakyla, 1997).

• Accuracy of Tape-Recorded Data and Adequacy of Transcriptions
Overall the sound quality o f the tapes was good.However, in places there was some loss o f data due to back ground noise.Accuracy o f the tran scriptions was checked by the re searcher and by an independent con sultant who checked both the transcrip tion and the translation o f the data.The consultant reported that there were some mistakes in the translation.How ever, in her opinion, the group discus sion was translated and transcribed sufficiently accurately and in sufficient detail for the purposes o f this research.
In the presentation o f findings, refer ence has been made to where data was lost or difficult to hear.

• Inclusiveness of the Data
Perakyla (1997) points out that taperecording can leave out certain aspects o f the interaction under study and thus impact on the inclusiveness o f the data.
In order to address this issue, field notes were made immediately after the focus groups in order to capture other information that was not captured on the tape (e.g.seating arrangements, non-verbal behaviour, events that hap pened before and after the tape-recorded session etc.).

Credibility
According to Lincoln and Guba (1985), credibility refers to how accurately the findings reflect the reality o f the study participants.In this study, member checking and peer debriefing were used to establish the credibility of the re search results.These methods are de scribed below.

• Member checking
Member checking involves taking the researcher's findings back to the par ticipants and asking them whether these accurately represent their mean ings (Gamer, 1991;Mason, 2002).In this study, the researcher took the prelimi nary results back to the group mem bers and they gave feedback indicat ing that they recognised their experi ences in the descriptions given.

• Peer debriefing
Credibility can also be established by asking a colleague to review the analy sis process and question the researcher on how he/she came to his/her conclu sions.This is referred to as peer de briefing (Lincoln and Guba, 1985).In the current study, the researcher's su pervisor was given the transcriptions to read and she was thus able to point out to the researcher when she had made assumptions about the data that might not be valid.

Transferability
Transferability/ generalisability can be established by giving detailed descrip tions of the context in which the re search took place.This allows the reader to identify similarities to other contexts, and thus apply the results to similar settings (Lincoln and Guba, 1985).In this research, descriptions of the participants are given as well as the context in which they are living.

Ethical Considerations
The Council for International Organi zations of Medical Sciences (CIOMS) (1993, p 10) requires that "all research involving human subjects should be conducted in accordance with three basic ethical principles: respect for per sons, beneficence and justice."In this project, a number of processes were put in place to ensure these ethical princi ples were upheld.The proposal was submitted to the UCT Health Sciences Faculty Ethics committee for ethical clearance b efo re the study com menced.Informed consent was taken before the process began.A detailed consent form in both English and Xhosa was given to the participants as well as being explained verbally.Par ticipants were informed of their right to refuse to participate and to withdraw from the study.Confidentiality was maintained throughout the study.No names or personal details that could identify the participants have been used in this article or any other docu ment relating to this study.The partici pants were warned that they might find the intimate sharing o f issues around their experience o f HIV/AIDS distress ing and they were assured that there would be support and counseling pro vided to them if necessary.

Discussion of Findings
The findings o f this study highlight some important insights into the proc ess o f engaging people in the PAR process and the experiences o f HIV positive people living in the context of poverty.In this section these findings are explored and compared to the find ings o f other studies.1990).Prestby et al. (1990) propose that in general people decide whether or not to participate by weighing up the per ceived benefits o f involvement and balancing it against the potential costs.

Engaging People to Participate in PAR
Cornwall and Jewkes (1995) point out, however, that people's motivation to take part tends to fluctuate over time.
As circumstances change and / or the participants discover that they are not going to get the benefits they expected to gain from involvement, motivation may dwindle (Cornwall and Jewkes, 1995).This is in fact what happened in this research study.
Zodwa: "1 think from last year there were researchers from UCT who gave us money, so I think that is our prob lem.We were thinking maybe you were going to give us money because we were not working." As the quotation above shows the par ticipants expected to be paid for their involvement.The expectation o f pay ment arose because the participants m isunderstood the relationship be tween the UCT Memory Box Project and the researcher's study.The re searcher entered the research site un der the auspices o f the UCT Memory Box Project and thus her study was assumed to also be a UCT project op erating under the same conditions in cluding payment of the participants.As a result, the participants entered the process with high motivation.This was not entirely motivated by payment: they also expressed excitement and hopefulness about the possibility o f gaining skills and knowledge through the research process.However, a few weeks later, once they had realized they would not be paid for their time, their motivation and excitement dwindled and they eventually decided to with draw.Thus, it appears that for the par ticipants in this study, the benefits of engaging in the PAR process were at tractive in the beginning, but not suffi cient to continue their involvement once the realized that there would be no financial/ material benefit.

Achieving Participant Control of the PAR Process
Cornwall and Jewkes (1995) suggest that the key difference between partici patory research and conventional re search is in the area of power.Cornwall and Jewkes (1995) and Reid ( 2004) ar gue that in true particip ato ry ap proaches, the control over the research process lies in the hands o f the partici pants.They point out, however, that transferring power into the hands of the participants is a challenging proc ess.Below, some o f these challenges are explored.

Power Imbalances: An Unequal Starting Point
Brugge and Kole (2003) point out that researchers come into a poor commu nity with a high level o f power com pared to the participants, because of their background o f education and ac cess to resources.This may be further enhanced in the minds o f the partici pants who may perceive the research ers to be the experts and lack belief in their own knowledge and abilities.Gray et al. (2000) point out the need to be aware o f this power differential when entering the research field.They advo cate that researchers take special care to actively involve the participants in every aspect o f the process, rather than just obtaining passive agreement to the researchers' agenda.

Contracting: The First Contact
The consent-taking process is often the first interaction that the researcher has with the participants and thus it has an important role in shaping the nature o f the relationship between the researcher and participants.The usual method o f taking consent in a research study involves using a pre-prepared consent form.However it is question able whether this is the most appropri ate and empowering way o f engaging people in a PAR project as it is mostly a one-way process controlled by the re searcher.This is contrary to the PAR principle o f shared control.This one way process may reinforce the expec tation in the minds o f the participants that the researcher is in control and they are to take a more passive role.It also emphasises the power differential be tween the researcher and the partici pants.It seems appropriate, therefore, that for this kind o f research, alterna tive ways o f engaging participants in the PAR process should be explored.Meyer (1993) suggests,the concept of "process consent".This refers to re tak in g co n sen t as the p rocess is adapted and changed.Thus it is not an alternative to the pre-prepared consent form, but it allows for flexibility and the involvement o f the participants in shap ing the process.

Facilitator Issues -"Letting Go of Control"
Kaplan and Alsup (1995) note that one o f the potential obstacles in the trans fer o f power to the participants,is the inability o f the researcher to "let go of control".Low, Shelley and O 'Connor (2000) report that, in spite o f having a sincere desire to relinquish control to the participants, they found themselves unintentionally taking over control at certain points in their study.Sometimes this was due to pressure from the par ticipants who wanted them to take con trol.At other times, this came out o f the assumption that the participants were not capable o f taking responsibil ity for certain tasks.Low et. al.'s (2000) experience bears similarity to what oc curred in this study.Although the re searcher valued the principle o f trans ferring power to the participants, she found herself unintentionally taking control for similar reasons.Research ers need to be constantly alert to this dynamic and to monitor the power dy namics in the process (Kaplan and Alsup, 1995).In Low et al. (2000), the researchers reflected on their actions in between each group session and at tempted to take corrective action and hand back control to the group mem bers in subsequent sessions.However, at times, the participants resisted this attempt to hand back the control to them.This raises another important aspect in power transfer, that o f par ticipant resistance.

Participant Resistance
Researcher: Okay, Nandi?... What would you like to happen [next in the re search]?Translater: (translating for Nandi) She say she would also like you to solve their problems.
In this study, the researcher found that the participants did not come up with ideas o f their own, but tended to defer to the researcher's suggestions and look to the researcher for solutions, as the quotation above shows.In refer ence to this problem, Cornwall and Jewkes (1995) point out that PAR par ticipants do not always want control o f the process.They tend to prefer the researchers to control and decide things for them and come up with the ideas (Low etal., 2000).This can be for various reasons including: lack o f be lief in their own abilities (Brugge and Kole, 2003) or fear o f the project be coming too time-consuming and bur densome (Gray et al. 2000).Gray et al.
( 2000) suggest that researchers need to take extra time and effort to actively encourage participation in all aspects o f the process.This includes nurtur ing the confidence o f the participants in their own abilities and holding back one's own desire to rush in and "fix" the problem.

Learned Helplessness and Learned Hopefulness
Prilleltensky, Nelson and Peirson (2001) observe that power and control are not inherited attributes that an individual either has or has not; they are qualities that are developed by continual inter action with the environment."Learned helplessness" theory, proposes that exposure to uncontrollable events will tend to lead to individuals losing moti vation and feeling helpless.In contrast, "learned hopefulness" theory p ro poses that if individuals are given op portunities to exert control over their lives, their motivation and sense o f control will increase (Zimmerman, 1990).
Participating in community organisa tions or projects that offer individuals opportunities to learn new skills such as problem-solving and taking part in decision-making, can facilitate the proc ess o f building "learned hopefulness" and consequently empowerment.In the current study, all the participants come from impoverished backgrounds and have experienced a number o f trau matic events in their lives over which they had no control (e.g. the death of loved ones due to HTV/AIDS, the loss o f homes and financial security, rejec tion by friends and family etc.) Accord ing to the theories presented above, these circumstances are likely to have led to the participants feeling power less and lacking motivation as is shown by the extracts below: Translator-translating for Thandi: (de scribing the day she found out about her HTV status) Thandi said that there is nothing that makes her happy.Eve rything changed, because she knows that any day she can die.

Themba: "To stay without a job, be cause I am not working anymore that makes me sad. I don' t want to sit do ing nothing, because I 'm going to steal from the others (LA UGHTER) and then I 'm going to ja il if I rob the others ..Staying without a jo b is something that makes me sad. Also think about every day, think about why I can' t get a job. "
In this context, it is understandable that they should have developed a low sense of self-efficacy.This further ex plains why they resisted control being passed over to them, as having a low sense of self-efficacy, they lacked the confidence to take control and make decisions.

Low Self-Efficacy
According to Bandura (1997), people with low self-efficacy are likely to be demotivated, pessimistic about their chances of succeeding in life and are more likely to suffer from anxiety, sad ness and depression.As a result, peo ple with low self-efficacy are likely to lack the confidence and the motivation to participate in empowerment projects and if they do participate they are less likely to take an active part in the proc ess.This presents a problematic situa tion: in order to be empowered, people need to take part, but in order to take part, they need to be empowered.This was a dilemma faced by this study.The researcher attempted to facilitate em powerment through giving the partici pants an opportunity to work on their problems with the support of the re searcher and the group.However, the participants were not greatly convinced that they could successfully act against the disabling circumstances in their lives and therefore their motivation to make an attempt was low, as the quota tion below reveals.
Researcher  (1993) suggest that people can be en couraged to take on projects if their confidence is built up gradually by giv ing them small tasks and then increas ingly larger responsibilities.Bandura (1997) suggests that another way o f building self-efficacy is for the people to watch someone similar to themselves succeed at a task or successfully solve a problem.

Hope and Despair in HIV
Zodwa: "I was worried that time my  (Cochrane, 2003).The participants in this study had already faced a great deal of uncertainty due to other fac tors in their lives, before their diagno sis with HIV/AIDS, for example, pov erty, unemployment, death o f loved ones.It seems that HIV/AIDS added another dimension to an already un certain existence.Kylma et al. (2001) point out that fluc tuations between hope and despair can have serious consequences for people living with HIV/AIDS.When hope is dominant, individuals are likely to have energy and motivation, but when de spair or hopelessness is the overriding factor, the opposite is likely.This has implications for the empowerment proc ess.Kylma et al. (2001) recommend that people working with people living with HIV/AIDS, should be acutely aware of these mood changes.They suggest that professionals ask their clients about their sense of hope and despair and use interventions that foster hope.These recom m endations, though aimed at nurses in a clinical setting, are also relevant for the empowerment proc ess.Knowing where the participants are on the continuum o f hope and de spair, could be important for structur ing the research process as periods of despair could negatively impact on the participants' ability to take part in the process.

Conclusion and Recommendations
PAR methodology offers an exciting alternative to the more traditional topdown, researcher controlled approach and holds out the possibility o f achiev ing more effective and sustainable out comes for communities.However, as this research has highlighted, there are many challenges involved in imple menting the PAR process.Important findings are discussed below and rec ommendations for future research are made:

Promoting participation among the disempowered
In order for people to benefit from em powerment projects, they need to par ticipate in these initiatives.However, as has been shown by this project, peo ple in poverty-stricken circumstances are unlikely to participate unless they are given some kind o f compensation or reward for their efforts, which con tributes towards their struggle for sur vival.
This can be problem atic as many projects lack the budget to give sig nificant financial compensation.It is therefore recommended that research is undertaken on how people can be attracted to participate.This could in volve studies which investigate how people's motivation is affected by non material benefits and how these could be marketed to potential participants.

The Contracting Process
In the current study, a particular area o f concern was the contracting proc ess.It was found that the contracting method used was not sufficiently par ticipatory and an opportunity for em powerm ent and skills transfer was missed.It is recommended that further research is undertaken on how to make this process more participatory and empowering.

Power relations in PAR
Handing over control to the partici pants is a vitally important objective of PAR as it is through this process that the participants are empowered and given the opportunity to learn new skills.However, in this study, this proc ess was blocked by a number o f prob-lems, including, resistance from the participants and their lack of confidence in themselves.It is therefore recom mended that further research is done on how this process can be effectively implemented and how the barriers such as low self-efficacy can be overcome.

CD4 was 147 and I was very very weak and I was worried, because I think maybe I'm going to be sick, but now I fe e l very strong and healthy...very strong! So I 'm always hoping that may be I'm going to be alive till the gov ernment get a cure fo r us...."
: risks and realities of child hood during the HTV epidemic.United Nations Development Programme: HTV and Development Programme.Issues P ap er 30.w w w .undp.org/hiv/ publications.html.Accessed 9 March 2009 PEIRSON, L 2001: The role of power and control in children's lives: an eco logical analysis o f pathways toward wellness, resilience and problems.Jour nal of community and applied social psychology.11:143-158.