Always leave the audience wanting more: An entertaining approach to stimulate engagement with health research among publics in coastal Kenya through ‘Magnet Theatre’

Background: Magnet Theatre (MT), a form of participatory community theatre, is one of several public engagement approaches used to facilitate engagement between KEMRI-Wellcome Trust Research Programme (KWTRP) researchers and public audiences in Coastal Kenya. We describe how we used MT as an entertaining forum where audiences learn about research, and where researchers learn about how the public views research. Methods: Drama scripts depicting community interaction with different aspects of research were developed iteratively with research staff, a theatre company and community members. Six fortnightly theatre outreaches per site over two months, attracting a total of 1454 audience members were held in Mida, a rural village 30 km north of Kilifi; and in Mtwapa, a peri-urban town 45 km to the south. Audiences were presented with dramatized health research-related dilemmas and subsequently invited to enact their responses. Evaluation comprised, notes and observations from meetings, rehearsals and outreaches, transcripts from a review workshop with repeat audience members (n=21), a reflection meeting with KWTRP engagement staff (n=12), and a group discussion with the theatre company (n=9). Discussions were recorded, transcribed, translated to English and analysed using thematic approach. Results: Despite being costly in terms of time and expense, we argue that MT in public spaces can assist audience members to navigate ‘border-crossings’ between everyday contexts and scientific/research concepts. This can enable audiences to share their views and concerns and enact their responses to research-related dilemmas. Conclusions: While reporting on MT’s successes, drawing from literature on rumours, we acknowledge the limitations of individual engagement activities in providing long-term solutions to address alternative interpretations and rumours about research, in the context of local and global inequities. MT, however, presents an opportunity for researchers to express respect to public audiences through making research more accessible and providing opportunities to listen to public views and concerns.


Introduction
There is a growing body of literature emphasising the importance of engagement between researchers, community members and the broader public for a range of reasons; including raising awareness and understanding of research, decision-making about research participation, raising researcher awareness of public concerns about research and incorporating public views into research [1][2][3] . A range of engagement approaches have been described 4 ; however, descriptions of the use of MT as an approach for engaging the public with research are rare.
Theatre has been used for decades as a creative approach for fostering community engagement with sensitive, technical issues. According to practitioners, theatre helps to break down hierarchies for nurturing trust over time, allow different parties to share their perspectives in a safe space and facilitate cross-learning in an enjoyable way 5 . Attitudes and behaviours are portrayed by antagonists and protagonists and explored jointly with participating audiences 6 . Theatre has been used in different health fields including health promotion 7 , health education 8 , policy engagement 6 and to a lesser extent to engage the public with health research 9,10 .
For health promotion, community theatre performances in Bangladesh aimed at raising awareness of eclampsia and means of effective treatment 8 . Though the evaluation only involved 15 audience participants, the authors present evidence that the initiative was well received and contributed to improved audience understanding related to eclampsia. In 2018, theatre was used for public engagement with antibiotic use in Myanmar 11 .
The authors report improvement in audience understanding of antibiotics and its use, and that audience preferred Forum Theatre as a form of engagement compared to conventional approaches to giving formal health talks.
While some articles convey the potential strengths of theatre as a means of health promotion and raising audience awareness of health issues, another use described in the literature is to engage audiences with health policy and health service implementation. For example, in Canada, theatre was used as a method of engaging public audiences in the development of health policy on preimplantation genetic diagnosis 12 . The team drew on audience observation and post-performance discussions to gather key public views about a controversial genetic procedure for selecting the genetic characteristics of embryos created through in vitro fertilization.
In describing the use of theatre to engage audiences with ethical dimensions of healthcare, Bowman 6 reports that using theatre enabled audiences in London, UK, to understand and grapple with balancing the risks against the benefits related to critical healthcare decisions. She argues that using approaches such as theatre and radio, which are more familiar to members of the public than research settings, can contribute to reducing barriers to communication of complex research ethics topics, and increasing trust in healthcare.
'Forum Theatre' was used for faculty development at the University of Michigan's (USA) multicultural medical school 13 . The initiative involved intermittent pausing of the drama performance for audience reflection, problem analysis and facilitated discussion, and the process enabled audience insights to be incorporated into the way courses were implemented. The approach raised audience awareness of key planning and implementation issues, changed behaviour of the facilitators and offered strategies of addressing dynamics in the classroom 13 .
The use of theatre in biomedical research over the last two decades is not widely described in the literature 10 . A few articles, however, report success with the approach, for example, in South Africa, a study team used theatre to disseminate findings of research they conducted in six villages on understanding causes of stroke 14 . Audiences in the six villages felt that portrayal of study findings was concurrent to real life happenings. During discussions, potential future research interventional thematic areas emerged.
In Cambodia, health research staff engaged with local communities towards developing a public theatre presentation aimed at educating communities about malaria transmission, prevention and treatment between health researchers and local communities [15][16][17][18] . This hybrid between engagement and health promotion involved working closely with 600 community members to develop a theatre production which reached audiences of 12,000 people across 20 villages. The authors present evidence that this outreach impacted audience members' understanding of malaria and its prevention.

Amendments from Version 1
We appreciate all the useful comments from the reviewers. We have replaced Magnet Theatre with abbreviation 'MT' throughout the manuscript. We have also edited the manuscript as stated below.

Methodology
Additional information has been provided about what informed the choice of using MT.
Rephrased the aim of MT and removed the word impact.
Explained the M&E team was not part of the MT implementation team and how evaluation activities were integrated throughout the implementation of the project.
Added a description and a table of the audience who attended the outreaches.

Discussion
Added some information on how we had planned to use MT in the future although this was not possible because of COVID-19.
Added a recommendation for institutions intending to use MT.

Conclusion
Added the word ' co-created' in the conclusion.

REVISED
MT is a specific form of participatory community theatre that is conducted at an identified venue and at a specific time, over an extended period of time; aimed at "initiating and maintaining interactive communication with, and within, a community" 5,19,20 . MT has a distinct component where the drama is paused at a point where the actors face a dilemma and the audience are invited to reflect on the problem, discuss possible solutions to the dilemmas and enact how the protagonists should proceed 5 . This approach recognises that local socio-cultural norms and contexts influence individual responses to dilemmas and draw on these to elicit culturally appropriate solutions to specific challenges. In this way, the audience become actors taking responsibility for the solution. Following the performance, audience members reflect on the drama and apply the learning to their own lives to resolve challenges and dilemmas. Performing the drama at the same venue and time over a specified duration provides an opportunity for tracking attitude change over time and provides a forum for the magnification of behaviour change. The approach has been used for decades in grass-roots community HIV interventions across different countries to raise awareness, promote social and behaviour change and incorporate public views into planning and implementation of interventions and is identified as a one of the best practices in HIV prevention 5,20 . For example, the Program for Appropriate Technologies for Health (PATH) has used MT in Kenya since 2000 to encourage community communication, interaction and learning around HIV and AIDS issues, consequences of early marriage, and TB prevention and treatment 20 . It has also been used to engage Zambian communities with malnutrition 19 .
While there is a range of different uses of theatre in development, given its widespread use and success in health promotion and policy engagement, KWTRP engagement practitioners felt that MT potentially provided a locally appropriate and entertaining means of engaging public audiences with research. At the outset, it was felt that as in other settings, MT could raise public audiences' understanding of research, nurture attitudinal change supportive of research and provide members of the public with fora to express their views and concerns in relation to research. There is, however, very little published literature describing the experiences of MT as a means of engaging the public with health research. This report focuses on MT as one of the public engagement approaches used at KWTRP and describes our experience of using the method.

Study site
KWTRP conducts integrated epidemiological, social, laboratory, clinical and health systems research, with results feeding into local and international health policies and practice. Research is supported by a Community and Public Engagement (CPE) strategy which includes several engagement approaches that have been implemented since 2006. KWTRP engagement is planned, implemented and evaluated by a multi-disciplinary team comprising the head and manager of engagement (NM and SM respectively), ethics and engagement researchers/evaluators (GS, IJ, DK and AD) and the Community Engagement (CE) staff. CE staff comprise a team of 15 trained and experienced male and female residents between the ages of 38 and 58 who come from Kilifi and speak the local languages fluently. For Kilifi residents, like in other settings, health research is complex, filled with technical scientific jargon, and often not considered to be a priority topic for many. MT was selected as one component of our broader CPE strategy to engage public audiences with health research in socially and culturally appropriate ways 5 . We acknowledge that research institutions may be perceived as intimidating to many people and MT was an attempt to break down those barriers and move engagement from the research campus to more familiar public settings. Given the popularity of drama in Kenyan culture, we felt that it might enable the audience members to cross cultural borders of day to day life to the culture and language of science 21 . The initiative, like KWTRPs other engagement approaches, aimed at nurturing colearning and mutual understanding between researchers and publics and thus, enhancing the ethical conduct of research at KWTRP.
In March 2018, two areas 100 kilometres apart were selected for the MT initiative in Kilifi County. The first, Mtwapa, an urban town 40km south of Kilifi town, where minimal research has been done, and secondly, Mida, a rural area situated in the north where KWTRP census and research activities have been conducted for two decades. In keeping with the participatory nature of MT, site selection was done through a "walk-about" and consultation with village elders and chiefs at the two locations and members of Jukwaa Arts, the collaborating theatre company. Potential sites were considered in terms of their accessibility and capacity to safely host over 100 members of the public, their distance from potential distractions, and their being at a sufficient distance from quiet environments such as schools and hospitals to avoid disruption. The MT outreaches were held between March and September 2018 in Mtwapa and Mida respectively.
Co-creating scripts "Jukwaa Arts", the local collaborating theatre company participated in a 3-day workshop at KWTRP to learn about health research and its social value, the range of on-going studies; and about research ethics. The workshop importantly, served as a platform for planning outreaches, discussing potential sites and introducing the research community to the theatre team and their work.
Theatre company members held informal talks with community members at the two sites to widen their understanding of community views about health research and inform themes and questions for the script content. From this, in consultation with engagement staff they drafted a script (see Extended data File 1 22 ). For example, the first draft script for Mtwapa, entitled 'Why do people refuse to participate in research', depicted a father who did not want his son to participate in research. Scripts were developed further and refined iteratively through discussion with the engagement team and during rehearsals. Scientific/research information was made as simple as possible, and technical jargon was avoided.

Magnet theatre performances
Prior to the six fortnightly outreaches held at each site, crowds were sensitised through a combination of village elder communication, posters, a public-address system playing music and relaying messages, and a mobile music and dance group called 'Simba Ropa' 1 . This approach attracted a crowd of about 70 people to each event, and this would increase as the play proceeded. Outreaches, run entirely by the Jukwaa Arts team and supported by KWTRP engagement staff, begun at 3.30pm and continued for 2 to 2 ½ hours.
Each MT performance involved several three-stage cycles. The first stage, acted by the theatre company for about 15 minutes, familiarised the audience with the drama's characters and their lives, gave the audience an outline of the context, and arrived at a dilemma or tension point for the characters. The performance was then frozen for the second stage where the actors would freeze like statues providing an opportunity for audience discussion and debate. In the third stage members of the audience would take up the roles of some of the characters depicting how they would address the dilemma. This cycle was repeated two or three more times over about 35-45 minutes, enabling audience members to air their views and address any questions and concerns through dialogue. After the cycles, a Jukwaa Arts facilitator would wrap up the session and, on some occasions, invite subject experts from KWTRP to speak at the end of an outreach to address information gaps raised during the performances. At the end of the performance, the actors mingled further with the crowd while packing up thus providing further opportunities to respond to more questions, refer individual health concerns to appropriate health facilities and to bond with the audience.
Post-performance debrief meetings were held between KWTRP engagement staff and Jukwaa Arts team to explore what worked well and what needed to be improved in the next outreach.
Monitoring and evaluation data collection All monitoring and evaluation of this work was guided by a Theory of Change (see Extended data File 2 22 ). Monitoring and evaluation methods aimed at documenting public views and concerns about research and assessing the MT's influence on public perceptions at KWTRP, and at guiding ongoing interventions, and comprised:

·
Minutes of planning meetings; · Field notes and structured observations of rehearsals and outreaches; • 12 drama performances' monitoring reports (guided by an observation and reporting tool) (see Extended data File 3 22 ); • A report of a mid-term review meeting with Jukwaa Arts team and KWTRP engagement staff; • Notes from post-performance de-brief meetings; · A final activity report from Jukwaa Arts; · A focus group discussion with 12 KWTRP engagement staff guided by a moderation tool (see Extended data File 4 22 ); · A half-day reflective workshop with 21 repeat audience members (who consecutively attended four or more of the six outreaches) guided by a workshop moderation tool (see Extended data File 5 22 ).
A monitoring tool was filled by a designated member of Jukwaa Arts during outreach, while the rest of the evaluation data was collected mainly by GS, NM, and IJ. KWTRP engagement staff and the evaluation team took field notes during all the rehearsals and outreaches and discussed their observations during post-meeting/performance de-brief meetings. Monitoring reports completed by the Jukwaa Arts team were verified by the KWTRP engagement staff and field-notes and notes from the de-brief meetings were added to complete the monitoring reports.
While the M&E team are part of the KWTRP engagement team, they played no part in the planning and implementation, which provided a degree of impartiality. M&E activities were integrated throughout the implementation of the project to enable critical reflection throughout and the workshop and FGD were done after the last performance.
The reflective workshop with the repeat audience and the Jukwaa Arts team was a half-day activity conducted in December 2018, three months after the completion of all MT outreaches. Reflective workshop participants comprised audience members who were observed by Jukwaa Arts team to have attended four or more outreach sessions at either Mtwapa or Mida. Though challenging within a busy public context, the Jukwaa Arts team provided a list of outreach regular attendees. The evaluation team attempted to purposively select attendees of the reflective workshop from this list to ensure diversity in views based on gender, estimated age and religion. They were invited through letters followed up by phone calls a few days prior to the workshop. Out of the 30 repeat audience members invited (15 each from Mtwapa and Mida) 21 attended (six from Mtwapa and 15 from Mida). The reflective workshop was facilitated by the evaluation team at the KWTRP, during which audience members were divided into small groups for a facilitated discussion guided by a topic guide. Discussions were audio recorded and notes were taken by the evaluation team.
Meetings were closed with a plenary session where the groups shared a summary of their views. This was important to enable participants to clarify and confirm aspects of the findings. In an attempt to reach data saturation from the repeat audience members, rather than facilitating several FGDs, we opted to invite as many repeat audience members as possible to one workshop in order to maximise the range of views.
An audio-recorded FGD was conducted with the KWTRP engagement staff in July 2019, which lasted 2 hours and 56 minutes with a short break in-between. During the FGD, engagement staff reflected on the summarised findings of the reflective workshop and shared their perspectives on the outreaches.
Audio recordings of the workshop discussions and the FGD were transcribed and Kiswahili content was translated to English.
Data was organized in QSR NVivo 10 software and analysed using content analysis approach. An initial coding framework was developed by GS and IJ and reviewed by DK. The improved coding framework was shared and revised following inputs by SM, NM, and AD where any discrepancies were discussed and resolved. Finalization of the coding framework was done iteratively with the co-authors (see Extended data File 6 22 ). Categories and themes were developed emerging from the data. A summary was then developed with the final themes derived from coding framework. The preliminary findings from the analysis of the FGD data was triangulated with the data from field notes and the reflective workshop. The findings were discussed in a meeting with KWTRP engagement staff for verification and comments which helped develop a final report.
Throughout the tool development, data collection and analysis, the researchers were aware of their positionality as KWTRP staff and how this could influence their personal views towards the evaluation of engagement activities. For this reason, researchers tried to be reflexive as much as possible ensuring they maintained neutrality during the interviews, carefully using probes and being conscious of and managing group dynamics during discussions; especially with their fellow staff; the KWTRP engagement staff and Jukwaa Arts team. Staff were encouraged to carefully consider any biases during post-meeting/ performance de-brief meetings.
This evaluation work was reviewed and approved by the Kenya Medical Research Institute (KEMRI) Scientific and Ethics Review Unit (KEMRI/SERU/CGMR-C/075/3416). Written informed consent (see Extended data File 7 22 ) was sought from repeat audience, Jukwaa Arts team and the KWTRP engagement staff who were involved in all interviews (focus group discussion, review meetings and discussions during the reflective workshop) and for audio-recording.

Results
Implementing MT KWTRP staff were initially unfamiliar with MT and two aspects of the initial stages assisted in the smooth running of the programme: the three day training and community walks with KWTRP engagement staff and Jukwaa Arts team; and performing a play for the scientific staff in one of the weekly seminar. The former enabled scripts to be co-authored and objectives to be agreed upon, while the latter was aimed at mobilising staff to participate.
"It was important for us to understand magnet theatre process so that when we are walking with them… we are able to support in meeting the objectives. For instance, this knowledge helped in the review of the script…". (P10, KWTRP engagement staff, Male) Despite performing for the scientific staff, only two research staff attended the 12 community outreaches. While experienced community engagement staff were able to address most questions and issues raised, in some cases, the presence of experts would have been desirable.
Challenges were encountered in identifying suitable sites which were busy enough to attract a sizeable audience whilst not being too noisy for engagement and maintaining a reasonable distance from places which could not be disturbed, for example schools and offices. Specific challenges were raised by the sites' proximity to busy roads, churches/mosques and informal bars selling palm wine, sometimes referred to as 'mnazi dens'.
"So for me the Mida site was not very far from the road which… okay when a heavy machinery is passing attention of the audience; they looked at the road unlike for Mtwapa where it was in the heart of the village, although next to a mnazi den…" (P10, KWTRP engagement staff, Male) Despite site selection challenges, early evening outreaches were attended by a total of 1454 adults, including some children living around the identified venues and it appeared that the public promotion of the events were effective. Estimated counts revealed that in both sites the majority of the audience appeared to be women as seen in the table below, and staff commented that they were mostly middle-aged.
Ice-breakers generally put audiences at ease at the beginning of the outreach; however, KWTRP engagement staff observed that at the first two performances, despite repeated probing by facilitators, audiences were initially hesitant to share views and experiences and to contribute to discussions. KWTRP engagement staff felt that this hesitancy was alleviated, firstly through the professionalism, willingness to learn and respectful conduct of the Jukwaa Arts actors, secondly through the MT approach itself, and thirdly through the audience learning It was felt widely that the crowd interaction spurred on by the 'freeze' point, an integral component of the MT approach, helped to nurture a rapport between the MT team and the audience. At intervals in the performance the facilitator halted the drama and opened the floor for discussion and invited two or three audience members to act out how they would respond to dilemmas. The growing audience discussion, reflection and laughter observed over subsequent performances evidenced this growing rapport.

MT stimulating audience engagement with research
In all outreaches, audience members were given opportunities to ask questions and raise concerns about KWTRP's research. The questions and concerns raised, spanned a wide range and comprised: that KWTRP promoted sex; questions about specific research procedures and blood drawing; the time taken from research conduct to getting results/findings (why some results are given immediately and others are not given); the difference between research and medical interventions (diagnosis and treatment); and aspects of research ethics, for example, KWTRP's response in the event of research participant death or refusal and voluntary participation/participation in research.
"What happens when one refuses to take part in research work, are there any consequences?" (Mida Outreach Audience Member, Male) Careful explanation of the purpose of the outreach was required to enable audience members to relax and participate actively in discussions. Jukwaa Arts actors observed that as audience members understanding of KWTRP grew, so did their willingness to participate in interactive components of the outreach.
In Mtwapa for example, where research is limited to a KWTRP satellite conducting HIV research with key populations (men who have sex with men and sex workers), more general awareness raising of KWTRP's work was required to create a rapport than in Mida where a considerably greater range of research and engagement takes place. A frequently raised concern for Kilifi community members is why blood samples are drawn for research and specifically that the volumes drawn from small children are too large 23,24 . This concern was also raised during MT outreaches. In Mida, an invited research staff explained the difference between clinical and research procedures and demonstrated the blood volumes taken for research purposes using different sample collection tubes, syringes, and storage bottles containing medium (liquid into which a blood sample is drawn into Another common concern raised by mainly women during community engagement activities, is that their husbands' suspicion is raised when women receive out of pocket allowance for research participation. A male outreach audience member from Mida reported that attending an outreach helped him gain a better understanding about why payments are made when an individual participates in research. Lastly, while community members reported that they learned from the dramas, some expressed a concern that the number of performances were insufficient. This suggests that careful consideration of the time and financial costs of MT are required for it to be employed as a long-term engagement approach.
"…I thought they [drama outreaches] were continuing for me to learn more but it came to an end…".
(Mida Repeat Audience, Female) "Leaving the audience wanting more" is however a common sentiment in theatre which can also be interpreted as audience enjoyment, and creating a thirst for future engagement.

Discussion
In our study we define a public space as a place that is accessible to all people and includes public squares, bus stops, roadsides and other places where people congregate and pass by. They range from quiet peaceful settings, to busy urban centres, abundant in disruption, noise and, of course, people. This contrast might precipitate an almost 'no-win' situation for engagement practitioners: a quiet site, while being ideal for performance and dialogue, may not attract large crowds, whilst a bustling noisy city square may attract a huge crowd but may be too noisy for meaningful engagement. To draw large public audiences, KWTRP engagement staff and Jukwaa Arts team opted for the latter, and drew from the experience and professionalism of the theatre company to address the consequences of their selection. Whilst practically challenging, a benefit of engagement in a public setting, according to some audience members, is that the public MT performances attracted audiences who would otherwise be missed by other common forms of engagement. That MT can potentially broaden audiences for engagement and convey science in familiar ways, enabling researchers to learn from wider publics is an important finding for public engagement with research.
Whilst distractions and noise are challenging, aside from the potential to draw large crowds, there are other benefits of engaging the public in their own environments. Several authors (see for example Bultitude and Sardo, and Gehrke) 26,27 describe the desirability of engaging publics 'organically' "in the places where they already exist and through those discourses and social practices by which they enact their status as publics" 27 . They argue that 'organic approaches' hand over control of the discussion to members of the public, enabling a more authentic sharing of views and perspectives. Though our evaluation is limited in its ability to explore the durability of gains in understanding over time, we feel that in the short term at least, MT enabled audiences to form opinions about research, based on a strengthened understanding, gained through watching and enacting familiar life vignettes in familiar spaces. In the field of science education, Aikenhead (1996) 21 likens the acquisition and understanding of scientific concepts, to crossing cultural borders from familiar 'life-worlds' to a largely unfamiliar 'science-world'. Situating research and scientific concepts and processes within scenes and dilemmas familiar to audience members are likely to have eased 'border-crossings' into the world of science and to some extent enabled learning. Jegede, in his article on the eco-cultural paradigm in science and mathematics education in Africa, describes this type of learning as "a state in which the growth and development of an individual's perception of knowledge is drawn from the sociocultural environment in which the learner lives and operates" (p. 124) 28 . Thus, we hypothesize that the three elements of MT (acting by the theatre team, freeze point/debate and acting by the audience) assisted audience members to border-cross into the world of research enabling them to grasp unfamiliar concepts: first, in comparison to engagement activities held in research institutions, unfamiliar and potentially intimidating to nonresearchers, situating engagement in familiar public locations is likely to have put audiences at ease with contributing their views and interacting; second, the co-creation process ensured that the vignettes presented during the plays closely resembled the everyday realities faced by people enabled them to relate to the characters and their contexts in relation to research; and third, those who enacted their responses to the dilemmas presented were able to incorporate their learning into practical actions, which is likely to affirm the learning.
Reflections of repeat attenders, and subsequent confirmation through community meetings suggested that acquisition of an understanding of KWTRP and its work may have addressed damaging rumours circulating in the community about KWTRP for some audience members. Geisler and Pool (2006) 29 argue that the origins of rumours about health research prevalent in sub-Saharan Africa and beyond, are likely to be more complex than just the result of knowledge deficits and that addressing knowledge gaps alone may at best only partially remedy the situation. They describe rumours as metaphors people use to express dissatisfaction and contempt towards inequities and wealth disparities between the global North and South, and between researchers, former colonial rulers and host communities. The periodic emergence of rumours surrounding research in Kenya, despite almost two decades of community and public engagement 23-25,30-32 , would confirm Geissler and Pool's sentiment. We acknowledge that individual engagement initiatives may not provide 'silver bullets' to address community and public views influenced, largely, by considerable global and local inequities; however, they may address part of the problem and provide temporary respite through expressing respect to communities by listening to their views and concerns. While inequities are unlikely to be ameliorated in the short term, and that research requires public support, community and public engagement will need to be sustained over a long-term using a range of approaches to support the development and testing of new health interventions.
We have presented evidence that involving 'expert researchers' in MT performances supports learning for both audiences and participating researchers; however, persuading researchers to participate proved challenging. Whilst recognizing the busy schedules of researchers, we feel that the importance of involving them at the 'frontline' of engagement cannot be understated. Their presence not only adds credibility to the messages conveyed, but equally important, provides a demonstration of willingness to be part of a community. That MT was a novel and untested initiative for KWTRP may have resulted in a hesitancy on the part of researchers to participate. We hope that a documentation of the challenges and benefits of MT as a credible form of engagement through this publication, coupled with being able to see a video of the outreach, may convey the value of participation to researchers in future MT initiatives. At the beginning of 2020, we had planned to incorporate MT as a component of our broader engagement. However, due to social restrictions caused by COVID-19 pandemic we had to place this on hold.
Implementing MT is expensive and time consuming, and for this reason research institutions with limited budgets for engagement may be advised to prioritise other engagement methods initially. While it may be tempting to think that experienced engagement practitioners may be able to save on selection and contracting costs through implementing inhouse MT initiatives, our experience was that the engagement benefitted considerably from the professionalism and specific skillset of the Jukwaa Arts team. Further, we interpret the audiences' 'wanting more' as an indication of their enjoyment of the performances and a willingness to attend future MT initiatives. It's likely that other forms of engagement, such as community meetings or open days could be more cost-effective, but whether they stimulate similar levels of empathy and enjoyment is unknown. Different engagement approaches have different purposes, and for this reason, as opposed to using it as a single engagement method, we recommend the consideration of using MT as a component of an engagement strategy, to stimulate enjoyment and co-participation in engagement.

Conclusion
MT provided a means of nurturing public engagement with health research. Conveying research concepts through co-created dramatized dilemmas, everyday issues, and moments of joy and conflict, provided an entertaining way for audiences to learn about research, and for researchers to learn about public questions and concerns. Though expensive in terms of time and money, we recommend MT as a complementary and enjoyable addition to research institutes' engagement strategies.

Data availability
Underlying data Data available: Data that may be made available include: data included in the manuscript in form of quotes; summaries of the main themes; and anonymized data transcripts of a half-day reflective workshop with repeat audience members, Jukwaa Arts team and group discussion with KWTRP engagement staff, in keeping with the conditions below.
What uses are applicable: As stipulated in the consent documents, data may be used to support public engagement by other researchers in Kenya or elsewhere, where the nature of the data might be considered relevant. For data not included in the manuscript, the consent form indicates that data sharing will require the approval of the KEMRI Wellcome Trust research Programme Data Governance Committee (see below).
Conditions under which data will be available: Data provided in the manuscript may be used without request but with reference to the full article including the data. Other data will be made available with the approval of the KEMRI-Wellcome Trust Research Programme Data Governance Committee (applications to Data_Governance_Committee@kemri-wellcome.org), only where anonymization can be adequately achieved to protect the privacy and confidentiality of the participants/respondents and any mentioned individuals and institutions, and where the proposed use is seen as relevant to the nature of the data.
Where the DGC recommend this, the national KEMRI Science and Ethics Review Unit may also be asked to approve the proposed use. Conditions for data sharing are outlined in a KWTRP Data Sharing Agreement, including that: -the requestor shall use the data only for the agreed purpose as stipulated in the application form and shall not use the data in such a way that causes damage or distress to the data subjects or communities involved in the research -The requestor shall agree to at all times to keep the data strictly confidential, and ensure that the data users maintain confidentiality of the data -The requestor shall not in any way attempt to seek to discover the identity of data subjects, to compromise or infringe on their privacy and confidentiality of their information.

Extended data
Harvard Dataverse: Replication Data for: Always leave the audience wanting more: An entertaining approach to stimulate engagement with health research among publics in coastal Kenya through 'Magnet Theatre'. https://doi.org/10.7910/DVN/ O9PNJY 22 .
-Extended data File 2: Theory of Change Magnet Theatre approach (PDF).
-Extended data File 6: Coding Framework developed during the analysis of the evaluation data (PDF).

Elizabeth Kimani-Murage
Maternal and Child Wellbeing Unit, African Population and Health Research Center, Nairobi, Kenya

General Comments
This paper presents interesting and very useful information on how participatory visual methodologies approaches can be used to engage the public on research purpose and obtain the views of the public to inform research and the practice of research. The paper details how Magnet Theatre (MT), a form of participatory community theatre, was used to facilitate engagement of public audiences in Coastal Kenya by the KEMRI-Wellcome Trust Research Programme (KWTRP) researchers. The KWTRP used MT to help the audiences understand research and to help the researchers also understand the perceptions of the public regarding research.
KWTRP researchers worked with participatory community theatre artists, Jukwaa Artists, to engage the public with MT. A co-creation process between Jukwaa Artists and the research team, with the engagement of the local community took place to ensure that the Artists understood the research issues and the prevailing views in the community to help develop the engagement strategy. Several magnet theatre performances were conducted in public spaces in the community and attracted huge members of the public. Monitoring and evaluation was later done to evaluate the influences and impact of the magnet theatre outreaches.
The results highlight how the participatory engagement helped the public understand research and dispel misconceptions regarding research that was conducted by KWTRP. It also helped the public to understand the conduct of research and why things happened the way they did. Drama helped the audiences, even those with low literacy to understand research and the purpose of research in an entertaining way.
Participatory visual methodologies are increasingly a great way of engaging the public to help them understand research concepts and the conduct and purpose of research in a way that is most relevant to the public. These participatory methodologies help empower communities and augment their voices -even the most faint voices in the community because they are participatory and less technical. They help to bring the public at par with advances in science, thereby making research more relevant to the public.
Comments based on the review guidelines for methodological papers:

Is the rationale for developing the new method (or application) clearly explained?
Yes. The rationale for the use of magnet theatre to help engage the public in an entertaining way so as to help them understand research and also to help the researchers understand the views of the public is well explained. However, more details could be provided on the factors such as political, socioeconomic and cultural contexts/factors that informed the choice of using MT methodology as opposed to any other participatory methodology.

Is the description of the method technically sound?
Yes. The process of us of magnet theatre to engage the public is adequately described and is technically sound. The description of the method described encompassed three main approaches to public engagement: public communication to transmit information on the topic at hand, public consultation to provide information through a co-creation process, and public participation where two-way interaction between 'experts' and the audience was promoted through the magnet theatre performances. Monitoring and evaluation methodology is also adequately described. However, given that that this was a participatory methodology approach, it would have helped for the team to provide some visual examples in the paper -in the methodology and in the results sections.
3. Are sufficient details provided to allow replication of the method development and its use by others? Generally details are given to enable replication. However, if possible, the authors could provide more information on: What a team should consider when planning and managing MT in public engagement. It would be good to provide more information on whether there was an independent and active evaluation integrated into the design process to give additional reflection. This is because it seems that the team reflexivity, workshops and FGDs were only done towards the end of the process. Also, it would be good to know if the evaluation efficacy and efficiency of using MT?
The team mention that the approach is expensive in terms of time and expenses. More details on this would be useful.
More details on the intended and unintended impacts that MT had on the public, scientists, research institutions and for other actors would be helpful.
Information on any expectations from the audiences and how these were managed would be helpful.
4. If any results are presented, are all the source data underlying the results available to ensure full reproducibility?
The available related data is outlined. We believe the data will be adequately available to those who need to access it.
The themes presented in the results section could be enhanced. For example, we could have themes around: how MT helped deal with misconceptions regarding research; experiences of the public with MT; experiences of the researchers with the engagement.

Are the conclusions about the method and its performance adequately supported by the findings presented in the article?
Yes to a greater extent the conclusions are supported by the data provided. However, this could include information on implications for research and practice.

Clarification queries
Did the repeat audiences represent particular population groups? How were they identified, invited and/or recruited? Related to the above comment on representation, were there any attempts to promote inclusion and diversity by involving marginalised and excluded groups as part of the repeat audiences? If so, how did the project reach out to them? If no, why not? What is the plan for uptake? Will wider uptake be promoted to contribute to wider attitude/mindset change on research processes, beyond the groups that were convened? How will this be done? How will the feedback from the public be used at KWRTP?
Grammatical/formatting edits Need for consistency in use of abbreviation to depict magnet theatre (MT). There are several instances where the full words are used even after the abbreviation was given earlier.

If applicable, is the statistical analysis and its interpretation appropriate? Not applicable
Are all the source data underlying the results available to ensure full reproducibility? Yes Comment: It would be good to provide more information on whether there was an independent and active evaluation integrated into the design process to give additional reflection. This is because it seems that the team reflexivity, workshops and FGDs were only done towards the end of the process. Also, it would be good to know if the evaluation efficacy and efficiency of using MT? Our response: Thank you for your comment. We have addressed this to some extent in the monitoring and evaluation data collection on page 8.

Comment:
The team mention that the approach is expensive in terms of time and expenses. More details on this would be useful. Our response: We acknowledge your comment. In the text on pages 14 to 16 we have outlined ways in which this was time consuming for staff and we feel that since the aim of the evaluation was not to do an economic analysis or an expense analysis cost, we feel that this will give a different feel to the paper and this was more about evaluating whether it was practically doable, how it was experienced by people.
Comment: More details on the intended and unintended impacts that MT had on the public, scientists, research institutions and for other actors would be helpful. Our response: Thank you for highlighting this aspect. In our findings section we have attempted to document all outcomes and influences of MT on audience views about research, both intended and unintended. For example, MT revealed public perceptions of the HIV clinic in Mtwapa, and how this changed as a result of the performances. While it was not entirely expected we didn't present this as 'unintended' consequence, but we do present as an important finding within the text, both intended and unintended.
Comment: Information on any expectations from the audiences and how these were managed would be helpful. Our response: Thank you for this comment. Audience expectations were met to some extent through initial discussions with community leaders about the themes for the dramas as described in page 4.

Comment:
The themes presented in the results section could be enhanced. For example, we could have themes around: how MT helped deal with misconceptions regarding research; experiences of the public with MT; experiences of the researchers with the engagement. Our response: Thank you for your suggestion. We made several attempts at laying the data out with the themes you presented but we found that our way provided the clearest framework for sharing our findings.

Comment:
The conclusions are supported by the data provided. However, this could include information on implications for research and practice. Our response: Thank you for this comment. In our conclusion we have emphasised the value of co-creation in the design and implementation of magnet theatre for engagement. We also recommend the use of MT to complement other engagement strategies.
Comment: Did the repeat audiences represent particular population groups? How were they identified, invited and/or recruited?
Our response: Thank you for this comment. We chose the MT approach in order to attract an urban population in Mtwapa and the more rural population in Mida. We didn't specifically target any particular marginalized group, but considered the fact that poor urban people and poor rural people in Kenya are marginalized in the sense that they are not able to access or be exposed to research information and that's why we went out to these specific places. This is explained in the 2 nd paragraph in the methods section on page 7.
Comment: Related to the above comment on representation, were there any attempts to promote inclusion and diversity by involving marginalised and excluded groups as part of the repeat audiences? If so, how did the project reach out to them? If no, why not? Response: As is consistent with many community theatre approaches, the implementation team targeted public audiences who happened to be present in the vicinity of the site at the time. As such this was an organically diverse group spanning a rage of ethnic and religious groups, ages and gender. A broad description of the audience composition has been given in the findings section.

Comment:
What is the plan for uptake? Will wider uptake be promoted to contribute to wider attitude/mindset change on research processes, beyond the groups that were convened? How will this be done? How will the feedback from the public be used at KWRTP? Our response: Thank you for this. We have described we had intended to keep MT as a component of engagement in the broader strategy, but we have not done so because of COVID-19 restrictions, as described in page 16.

Comment:
There is a need for consistency in use of abbreviation to depict magnet theatre (MT). Our response: Thank you for pointing this out. This was noted and the necessary edits were made throughout the manuscript.
including consideration of the logistics that may enhance or impede engagement through this approach. The benefits of Magnet Theatre are clearly demonstrated and the challenges in implementing the approach considered. The paper focuses largely on the concept of this approach rather than on the actual impact of the approach in terms of increased knowledge and understanding of health research. Whilst it emphasises the overarching benefits of increased familiarisation and trust in health research conducted by the Research programme it does not attempt to specifically measure the extent of the influence of MT on research engagement. As such the paper is a useful and interesting introduction to the relevance of theatre as a tool for learning about research but does not provide evidence of its instrumental utility as a tool for increasing knowledge of research. There is emphasis on the cost (time and monetary) of implementing the approach which leads the reader to conclude that such an activity may be worthwhile in settings such as KWTRP defined by long-term research and therefore community engagement needs but perhaps less relevant in settings where there is less long-term commitment to the research population. I would suggest that the approach is highly suitable for programmes such as KWTRP, and indeed partners across the WT overseas Africa and Asia programmes but perhaps less so for other settings, unless the purpose of the MT itself is to engage communities about a specific research aim such as the examples cited in Canada, Bangladesh etc. This type of generic approach to familiarising communities to health research is, despite this, a valuable, entertaining and effective method.
There are a couple of areas that could have been expanded in the paper, providing greater detail in order to promote the consideration of this approach amongst those who may be less convinced. Firstly the provision of a more structured evaluation to promote some measurement of spread including a breakdown of audience by sex and age, a longer term follow-up with community members as opposed to just those who were repeat audience members (by definition likely to be more engaged with the institution) to explore the sustainability of the method on future engagement specifically and the potential use of the MT method to 'spread' messages in the wider community (reflecting and potentially countering the spread of rumours in these settings). A more explicit comparison across the two sites might have been interesting especially given their descriptions as relatively more or less exposed to health research conducted by KWTRP. Whilst this was considered in the follow up evaluation (FGDs and workshop) it wasn't explicitly evaluated more formally.
However, I acknowledge the challenges in conducting a more rigorous evaluation of the impact of MT and recognise that the authors did not set out to measure impact and the paper provides an interesting and highly relevant perspective on the use of MT as a general and effective engagement tool for increasing trust and familiarity with health research in a long-term research setting.

Is the work clearly and accurately presented and does it cite the current literature? Yes
Is the study design appropriate and is the work technically sound? Yes

If applicable, is the statistical analysis and its interpretation appropriate? Not applicable
Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Yes and we have amended the final paragraph of the discussion.
Comment: I would suggest that the approach is highly suitable for programmes such as KWTRP, and indeed partners across the Wellcome funded Africa and Asia programmes but perhaps less so for other settings, unless the purpose of the MT itself is to engage communities about a specific research aim such as the examples cited in Canada, Bangladesh etc. This type of generic approach to familiarising communities to health research is, despite this, a valuable, entertaining and effective method. Our response: Thank you for your comment and for acknowledging the place of MT in engagement.

Comment:
Firstly the provision of a more structured evaluation to promote some measurement of spread including a breakdown of audience by sex and age, a longer term follow-up with community members as opposed to just those who were repeat audience members (by definition likely to be more engaged with the institution) to explore the sustainability of the method on future engagement specifically and the potential use of the MT method to 'spread' messages in the wider community (reflecting and potentially countering the spread of rumours in these settings). A more explicit comparison across the two sites might have been interesting especially given their descriptions as relatively more or less exposed to health research conducted by KWTRP. Whilst this was considered in the follow up evaluation (FGDs and workshop) it wasn't explicitly evaluated more formally. Our response: Thank you for this pertinent comment. As explained in previous responses the evaluation did not set out to formally compare the two sites. However, we do present evidence of the differences, for example, in Mtwapa where there is an HIV clinic which influenced community perceptions about KEMRI. We have added Table 1 on page 10 showing the total attendance for all the outreaches in the two sites.
Competing Interests: No competing interests.