In the second of the DYOC series, Aparna Maladkar writes about data visualisation:
A number of motivations and factual instances can drive the use of visualisation, and in this second ‘Draw’ your own Conclusions Series (DYOCS) blog, we will consider the first question within the ‘Four W’s of Visual Research’ (4-Ws) diagram tool, “When to use visual graphics?” The first DYOCS blog looked at the 4-Ws tool and some historical examples. The blog is split into two parts, Part I outlines theoretical examples and ethical considerations, while Part II will look at factual examples.
Feel free to skip the written words, and jump to the last section ‘Summary’ and the image ‘Key to WHEN’ (Figure 7), which visualises the important points discussed within the two parts of this blog to embark on your visualisation journey.
The key reason for William Playfair’s invention in 1786 of the first known bar and pie charts was the lack of data.[1] John Snow’s 1854 pioneering cholera map that completely changed the way the world saw epidemics, was designed due to Snow’s scepticism about the miasma theory which was believed to have been responsible for the transmission of infections like cholera. Snow discounted this theory that pollution or noxious form of ‘bad air’ caused cholera, and he rightly confirmed this through the map proving that the water pump in Broad Street, polluted by cesspit sewage where a baby's contaminated nappy was dumped, was the source of the outbreak in Soho, London.[2] The map used diagram, words and numbers simultaneously to present evidence and make an argument. It has influenced our understanding of disease mapping even today during coronavirus outbreak[3] as well as making an entrance into the academia.
Part I Theoretical examples and Ethical considerations
Theoretical examples
A study was undertaken in 2013 to understand the extent of the use of visual display in articles published within three reputable qualitative research journals. Despite data visualisation being considered an important step in qualitative analysis and is said to ‘add life’ to qualitative data, the results following a review of 784 articles published between 2007-2009 showed that only 27% used data display. Most commonly used displays were matrix (60% of all displays), and majority were found within ‘Results/Findings’ section followed by the ‘Method’ section. Though diagrams were used for diverse purposes, in general, visualisation was considered to be largely underutilised.[4]
Edward Tufte, one of the pioneers of data visualisation, noted that the use of ‘abstract non-representational pictures to show numbers’ has not been in much use for various reasons.[5] Until recently, the use of diagrams in earlier stages of the research process was not considered to be the most common approach.[6] A recent journal issue focusing on diagrams argues that though used as thinking tools, and with the ability to ‘transform abstract ideas into understandable and intelligible images’ that can be actioned, ‘the role of the humble diagram has been slow in earning recognition’.[7]
Charts and diagrams are thought to be easier for non-experts to understand in comparison to numbers. Walliman differentiates graphical displays into matrices, which are two dimensional arrangements of rows and columns, and networks that include maps and charts, which can be based on time-, role- or conceptual-orders.[8] An effective format of note-taking, ‘pattern-notes’ is strongly encouraged to assist in visually connecting information by making diagrams in the form of spider diagrams or mind maps.[9]
Grounded theorist, Charmaz strongly advocates pre-writing exercises such as clustering technique that makes use of diagrams to represent relationships as an essential step in theory building. Considered to be a non-linear approach to free the researcher from conventional linear thinking, clustering is thought to be ‘quick, active and changeable’. Fluent writers and those with writers’ block have found clustering beneficial in rethinking and reorganising ideas to redefine essentials, view emerging patterns, speed up writing, all with the intension of making the writing process less onerous.[10]
Though the use of visualisation is likely to have increased in recent years, there are still frequent discussions around when, and more precisely, at what stage should visualisation be used and is most effective. There is also a common misconception that visualisations are typically created for analysis of the data. My own experience has proved that at any stage of the research, visualisation has the ability to enhance the overall research process.
The motivation to use doodling as a methodology for my master’s dissertation[11], was greatly influenced by Ledearch’s research on the importance of visualising data, specifically to encourage ‘creative’ conflict transformation (CT).[12] Ledearch believes that ‘creative CT’ should be able to bring into focus effective ways to build healthy relationships and communities by fundamentally changing the way we think. By listening, interpreting and re-interpreting information in various formats, communities will likely be able to see some change and give new meanings to old problems.
To effectively achieve this, Lederach places enormous importance in translating talks i.e., data into figurative results. He advocates doodles primarily to channel the heard words from the brain’s ‘rational and conceptual view’ to the heart’s ‘emotional view’ and finally surfacing through the hand ‘connecting to form powerful, constructive and practical ideas’. Lederach believes that the data becomes enriched once it is shown in a visual format while simultaneously connecting spaces, relationships, processes and changes. Additionally, when people struggle to understand, describe and create through discussions alone, visualisations help to improve communication, as ‘people talk in images.’ [13]
Firstly, to appreciate the analogy of the creative CT process and the three conceptual views, and secondly, to understand how to effectively use this within my research, I illustrated Ledearch’s concepts and its impact on the people in the West Bank (Figure 1). Though basic in its form, this first diagram for my research significantly clarified the methodology, theoretical framing of the research questions (Figure 2) and understanding of the conflict topography (Figure 3) in the West Bank.
Visualising this theoretical data was key to holistically appreciate relationships, processes and changes in the West Bank, and enrich discussions during field work. It also stressed the need to constantly keep evolving as a researcher to keep up with the changing society and conflict in the West Bank. Throughout fieldwork, the visuals assisted in recognising the need for an additional outlet for articulating thoughts and feelings difficult to voice, highlighting crucial issues and potential strategies for effective and creative transformation.
CENDEP’s current PhD student, Fatima Hashmi, has been utilising visualisation including clustering to capture multiple perspectives around stigmatisation and resilience encountered by some of the minority groups in Pakistan and Colombia.[14] Clustering along with freewriting has helped her enhance her reflection skills, assist in early categorising and identifying of patterns, getting improved results for selective coding, developing theoretical insights, and organisation of findings.
Additionally, as part of her field research, participants were encouraged to create a ‘Stigmatisation tree’, which gave them a platform to discuss the different forms of stigmatisation they face, and motivated them to think about how resilient they are as individuals, family and community members. Developed by Dr Brigitte Piquard, this visual exercise has been a crucial methodology of CENDEP field visits in Oxford (2015), Colombia (2016) and Lebanon (2017), giving voice to the participants, and where students engage with the locals through a creative participatory method to visualise conflict, trauma, impacts and responses. This process has the ability to assist transformation, healing and empowerment of individuals and groups.
Ethical considerations
Once the decision to sketch out the written word has been taken, some important ethical considerations should be deliberated upon. As with conventional research, during visualisation process, it is important to be mindful of the ‘Do no harm’ principle, understanding and following cultural diversity and etiquettes, and treating sensitive information carefully. Other aspects where vigilance is necessary while illustrating data include avoiding creating images that patronising, or portray bias, stereotyping, marginalisation, discrimination, gender stereotyping, prejudices and intolerance.[15] It is crucial to note that if the data depicts vulnerable populations, necessary measures are in place to ensure protection of the sensitive populations, their identities and their information, and actions have been considered to mitigate or eliminate potential risks.[16]
It is also crucial to understand cultural meanings of colours, symbols and images being used so as to not to upset any individuals or groups represented within the data nor be disrespectful of any culture or religion. During the process of visualising a tool-box for field staff of local humanitarian organisations, we realised that the best way to approach this issue without hurting any cultural or religious sentiments was to have open discussions with the local populations and field staff to understand implications of the symbols and colours we were using within the diagrams. Based on the consultations, we could reach consensus that the final diagrams would, in no way, be disrespectful or threatening to all parties involved in the project including locals, people impacted by the data, NGO and the donor.[17]
Alberto Cairo, who wrote the book ‘How charts lie’ believes that graphicacy is a language which can present a lot of limitations. He believes that ‘every visualisation gives great power and with great power comes great responsibility’: therefore, being responsible of the content within the visualisation, and being responsible to ensure that the audience understands the visualisation without confusions and misunderstandings.[18] Coreell highlighted additional concerns such as stifling critical or contradictory voices, echoing biases and inequalities, and excluding representation of factors like the uncertainty of the data. This increases the likelihood of the visualisation resulting in ‘cruel and inhuman’ image and in some cases potentially overlooking human suffering. This misinterpretation is further likely to create a separation between the people impacted by the data and the people consuming the data. It is therefore essential to provide necessary context and explanations to ensure that these empowering tools do not give rise to bad, incorrect and flawed visuals with damaging and harmful results.[19]
Last but not the least, is the concern around ‘data’ that is being illustrated, its quality, legitimacy and limitations. The visualisations can only be as accurate as the data; misrepresented data will likely create biased diagrams. For an unbiased visualisation, it is necessary to understand and verify the data trail to avoid flawed visualisations.
Part II: Practical examples
In 2018, Everyday Action, a digital platform for donor management, surveyed over 460 non-profit sector individuals about their organisations’ ‘habits, culture, and outlook on the state of data’. The key takeaways were that in the current time of big data, NGOs were collecting large amounts of data, but few knew what to do with it, how to use it or process it. They additionally believed that the data was strategically important but it was not being used effectively.[20]
A lot has been written about how data visualisation has an impact on the business and marketing sectors and in the media, but there is limited research regarding how it affects research processes and the third sector. However, there is consensus that visual information allows people to interact with it, gain insight, draw conclusions and make better decisions. Hans Rosling, a global health expert was a champion of visualising data and statistics to better understand the world and fight against pre-conceived notions and ignorance. He strongly believed that decision makers needed to ‘see’ how the world is changing.[21]
Visualst, a data design firm along with a mobile-based research provider, GeoPoll designed a series of workshops creating visual stories and maps (Figure 5) post 2019 Cyclone Idai’s devastating destruction in Mozambique, Malawi, and Zimbabwe. The Cyclone Idai Data Stories Workshop Series firstly, wanted to facilitate reflective and meaningful dialogue between affected populations and responders to better respond to natural disasters and climate refugees, and secondly, to explore supply and demand of the relief efforts. Informed decision making around relief, return and resettlement issues were thought to be greatly influenced by these participatory and creative exercises that enabled people to interact with data in a ‘non-intimidating way’, and find greater value by not getting overwhelmed by vast daunting numerical data.[22]
Human rights defenders strongly advocate data visualisation to effectively convey messages and to engage and educate with easy to read and understand graphics. Results showed that visualisation is particularly useful in situations where there are abuses of power and where there is a need to enable immediate communication. Additionally, ‘tailoring’ visualisations for target audiences was seen as effective where there are polarised attitudes. A key challenge to visualisation was thought to be the difficulty of quantifying data in the field of human rights violations.[23] DME for Peace promotes visualisation in the field of peacebuilding as it gives an opportunity to effectively present hard evidence for peacebuilding policies and programs in conflict-affected communities and fragile states.[24]
Harvard Humanitarian Initiative’s PeacebuildingData.org features interactive visual analyses of data from countries affected by mass violence. It aims to give a voice to the people involved in peacebuilding and reconstruction processes and bridge the gap between peacebuilding as intended by decision makers and realities on the ground. Visualizing Peace, created using Political Settlements Research Programme’s (PSRP) Peace Agreements Database, is a timeline of the history of peace agreements across: time (from 1990 to 2015), geography, and different categories addressed within the peace agreements e.g., women and gender, human rights framework, transitional justice, etc. This timeline aims to inform and enhance PSRP’s understanding of research on peace processes to make them more inclusive and help bring an end to violent conflict. Based on its historical database from 1960 to 2012, UNHCR created a narrative visualisation to illustrate the migration of refugees, populations and countries that refugees have moved to and whether the host countries have granted them refugee status. The historical examples tend to improve understanding of issues around accountability and transparency.
In addition, visualisation helps in understanding information appropriately and in timely manner; identify emerging trends that can influence goals, and communicate human stories. With regards to its persuasion powers, one research found that 68% participants believed a scientific claim without visualisation, while 97% believed the same claim when visualisation was added.[25] This is also because seeing and visual perception in the visual cortex has been proved to be quick and efficient in comparison to the thinking cerebral cortex.
Visualisation has proved to be effective for quick analysis and decision-making to understand policy outcomes and preferences. It can help identify gaps, enhance the understanding of information, initiate meaningful communication with other actors, and promote dissemination of findings.[26] It provides an opportunity for people to reflect and debate upon their problems and struggles within the society, and in the process visualise constructive social change. With the rise in social activism and in its commitment towards social transformation, visualisation can become key in supporting the vision for transforming action to activism. Through its inherent quality of being accessible and universally understandable, graphics can empower activism as it moves ‘beyond participation, developing solidarity, challenging power relations,’ and ultimately ‘building emotional connections’ between people.[27]
Summary
The key motivations, facts and considerations are highlighted in the extracted version (Figure 7) of the 4-Ws diagram tool, that can act as a quick guide when commencing with the visualisation process. I would encourage all of us to pause and think at different stages or rather every stage of research, and question ourselves, “can a diagram help me here?” Who knows, we may find our eureka moment while exploring visualisation!
Previous DYOC Series
1# Introduction and History of Visual Research
Next
#3: ‘WHAT’ (4-Ws diagram)
#4: ‘WHO’ (4-Ws diagram)
#5: ‘WHY’ (4-Ws diagram)
#6: Words that end in diagrams
[1] Sack, H., 2020. William Playfair and the Beginnings of Infographics. [online] SciHi Blog. Available at: <http://scihi.org/william-playfair-and-the-beginnings-of-infographics/> [Accessed 21 February 2021].
[2] Rogers, S., 2013. John Snow's data journalism: the cholera map that changed the world. [online] the Guardian. Available at: <https://www.theguardian.com/news/datablog/2013/mar/15/john-snow-cholera-map> [Accessed 21 February 2021].
[3] FierceBiotech. 2020. FROM CHOLERA TO COVID-19: How integrated data visualizations inform decisions and impact actions. [online] Available at: <https://www.fiercebiotech.com/sponsored/from-cholera-to-covid-19-how-integrated-data-visualizations-inform-decisions-and-impact> [Accessed 22 February 2021].
[4] Verdinelli, S. and Scagnoli, N., 2013. Data Display in Qualitative Research. International Journal of Qualitative Methods, 12(1), pp.359-381.
[5] Tufte, E., 2001. The visual display of quantitative information. 2nd ed. Connecticut: Graphics Press LLC.
[6] Umoquit, M., Tso, P., Burchett, H. and Dobrow, M., 2011. A multidisciplinary systematic review of the use of diagrams as a means of collecting data from research subjects: application, benefits and recommendations. BMC Medical Research Methodology, 11.
[7] Engelmann, L., Humphrey, C. and Lynteris, C., 2019. Diagrams beyond Mere Tools. Social Analysis: The International Journal of Anthropology, 63(4), pp.1-19.
[8] Walliman, N., 2006. Social Research Methods. SAGE Publications.
[9] Ridley, D., 2012. The Literature Review. SAGE Publications. 2nd ed.
[10] Charmaz, K., 2006. Constructing grounded theory. SAGE Publications.
[11] Maladkar, A., 2017. Understanding the Role of Murals in Conflict, Occupation and Forced Displacement: Arterial Walls of West Bank. Master of Arts. Oxford Brookes University.
[12] Lederach, J. (2003). The little book of conflict transformation. Intercourse, PA: Good Books.
[13] Lederach, J. (2005). The moral imagination. Oxford: Oxford University Press.
[14] Hashmi, F., n.d. Understanding forms and processes of stigmatisation and populations’ responses: A multi-site case study of minority groups in Pakistan and Colombia. [ongoing] Ph.D. Oxford Brookes University.
[15] Walliman, N., 2006. Social Research Methods. SAGE Publications.
[16] Newtactics.org. 2017. The Use of Data Visualization in Human Rights Advocacy | New Tactics in Human Rights. [online] Available at: <https://www.newtactics.org/conversation/use-data-visualization-human-rights-advocacy> [Accessed 23 February 2021].
[17] Developing ‘Toolbox for Data Management and Reporting in CCCM’ that could be easily carried on field, for a local NGO field staff involved in humanitarian actions, Research lead Dr Brigitte Piquard, 2017
[18] Cairo, A., 2019. How Charts Lie.
[19] Correll, M., 2018. Ethical Dimensions of Visualization Research. [online] Seattle, WA: Tableau Research. Available at: <https://arxiv.org/pdf/1811.07271.pdf> [Accessed 24 February 2021].
[20] EveryAction. 2018. 5 Things You Need to Know about Nonprofits + Big Data. [online] Available at: <https://www.everyaction.com/blog/5-things-you-need-know-about-nonprofits-big-data> [Accessed 21 February 2021].
[21] Rosling, H., 2006. The best stats you’ve ever seen.
[22] McCrocklin, S., 2019. Data Visualization for Humanitarian Crisis Relief - GeoPoll. [online] GeoPoll. Available at: <https://www.geopoll.com/blog/data-visualization-for-humanitarian-crisis-relief/> [Accessed 24 February 2021].
[23] Newtactics.org. 2017. The Use of Data Visualization in Human Rights Advocacy | New Tactics in Human Rights. [online] Available at: <https://www.newtactics.org/conversation/use-data-visualization-human-rights-advocacy> [Accessed 23 February 2021].
[24] Duncan, E., 2014. Getting Data Visualization Right. [online] DME for Peace | Design, Monitoring and Evaluation for Peacebuilding. Available at: <https://www.dmeforpeace.org/resource/getting-data-visualization-right/> [Accessed 23 February 2021].
[25] Chun, R., 2015. 6 lessons academic research tells us about making data visualizations - Poynter. [online] Poynter. Available at: <https://www.poynter.org/reporting-editing/2015/6-lessons-academic-research-tells-us-about-making-data-visualizations/> [Accessed 25 February 2021].
[26] Gatto, M., 2015. Making Research Useful: Current Challenges and Good Practices in Data Visualisation. [online] Reuters Institute for the Study of Journalism, University of Oxford. Available at: <https://www.alliance4usefulevidence.org/assets/Making-Research-Useful-Current-Challenges-and-Good-Practices-in-Data-Visualisation.pdf> [Accessed 23 February 2021].
[27] Chatterton, P., Fuller, D. and Routledge, P., 2007. Relating action to activism: Theoretical and methodological reflections. In: S. Kindon, R. Pain and M. Kesby, ed., Participatory Action Research Approaches and Methods, 1st ed. Routledge.
[28] Koblin, A., 2011. Visualising ourselves…with crowd-sourced data.
[29] Jordan, C., 2008. Turning powerful stats into art.
