Co-Creative Visual Poetic Inquiry for communicating lived experience of HIV self-management and self-care

Abstract We introduce our innovative analytic approach that draws from poetic and visual forms of inquiry: ‘Co-Creative Visual Poetic Inquiry’. We report on the development of this approach in the context of a design research project that explored individuals’ lived experiences of self-care practices for living well with HIV. We present a set of visual poems that were co-created between the research team members as a means to capture and communicate idiographic insights from this project. We aim to contribute methodological insights through a case study, about how Co-Creative Visual Poetic Inquiry can support the qualitative analysis and dissemination of research. We highlight the value of the approach for helping researchers co-define a research agenda early on in a project, in dialogue with members of the research population. We also show how poetic, visual and time-based forms of representation can be used in individual and collective analyses, for phenomenological engagement and to communicate insights to stakeholders in ways that retain emotional and performative qualities of expression. We conclude by critically reflecting on the scope and limitations of our novel combination of methods, and encourage other design researchers in health and wellbeing research to appropriate and develop this approach.


Introduction
In this paper, we introduce an innovative analytic approach that draws from poetic and visual forms of inquiry: 'Co-Creative Visual Poetic Inquiry' (Table 1). Our approach has been informed by a maturing discourse on Poetic Inquiry (Butler-Kisber 2010; Johnson and Wimpenny 2018), which involves the use of poetry as a qualitative research method. Also drawing upon visual and time-based forms of communication, we, the three authors of this paper, co-created a set of visual poems for a design research project. The project aimed to understand individuals' lived experiences of self-care practices for living well with the Human Immunodeficiency Virus (HIV), to inform digital forms of support. Herein, we reflect on the strengths and limitations of using creative modes of data analysis for pursuing shared understanding within a research team holding interdisciplinary skills. We posit that Co-Creative Visual Poetic Inquiry provided us with the means to collaboratively explore and make sense of the research data. This paper presents the visual poems that formed our analytic response to data collected from sensitizing workshops conducted for our first study in the project.
We previously reported on the workshop method for this study (Claisse, Kasadha, and Durrant 2020), and how we used Found Poetry (Butler-Kisber 2010) in part of our data analysis to compose a set of 'Individual Poems' that each conveyed sense making on individual participants' experiences. Found Poetry is a method whereby participants' utterances are rearranged and reframed as poetry. In this paper, we expand our report of combining Found Poetry with Generated Poetry and visual forms of inquiry (including collage, illustration and animation). Generated Poetry is another method whereby researchers use their own words to describe their interpretation of research data (Butler-Kisber 2010). Herein, we report on the novel analytic approach to Poetic Inquiry that we developed, offering methodological reflections.
The paper is structured as follows: first, we provide context about the project and conduct of the sensitizing workshops. We then describe how our analytic approach is informed by poetic and visual forms of inquiry (Butler-Kisber 2010). We report on the particularity of our co-creative analytic steps where we drew from each other's multi-disciplinary skillsets, including poetry, visual communication and animation. Two of the poems created by Kasadha are presented in this section to illustrate our use of Poetic Inquiry for capturing and communicating idiographic insights. We then demonstrate how Poetic Inquiry was enhanced by additional creative skillsets that enabled us to further co-analyse and disseminate empirical understandings that guided us within the continuing research project, towards the design of follow-on studies and design prototypes about providing digital support to Table 1. Analytical steps and authors' involvement in doing Co-Creative Visual Poetic Inquiry.

Co-Creative Visual Poetic Inquiry
Author 1 (Claisse) Author 2 (Kasadha) Author 3 (Durrant) Poetic inquiry Thematic analysis X X Found Poetry X X X Generated Poetry X Visual inquiry X X people living with HIV. To help illustrate our approach, we include examples of the Visual Poems and the Visual Poems Animations that are available to view on the project website (https://intuitproject.org/poems/), and offer insights on discussing the poems with participants and stakeholders. In our Concluding Discussion, we critically discuss the transferrable value of our novel approach for design researchers working in health and wellbeing. We draw out methodological reflections with practical guidance on how Co-Creative Visual Poetic Inquiry may be adapted by others studying complex socio-medical subjects. Our objectives in this paper are: (i) to advance understanding on co-creative methods for conducting design research on health and wellbeing, through a case study of Co-Creative Visual Poetic Inquiry in practice; and (ii) to demonstrate the potential value of time-based media production for disseminating performative research outputs to stakeholders in a way that efficaciously supports knowledge exchange. In doing so, we aim to promote the use of a co-creative mixed methods to analytic practice, whereby analytic procedures are enhanced by creative skills and techniques (e.g. collage, sketching, animation) to facilitate co-production in a way that recognizes the multi-disciplinary skillsets of each researcher.

Research background
This study formed part of a broader interdisciplinary programme of work named 'INTUIT: Interaction Design for Trusted Sharing of Personal Health Data to Live Well with HIV' (https://intuitproject.org/). A key aim of the project was to gain understandings of routine HIV self-management and selfcare to inform the design and evaluation of digital tools that better facilitate the sharing of personal health data with healthcare professionals and others in networks of care, to self-manage HIV as a long-term condition. By focussing on understanding lived experiences of HIV self-management and selfcare, we sought to design appropriate interactive tools to support people in living and ageing well with HIV, and achieving Quality of Lifea current global goal (Lazarus et al. 2016).
Early on in this project, we used participatory methods to conduct sensitizing workshops that explored, as a generative endeavour, how participants who live with HIV define self-management and self-care and to identify key matters of concern. We conducted three workshops in 2019, one at the offices of the Terrence Higgins Trust (THT) in London, and two at Blue Sky Trust (BST) in Newcastle upon Tyne. In total, 17 adults (11 women; 6 men) living with HIV participated. Participants were guided through a series of creative and 'hands-on' activities, including badge or brooch making, hand-drawing and presenting maps of networks of care ( Figure 1). These activities invited individual and collective sense making on the following phenomena: HIV and Self-identity; HIV Self-care; Networks of care (i.e. who else may be involved in an individual's practice of self-care); and Experiences of receiving clinical care.
Workshop outputs informed how we practiced Research-through-Design (Durrant et al. 2017;Stappers and Giaccardi 2017) in the remainder of the INTUIT programme. Significant for our argument herein, our combined cocreative methods arguably supported a continuing dialogue with members of our research population and stakeholders, for communicating early outcomes in order to co-develop further studies with their input. In this paper, we provide a reflective account detailing our development and use of Co-Creative Visual Poetic Inquiry, for analysing workshop data and disseminating study findings.

Approach
Following a phenomenological methodology, we focussed on empirically understanding the felt experiences of our participants. Our co-creative approach builds on 'Collaborative Poetics' (Johnson and Wimpenny 2018), recently developed in the social sciences, in which poets and social scientists work together: The approach focuses on expounding the subjective, lived experiences of coresearchers, and thus also harnesses a third base of expertisethat of personal experience 1 In our case, the three authors worked closely together to devise, conduct, analyse and disseminate the research. Our team comprised of a performance poet who was also a peer researcher on INTUIT, and two interaction designers, one also holding expertise in graphic design, the other also holding expertise in visual arts and social psychology. A peer researcher is a person who shares an experience with research participants as well as holding technical research skills and the insight of lived experience. Within the context of INTUIT, the peer researcher also lives with HIV.
Co-creation in our team meant that we drew from each other's multi-disciplinary skillsets to pursue and progress our research. Following our collective qualitative analysis of the workshop data (Braun and Clarke 2006), we drew inspiration from Kasadha's experience in performance poetry and explored Poetic Inquiry (Butler-Kisber 2010) for capturing participants' lived experience. We drew from the creative design practice and design research skills of Claisse and Durrant to augment the metaphors and evocative utterances captured in the written poems with multimedia: Claisse used her skills in graphic communication to produce visual interpretations of the poems; and Durrant used her animation skills to afford the performative and timebased communication of the visual poems for dissemination. Through each of these creative, analytic steps, the medium afforded a process and output that retained emotional resonance and performative quality in expression. The Visual Poem Animations were shared at an online meeting with a large group of INTUIT stakeholders, who were unable to convene face-to-face due to the COVID-19 pandemic. Overall, these were well received, indicating potential value in this multi-media form of dissemination. Resultantly, we have recognized the qualities of a Visual Poetic Inquiry approach and how the co-creative methods effectively supported both collaborative sense making and knowledge exchange through forms of communication that arguably retain a closeness to the original voiced expressions of experience.

Poetic inquiry
We experimented with Poetic Inquiry whereby poetry is used as an analytic method for examining research data (Butler-Kisber 2010). We refer to the term 'inquiry' in terms of analytic inquiry and as a practice of making sense of our data. There are two ways of conducting Poetic Inquiry: 'Found Poetry' and 'Generated Poetry' (Ibid.). We have found value in using both methods in compliment, to make sense of and communicate participants' lived experiences captured at the workshops. We now summarize three interlinked steps of our Poetic Inquiry process: Thematic Analysis, Found Poetry and Generated Poetry.

Thematic analysis
To begin, Claisse and Durrant annotated and coded transcripts of the workshop data, using Inductive Thematic Analysis (Braun and Clarke 2006). This involved producing a visual map of initial themes and subthemes, printed on paper and laid out on a table together with printed codes and coded participant excerpts. The map was populated with 'cut-out' paper strips of codes and excerpts; themes were developed through visual sense making and clustering content. We first generated a list of initial themes supported by coded excerpts. Next, Claisse and Kasadha further coded the data guided by initial analytic themes; this engagement included paying special attention to non-verbal cues in participant utterances (such as changes in pace or pitch and pauses), also captured in the transcripts. Listening back to the workshop recordings supported this process, enhancing our analysis in two ways: it made us feel closer to the data whilst helping us reflect on participants' emotive experience; we felt confident and able to use the thematic insights as provisional and aspirational materials for the creation of poems. After sessions between the three of us, we finalized a summary of the themes supported by key excerpts and annotations.

Found poetry
We now describe our use of Found Poetry in analysis to produce a set of 'Individual Poems', through which we intend to convey a participant's unique experience of HIV self-management and self-care. For this report, we focus on how we worked with the accounts of three participants to further reflect on one of the themes: Sense of Self. These accounts reflect individual voices and the diversity experiences of people living with HIV that was reflected in our data.
Following the Found Poetry method, Claisse and Kasadha identified phrases and words in the transcripts for each selected account, that reflected a unique journey and experience around self-management. Kasadha then organized this text in a poem, choosing to work with free verse in order to be led by participants' words, rather than feel restricted by a poetic structure. Repetition, as captured, and breaks in lines and spacing within the text, was intended to reflect the participant's change of subject, or their own pauses when speaking. The text was also ordered for accessibility and ease of understanding by diverse stakeholder audiences. The last step in the process was naming each poem; the title reflected the participant's expressed self-identity or characteristic of self-care practice portrayed in the poem (e.g. 'The Teacher', or 'To Persevere' (Figure 2), respectively).

Generated poetry
We complemented Found Poetry by using Generated Poetry to produce a second set of poems, which we refer to as 'Collective Poems'. Whilst Individual Poems focus on an individual's unique experience, each Collective Poem captures images and sentiments expressed by multiple participants. Kasadha responded to the outcomes of the original Thematic Analysis to create these poems, representing emerging data themes and sub-themes. She primarily used her own words to interpret and synthesize participants' accounts of self-care and care provision; however, some phrasing was incorporated from transcripts and the account of findings.
Four Collective Poems were composed to communicate thematic insights: 'Meanings of Self-Care', 'Understanding Care Provision' and 'On Communication and Sharing'. The 'Meanings of Self-Care' poem conveys a reflective internal dialogue, and the journey to finding support within a clinical setting ( Figure 3). 'Understanding Care Provision' conveys the importance of peer support and advice from others living with HIV. Twelve out of 17 participants had stated that they would like to discuss relationships more with the medical team providing their HIV care. In response, the first part of 'On Communication and Sharing', conveys peer-to-peer learning and the value of knowledge exchange outside of a clinical setting; the second part conveys a voice asking their healthcare provider about relationships and a public health campaign.
All of the Collective Poems were composed in free verse, except for 'Meanings of Self-Care', in which the stanzas alternate between a Haiku repetitive chorus and an imperfect rhyme stanza; this alternative structure for the latter was used to convey the sub-theme, 'Routines and goal setting enhances self-care'stanza structure representing routine structures. Other metaphors were reflected in structure. In 'On Communication and Sharing', three voices are conveyed; this polyvocality was intended to reflect the interactions within the workshops and more broadly experiences of HIV support groups. The Collective Poems are available to view at this link: https://intuitproject.org/poems/. These show how a poem's layout and structure (both in terms of use of space on the page and use of rhyme and form) has been used to convey the analytic theme(s). Whilst Butler-Kisber 2010 suggests that Generated Poetry may afford more creative freedom than Found Poetry (Butler-Kisber 2010), we have found that Generated Poetry can feel more overwhelming or daunting because the poet must find the words from scratch. With Generated Poetry, researchers are invited to use their own words to convey thoughts, images and sensations that may speak to many, relying on their personal experiences of conducting the research (Yuan and Hickman 2015).
In our study, we found that the two methods enabled us to work in different ways with our data: we used Found Poetry to stay close to an individual's voice and unique journey, and Generated Poetry to synthesize themes that spoke to participants' collective experience of HIV self-management and self-care. We found it productive to use both methods, to work with raw data and Thematic Analysis outcomes. As with Butler-Kisber (2012), we found Poetic Inquiry to enhance the qualitative analysis, by: enriching our interpretative work with metaphors; co-creatively engaging with ambiguity in generative acts and adopting an exploratory tone rather than a summative one, for multiple readings of the data; and acknowledging researcher reflexivity and subjectivity. These interpretative qualities were further developed through using Visual Inquiry.

Visual inquiry
We used Visual Inquiry to creatively respond to the written poems. In art research, Yuan and Hickman (2015) build on Butler-Kisber's work (2010,2012) to encourage 'Poetic-Visual Inquiry'. These authors draw the connection between the poetic act and how 'visual inquiry acknowledges that we can express our ideas and enrich our understanding of the world through visual materials' (Ibid.,9), highlighting the difference between using visuals as supplementary illustration to written text and their application as a means of storytelling, argumentation and analysis. In our study, we engaged with visual communication in the latter mode, drawing from Collage Inquiry, a visual-based method valued for communicating nonverbal and latent themes (Butler-Kisber 2010). Arguably, composing collage clusters (compositions) from visual data can help researchers make sense of it (Ibid., 102): Collage clusters can help to conceptualize dimensions of understanding that were previously unconscious, and how collage creation can be a way of making thoughts concrete, facilitating the thinking, writing and talking about the inquiry.
Butler-Kisber (2010) also describes how, through Collage Inquiry, one might work from 'the heart to the head', through a process that is inherently fragmentary and embodied.
Collage Inquiry has been used in different ways and we found limited examples that combined both visual and poetic modes of representation. One example (Yuen 2016) captures the use of both in participatory research with Aboriginal women to study experiences of healing: Poetic Inquiry guides the researcher's collage response. Visual materials for creating the collage included raw data; Yuen includes visuals of body maps created by the women, and describes how this process led to polyvocal understandings whereby a single metaphor emerged into multiple metaphors and meanings. Gillies (2007) provides another example of Visual Poetic Inquiry, using visual sketching to convey experiences of transition for university graduates with disabilities. In this case, one illustration was generated to represent metaphors and themes inspired by data; the author describes exploring visual representations after reading poems, finding that the visual and poetic components added depth and clarity of understanding to one another. Most recently, Lab4Living launched a book with Visual Poetry to communicate psychosocial aspects of living with Dementia, comprising 'beautiful illustrations which take the reader on a visual journey' 2 . These three examples demonstrate the identified value of combining visual and poetic modes for making sense of and disseminating research.
We aim to build on this work by reporting on how we collectively enhanced our analysis with a methodological innovation. In our multi-stage analysis, we drew on the multiple creative skills that we each hold in addition to our research skills that are formally aligned with our roles on the INTUIT project. Our involvement as 'creative practitioner-researchers' compares to the approach of Sjollema and Yuen (2017), in which they commissioned a poet external to the research team to engage, and the Lab4Living researchers who contracted an illustrator for their book publication.
By drawing on our creative skillsets, we have recognized the hybrid nature of our professional expertise, and the relevance of reporting on these hybrids for considering co-creative analytic practice.
We have further extended Visual Poetic Inquiry methods by considering time-based communication in the analysis and dissemination of findings. Video production incorporating animation with audio and visual media, delivered research outputs that are performative in their creative expression. We now expand on these novel steps.

On Co-Creating Visual Poems
We describe how we responded to the two set of poems using visual and time-based media. We further reflect on the effectiveness of the multimedia production of the visual poems for 'performative' dissemination.

Individual poems for communicating one's unique experience
A set of three Individual Poems were selected for further engagement; these are available to view online (https://intuitproject.org/poems/). The poems represent the sample in their diversity, plus the analytic themes; we also selected one individual's account from each workshop. Each poem was based on a participant's personal journey of living with HIV. The first poem 'To Persevere', communicated challenges of living with comorbidities such as HIV and Type 1 Diabetes (see Figure 2). The second, 'The Teacher', conveyed the importance of connecting with others and sharing experiences as a source of self-empowerment. The third, 'Becoming the Advocate', illuminated emotional growth and self-reflection resulting from building resilience and growing confident with HIV.
Claisse responded to the written poems using visual methods; she made mood boards and sketched drawings to develop a visual analytic response ( Figure 4). As the poems focussed on an individual's experience, it was important for the visual response to personify expressed perspectives, beliefs and values portrayed in the poems. This process resulted in three visual 'figures' that accompanied each poem: 'The Teacher'; 'The Perseverant' and 'The Advocate' (Figure 4).
Reflections by Claisse (Figures 5 and 6) convey how she developed a visual language constituting empathetic engagement and further consolidation of insights.

Collective poems for representing the shared expressions of a group
Claisse used collaging and bricolage in her visual response to the Collective Poems. She searched on the web for images with free copyright and worked the media into collage clusters. The structure and form of each poem inspired the collaging method. In the 'Meanings of Self-Care' poem, one collage cluster was created for each stanza (a group of lines), whilst only one collage was created in response to 'Understanding Care Provision'.
Participants voiced that receiving clinical care was of central significance for self-care. The poem 'Meanings of Self-Care' communicates how self-care interlinks with clinical care, and is closely linked to setting up routines, taking medication at the right time and adhering to treatment.  The collage cluster was inspired by the first stanza of the poem (Figure 7).
Claisse represented the context of the consultation where the patient is waiting to speak and share their experiences, considering the dynamics of patient-clinician interaction. She composed a view of the clinician's desk with symbolic artefacts placed on it, including a stethoscopea tool known for listening to a patient's heart and breathing. Traditionally used in clinical consultations, this stethescope became a metaphor for listening closely to a participant's feelings. Language barriers expressed by participants for sharing personal information with clinicians (e.g. where English is not a person's first  language) were also represented, through sketches of a mouth opening and closing (Figure 7).
With this collage, Claisse responded to experiences of self-reporting at the clinical consultation: 'I wait, notes in hand'. A patient is represented holding their smartphone with notes to share with their clinician. Claisse chose to include a phone instead of a notebook to invite the viewer to think about what it might be like to use a phone in this context, portraying the patient looking down onto it. In dialogue with Kasadha and Durrant, the visual composition was explored to highlight potential tensions of using a smartphone at the consultation. Making the collage provided us with an interpretative space for considering how interactive technology including the smartphone may support self-reporting and sharing with clinicians in routine communication. We also started to think about interface design and used collaging to imagine what could be featured on the screen. This re-engaged us with what our participants shared at the workshops about having different preferences for capturing and sharing their personal health data in networks of care. The resultant image intends to convey how such a device may hinder patient-clinician interaction as it could become a potential distraction instead of supporting facilitation. Figure 8 features a collage created for the poem 'Understanding Care Provision' (https://intuitproject.org/poems/). The poem reflects what participants expressed about the importance of peer support for HIV self-management and self-care.
The collage illuminates insights about peer support as a source of empowerment, for reducing isolation and building a sense of community. Claisse created one collage that portrays a place where peer support is facilitated, communicating the qualities and values described by participants, such as feeling at home, in a safe space. Claisse used colours and textures that portrayed feelings of warmth and a welcoming atmosphere, inspired by one of our workshop settings. We did not include any reference to technology in our collage response, emphasizing that peer support is facilitated through face-to-face interaction.
In sum, the Individual and Collective Poems acted as a catalyst for visual sense making and reflection. The Visual Inquiry was supported by dialogue between the three authors as researchers and creative practitioners, honing empathic understandings of participants' experiences, and highlighting metaphors captured in words, lines, and phrases that shone forth to us in the analysis.
The third creative analytic step was to draw on time-based communication to support the performative dissemination of the research outcomes. This resulted in Visual Poem Animations (https://intuitproject.org/poems/), video files that, as discrete media artefacts, could be further shared for sustained engagement with the findings.

Sharing the poems with participants and stakeholders
As part of our co-creation process, we intended to share the Individual Poems with the participants who inspired them, before finalizing them. We were able to share them with two of the three participants. As the researcher-poet could not be collocated for the meeting with participants, she audio-recorded the poems for presentation and feedback. Claisse met with the participant who inspired 'The Teacher', played the recording of the poem to him, inviting his feedback. He described the talent of the researcher-poet to capture 'the message' of his study contribution in a poem that was 'short' and 'sharp'. He saw the poem 'as a way to get the message out there', with potential 'to change world'. He specifically liked the title that alluded to his background of working as a teacher. He suggested to change a word that had been misheard, which was made. Claisse and Durrant met with the participant who inspired 'To Persevere', and Kasadha joined via video-conference. After hearing the poem, the participant recognized her words and began sharing names of colours and feelings that the poem evoked for her. These guided our visual response. This participant also suggested we edit a word that had been misheard, which we effected. Sharing the poems in this way not only provided an opportunity for shared sense making but also demonstrated to our participants that they had been 'heard' and how their words had been interpreted.
We recognized the value of the recorded performance of poems in knowledge exchange, and selected three of them (1 Individual, 2 Collective Poems) to be developed into animations by Durrant. Selection of poems was guided by our analysis and the salient themes. Creating animations also supported research dissemination at two upcoming stakeholder project meetings. Due to the COVID-19 pandemic, these meetings were organized online using a video-conferencing platform.
It felt important to Durrant that the composition of the animations, in terms of content and pace, was guided by the readings of the poems by Kasadha. First, Kasadha audio-recorded her performances of each poem. Durrant imported this media into a video editing programme and used it as the 'backbone' for arranging and editing with visual media. In places, Durrant edited the recorded audio track to increase pauses for reflection on the mixed media presentation of spoken word and animated imagery.
Durrant worked with Claisse to co-create visual content, working with the collage context, and creating additional files constituting sequences of frames for a moving graphical image. To illustrate, we can refer to the sequence of images in Figure 7 of a mouth opening and closing 'Meanings of Self-Care'; this sequence was extended and animated to render the mouth moving in time. Selected graphical elements of the visual poems were developed into animated sequences, to enhance the metaphorical expression of felt experience as captured in the findings. In 'Understanding Care Provision', the sketched blue lines reflecting a blue tone, depicted in Figure 8 (bottom left hand corner), were developed into an animation sequence that portrayed them moving like air or clouds. Other graphics in the poems were developed and animated to move and vibrate. With 'The Perserverant' animation, Durrant animated the fine line drawings-intended by Claisse to portray 'delicate but strong feelings' ( Figure 5); and where line-drawn hills were intended by Claisse 'to symbolize the ups and downs of the journey', these lines were animated to depict gentle, undulating movement. We intended to communicate a liveness or anima in the images that would resonate in synchronicity with the spoken words in the recordings. Table 2 includes selected verses from 'Understanding Care Provision' to show how time-based communication through animation was intended to enhance the performative dissemination of the visual poems. Creative decisions and metaphorical expressions that were made in producing animations were grounded in a sustained analytic engagement with the research findings. We decided not to animate text in the poems (known as Kinetic Typography), to ensure that the animations accentuated rather than detracted from the performative readings. Resultant animations were produced as MPEG-4 files, a widely used digital multi-media container format.
The three animations were presented at the INTUIT stakeholder meetings in Spring 2021. Each video was played online in a plenary meeting context on a teleconference platform. Attendees ranged from academics, healthcare professionals, representatives from HIV organizations, and HIV Community Representatives who live with HIV. We invited feedback on these presentations and they were found to resonate, indicating potential value in this form of dissemination. A UK University lecturer commented on how the animated visual poems enabled her to 'stay centred' on the represented experience, adding: Allows us to keep the emotion, whilst not requiring people with experience to (re)rehearse pain they have felt of feel. Remaining in a relationship with their experience even when at a distance -all very, very powerful -politically, discursively. Keeps design anchored in experience, particularly where you need broader stakeholder input to progress and embed design.
Making reference to a Collective Poem, an attending Community Representative who also took part in one of the workshops commented that she found the poem 'really moving', because she could 'recognize her own experience embedded within others'. She added how the visual poems 'gave voice' in a way that other research approaches often can't, 'making co-constructed narratives more visible'. An HIV clinician and academic added that she 'would like to see research pursued through this approach presented at clinical conferences, to demonstrate the value of qualitative, creative methods in healthcare research, beyond interviews and focus groups'. Reflecting on how the Visual Poem Animations were received, we have found value in this format for disseminating research outcomes that have a performative quality, widely and effectively, when live performance is not always possible.

Concluding discussion
We have reported on our development and use of a novel analytic approach dubbed 'Co-Creative Visual Poetic Inquiry'. This approach enabled us to Birds fly off the peak of the mountain to enter the room. They move down and then up.
A sense of community, togetherness; going through things together. We hope you can find a home here A second one, a first one, a new one The kettle is depicted boiling water; the cup is simming hot.
Feelings of warmth and comfort associated with making a cup of tea.
It may not feel like it now, but one day You might be sharing your story The lamp turns on in the room and spotlights the chair.
Forms of empowerment and acts of listening via sharing experiences.

It's okay Cry Scream
Another lamp turns on to spotlight the cushion that shakes slightly.
Embodied sharing in a safe space; physicality of emotional expression; coming to terms with things.
respond creatively and collectively to interpretations of a qualitative data set, which was collected in the early stages of a design research project. The paper's three authors drew on their skillsets as creative practitioners to augment their qualitative research skills; they explored the application of these skills in a collective analytic process that was generative. We have portrayed a logical progression of practice, from composing the poems in response to the thematic analysis, and then creating collages and animations in response to these poems, and also explicated how, at each creative step, we sustained our engagement with the data and with those who took part in the workshops. Arguably, the co-creative engagement helped us foster a 'dialogical understanding' (Durrant et al. 2017) with each other and with our research subject. Key to this, we have concluded, is how the practice and outputs afforded communication of emotion in analytic expression.
When working with a qualitative dataset, a researcher can often quite quickly feel distanced from the data collection context; transcripts and the process of identifying and writing about thematic insights can also feel abstract; this is our experience. We found using creative methods in the reported study valuable for retaining a closeness to the data at each analytic step. This is epitomized by how participants' utterances (including inflections of tone, emotional resonance) have been retained and rehearsed through the performative media and its time-based communication; how a stakeholder engaging with a presented animation recognized her experience in the media.
As researchers who each hold professional skills sets in research methods and in creative practice, we had found it opportune and intuitive to draw on our creative skills in the analysis, and to connect with making sense of participants' experiences through creative interpretation. Co-analysis and empathetic connection were also deeply shaped by Kasadha's readings of the poems who shared lived experience with the research population. This connection was achieved through an engagement with metaphors; creative practice supported translation work with these metaphors, to give them form in ways that made them feel more live and relatable.
We may also critically reflect on the potential transferability of our novel combination of methods to other researchers considering using Visual Poetic Inquiry or other creative skills for research analysis. We appreciate that not all research teams are interested to adopt this approach will also have skills in creative practice; this is a potential limitation of the approach. Within the social sciences, researchers have collaborated with creative practitioners to conduct Poetic Inquiry and produce creative outputs (e.g. Sjollema and Yuen 2017). However, such collaborations have typically taken place after data has been collected and analysed (Ibid.). One consideration for researchers exploring this approach in future studies is to consider opportunities for involving a creative practitioner early in the research process, whether directly involved in and informing procedures of data collection and analysis, or to facilitate or observe them. We highlight the rich scope of opportunity for researchers to consider employing other creative skills in analytic endeavours, besides poetry, graphic design and animation: to work with metaphors in an interpretative stance, in dialogue with others.
A third related consideration is about hybrid professional identities supporting interdisciplinarity. Our project required multi-disciplinary expertise for conducting research that will also deliver interaction design insight; Claisse and Durrant were required to be both researchers and creative design practitioners. When practising design, we typically find it important to be able to draw upon research outputs that inspire a creative response. Whilst traditional forms of dissemination such as research papers do inspire this, we find that multimedia renders findings more tangible as resources for design. In writing, we mindfully build on a mature discourse in Design Studies and related discourses that critically reflect on the value of practice research and creative methods in Design Research (e.g. Frayling 1994). Visual, video and multimedia formats have also gained much traction in Ethnography and the social sciences, (Pink 2002). In Digital Health Studies, the value of qualitative methods for focussing on sensory, embodied interaction and felt experience is increasingly demonstrated, researchers incorporating creative skills and sensibilities (Lupton 2016). In the HIV field, there is a growing body of interdisciplinary work utilizing creative methods (van Rooyen et al. 2019).
Furthermore, we link our insights with a Research-through-Design discourse that considers alternative formats for disseminating practice research (Durrant et al. 2017); highlighted here is the hybrid professional identities of practitioner-researchers, typically working in teams to conduct research for or through design (Ibid.). We offer a worked example of putting a hybrid skillset into practice, for a co-creative analytic endeavour and studying a complex socio-medical subject.
Leading from this, we consider two further methodological reflections. The first is about finding creative expression useful for reflexive analytic practice. Drawing and Collaging are recognized techniques for what Donald Sch€ on (1984) describes as reflective practice. Through Co-Creative Visual Poetic Inquiry, we co-created artefacts that could then 'talk back' to us (Sch€ on 1984). We co-developed representations of our analytic process, involving phases of individual plus collective analyses. After Butler-Kisber (2010), our practice and its outcomes helped us to make sense of our design space, giving form to dialogue; and furthering reflection on what we chose to illuminate about participants' lived experience and why. We raise a critical reflection on this interpretative mode, and of working performatively and iteratively with metaphors and multimedia: in rendering the visual poem animations, we may risk inadvertently closing down an interpretative space for others to engage with the research.
Our final methodological reflection is to raise the visibility of another discourse that this work may incrementally advance, namely Performance Inquiry: research inquiry through Performance (Butler-Kisber 2010); we signpost the arguably underexplored potential of performative modes of inquiry for the subject of health and care, especially for focussing on understanding embodied and emotional experience.
Overall, we found that 'Co-Creative Visual Poetic Inquiry' was useful at the early stage of a project to help us co-define a research agenda and identify matters of concern, in dialogue with members of the research population. Whilst the visual poem animations are research outputs in their own right, they have also been important for helping us shape research-through-design methods for follow-on studies.
To conclude, we contribute an account of a recent sensitizing study of self-care for people living with HIV in the UK, in which we devised a novel, co-creative approach to data analysis that has delivered emotionally resonant media for performative dissemination.

Notes
Bakita Kasadha is Peer Researcher at Terrence Higgins Trust, a UK HIV and STI charity. She is also a performance poet.
Abigail Durrant is Professor of Interaction Design in Open Lab at Newcastle University, predominantly working in the field of Human-Computer Interaction.