That which we call a user, by any other word would smell as sour

(With apologies to Shakespeare, W. Romeo and Juliet. 1595. London)

Introduction

I am intrigued as to whether the classification of the world’s 7 billion individuals as “users” has hindered the success of both HCI (Human Computer Interaction) and the Public Health aspects of Digital Civics. Not that the actual word “users” concerns me, I will happily continue to use it within this blog post, it is the reductive expression of the individual that I believe is the issue.

In the role of designers, we build systems, artefacts, services and infrastructures for people and publics. All of these socio-material assemblies that can be labelled as “Things” [1], are hidden behind or expressed by a service interface that we, as designers, design. A point of commonality and an issue of importance with these interfaces is the frustration, confusion, anger and anxiety [2] they can inflict on their users, which can be caused through miscommunication, inability to convey meaning and poor signposting.

This blogpost will begin by setting out the contexts and key definitions, then move onto examine the history and issues, first for users within HCI and then for users within Public Health. I will close with potential areas for research, for Public Health to reflect on its first successes and for HCI to reflect on a means of addressing the problems caused by aggregating entire populations.

 

Contexts

Both HCI and the Public Health aspects of Digital Civics are wide ranging topics, so I want to focus where long term relationships are required and the power flow is from a large organisation to an individual. For Public Health, I will focus on the scenario of people with chronic, disabling conditions who depend on long term home care within the UK. For HCI, I will focus on the type of interfaces we use every day on our computers and phones in the UK, examples being word processing software, phone apps and systems for our work.

 

Who is a Designer?

An aim of designers is to plan how we, as users, will interact with the Things they build, what opportunities they will give us, what constraints they plan for us to overcome. An issue being that when they describe their step-by-step plans for us, we do not blindly follow as the designers intend, but use these plans in a discursive fashion, a reference amongst others that we use to achieve our goals [3].

In addition, we may never see the designers in these contexts. They are likely hidden from us, distant geographically, temporally, socially and culturally. So when the designers think of the user, they do this within their own set of beliefs. Lucy Suchman frames this scenario as a Situated Action [3], which I discuss in an earlier blogpost, examining how our face to face conversations with all our gestures, facial expressions, ums and ahs, all form a highly complex interactive set of competencies. With misunderstandings still occurring when we are face to face, then the communication issues we experience are perhaps not a surprise when the user and the designer are so separated in time, distance and culture.

 

Who is a User

The “user” here is the consumer of the Thing the designer designs. There may also be other people involved, these could be home carers or an IT Support team but they would still live within the interface and the plan laid out by the designers.

Within Public Health these can be known as ‘patients’, ‘service users’, ‘disabled people’, ‘the client’. HCI knows them as ‘end users’ or ‘users’.

 

Let’s look at HCI, Past and Present

The user has played a key role in the history of HCI research [4] and also a personal history for me within commercial HCI, as my first work experience involved coding and designing on mainframes. Bannon’s “From Human Factors to Human Actors” [5] echoes my experience, as in the era of computing to the 1980s, we paid minimal attention to the users. They were seen as an operator of hardware and software, rarely visited and often disregarded simply as ‘wetware’. However, the advantage was that the users were a known quantity, in that it would only be the company’s staff that would have access, they would be trained on the system and have close access to colleagues for help and support. Of far greater priority to the designers were the technology constraints at the time: we had limited display capabilities such as 32 rows by 80 columns (see Figure 1), no graphics, no mouse, simplistic file structures, minimal storage and minimal processing power. The designers ordered and the users complied, adapted and changed their natural behaviours to that of the interface.

Figure 1: Visual Display Units from the 1980’s

The next era for the user was a mouse with a graphics capable screen as the Personal Computer had arrived. Storage and power were still limited, so too were design principles (see Figure 2). as we had only just embarked on the HCI path [6]. The new issue was that with personal computing, we no longer knew who all the users were, their capabilities, their location, their skillsets. We still designed an interface as how we thought best, with users still regarded as a troublesome part of the development lifecycle.

Figure 2: Early version of Microsoft Windows

In the present day we have handheld devices, seemingly unlimited power, storage and communications. We have trained designer teams who hold user workshops, perform A-B testing with users to compare web pages, run usability labs and work with proxies such as user personas [7]. But now we have a potential world of 7 billion users, which means we cannot talk to any significant percentage of them. The result is that designers still design to what they believe is correct for the users. This design is highly skilled, knowledge based, but living within the culture of the designer as a Situated Action. To further complicate this, the purpose of the interface simply to interact with a system has been transformed, with users now talking to users, that is, a social organisation rather than a computational one [4].

 

An Issue Within HCI

I feel we have an issue with HCI performing to its purpose of empowering the user with the technology. Examples such as poor communication within HCI (see Figure 3) and even difficulties performing such common tasks as setting digital clocks [8] are widespread, with users feeling frustration, creating inefficiencies [2] and an inability to use the technology to its fullest capabilities.

Figure 3: Error message from Tyne & Wear Metro app

To solve these issues, HCI has looked at User Centred Design [9], Tangible User Interfaces (check my blogpost here), Humanistic HCI, Feminist HCI, Experience Centred Design, post-colonial HCI [10] and other variants, incorporating a multi-disciplinary approach ranging from the social sciences to the arts to philosophy [11]. However, we still seem to be struggling with our interfaces as we design for this mythical user, who has been distilled from the world’s population to a single node, who has been abstracted from their unique character and interaction with the world [6]. This distillation can be somewhat explained through feminist epistemology, in that long-established patriarchal patterns have been and are being used to conflate attributes as outlined in my blogpost here.

This is intriguing, as the designers are not wholly apart from the world of users, they themselves are users of a wealth of computer interfaces. So why don’t they convey the lived experience they inhabit? Perhaps one answer lies in a skills assumption. I would expect designers to be highly IT literate by the nature of their role. The OECD [12] researched across 33 countries and found that 26% of the adult population could not use a computer, with just 5% operating at a high level of IT literacy. If, as a designer, we are not designing for this range of skill sets, then we are likely to fail. Again, Lucy Suchman’s Situated Actions [3] come into play as designers design for how they picture the user.

This is a problem and borrowing from the field of social policy, it is a wicked problem. Wicked problems are difficult to describe, perhaps do not have an addressable right or wrong and are highly interactive between their influencing factors [13]. The designers’ reduction of the world to a single user is a wicked problem that HCI is there to address [14].

 

Let’s look at Public Health, Past and Present

Within the context of long-term care, the history of user-led models within Public Health is a relatively recent one. Muir and Parker’s “Many to Many” [15] discusses that up to the 1970’s, the designers were located at government level, where the organisational model in place for public health was one of bureaucracy. In this model, the expert elite make the policy decisions as how best to enact the orders of the elected leaders. These decisions travelled from the centre of government down to local authorities who were trusted to deliver to the users. This bureaucratic model works with simple problems that have standalone input and outputs, with outcomes easily planned and measured. Wicked problems exhibit cross organisational silos, demand individualisation and cannot be easily understood and fixed. Wicked problems describe the Public Health users in our context.

Historically, the UK handled this wicked problem by either the family providing all care or purposing institutions such as asylums (see Figure 4) to remove the users from society, with little regard for the users’ personal needs [16].

Figure 4: 18th Century Lunatic Asylum – Wellcome Images (Creative Commons Attribution)

As we moved on from the 1970s, we turned our back on these institutions [17] and the designers moved their users into the community to be cared for by the State and supported by their families. This was part of a market led approach, where a neoliberalist drive implemented targets with users now classified as consumers [15]. The result, as Wendy Brown states, was “market values are crowding out all others” [18]. The State still dictates the design through Situated Actions, keeping to a mechanistic [17] formula that disregarded the individual, their needs and wishes, e.g. here are some carers, here is some equipment.

 

An Issue Within Public Health

As for HCI, an outcome can be highly stressed users fighting against distant designers that are not catering for them as an individual. They have to “battle the system” [15], where they or their families have to undertake their own research to understand and navigate the system, resulting in inefficiencies with their time and that of the authorities, as the user tries to locate the person with the appropriate knowledge and power. Users, even though submerged in a neoliberalist system, cannot materially pick and choose between solutions, they can perhaps change the agency that provides their carers but the designers’ plans remain the same, meaning the level of service remains the same.

How the designers solve the myriad of issues that belong to a myriad of users is again, a wicked problem.

 

Commonalities. From the Transactional to the Relational

I see we have a common issue with the user experiencing frustration and inefficiencies with the service that has been designed for them, i.e. the service is delivered to the user, rather than with the user. For Public Health, the previous models of bureaucracy and markets have resulted in a transactional relationship from the State to the user. The user calls upon the State for a service, the State assesses the type and degree of service that they feel is best suited for the user, and then the State applies that service to the user. For HCI, the anonymisation & distillation of the individual to a designer held Situated Action, has also resulted in a transactional service. The user calls upon an interface for a service and then the interface, as a proxy for the designer, returns what the interface believes the user has requested.

As a method to ease these issues, both HCI and Public Health are looking to change from the transactional to the relational. This will not result in a common method and the relational model will not apply in the same way to all users, but it is a way forward for these specific scenarios. The solution being that the users are brought into the problem solving alongside the designers.

For Public Health, NHS England have embarked on a coproduction programme [19]. Its aim is to recognise the user as the key resource when a service is required, enabling that user to design their own health outcomes, build strong relationships with a multidisciplinary team, and support others in their community. This aligns to Muir and Parker’s “Many to Many” report that calls for these deep relationships, changing the State from a managing function to an enabling function [15]. Digital Civics has also built a corpus of work targeting coproduction with small publics, with the Malmo Living Labs in Stockholm being a leading example demonstrating that coproduction can deliver benefits.

For HCI, there are the methods of User Centred Design and Participatory Design that incorporate users into the standard software development lifecycle. These methods gain knowledge of the users through use of multidisciplinary teams, brainstorming workshops and ethnographic principles [20] to help us better understand the users. But we are not at the level of the individual, we are still adopting a process of reductive representation [10] to aggregate the world’s population down to that of a user. Involving users in the design process is a step forward to unpackaging the designers’ Situated Actions but still falls short of addressing the needs of every individual.

 

The Way Forward for Research

Within the NHS England coproduction programme, the State uses frontline staff to take on the role of an expert designer to design and innovate with their users, to nurture rich and long lasting relationships [15]. This participatory model has started to roll out in England but with an issue. Coproduction is a non-trivial change for frontline Public Health workers, who “need to change their attitudes, priorities and training” [21] and is seen as deeply suspicious by some [22] [23], resulting in a lack of frontline support as “uncertainty [persists] about why and how to do involvement well and evaluate its impact” [24].

This is a variance from that discussed by Manzini, who lays out the role of the expert designer as the traditional designer transformed into a social actor, supporting the design process [25]. Manzini may perhaps envisage that if there are no design experts on the frontline, then the user becomes the primary diffuse (or non-expert) designer, making use of their natural design abilities supported by their in-depth knowledge of the service required. The user would then be supported by further diffuse designers from the State who will exist within their different domains, each bringing their siloed knowledge.

When we map this scenario onto Manzini’s Design Node Map for the initial deployment of coproduction, I see users and the State actors residing within different quadrants (see figure 5). Research could take place within this node map (and Manzini’s Participation Involvement and Interaction Quality maps), confirming the placement of actors, examining the issues where the expert designers are absent, and the movement of actors over time as they gain competency and transition to sense making.

Figure 5: Personal view of Manzini’s Design Node Map: an initial deployment illustrating the actual and desired

Within HCI, the concept of adapting the interface to the user is far from new, with many papers examining this topic in detail such as Barrera et al [26]. We cannot follow a face to face participatory design model for such worldwide systems, as the designer cannot work alongside every individual (see figure 6).

Figure 6: Current interaction from the designer to the user

However, this does not obviate the need, as users are heterogeneous in time and over time, so perhaps we could look to a virtual participatory design. With machine learning and artificial intelligence acting as the expert designer, research could take place to understand the issues and possibilities. This virtual designer could deliver a relational model by learning the capabilities of the user, their needs and the users’ preferred way of working (see figure 7). The complexity of this cannot be denied, especially as the user will continually evolve as they became accustomed to the software, and, change as their life experiences accumulate [6].

Figure 7: Potential relational interaction between the virtual designer and the user

 

Conclusion

Using the scenarios defined at the start of this blogpost, I have discussed the histories that have led to designers within HCI and Public Health creating interfaces that have caused issues on their users. These designers have delivered in a transactional manner, aggregating their users to a node that the designers have pictured in their own minds through Situated Actions. A more relational, participatory approach is now being adopted that has opened up questions for research.

For HCI, we could examine how employing a virtual designer could enact participatory design with the world’s 7 billion users. The research would first set out to examine the issues and demands of such an approach to establish a way forward.

For Public Health, where participatory design is at an early stage with ambitious plans to expand [27], research could examine the roles of designers currently in place, determining the gaps and the issues arising from such gaps. This research could be used to establish the path towards deploying expert designers in the places they are most needed.

Although I have selected negative aspects to discuss, it must be recognised that positive outcomes do exist, where the distance between the user and the designer has been narrowed, to fade the interface away [28].

 

References

  1. Erling Björgvinsson, Pelle Ehn, and Per-anders Hillgren. 2010. Participatory design and democratizing innovation. PCD’ 10 Sydney, Australia. https://doi.org/10.1145/1900441.1900448
  2. Jonathan Klein and Y Moon. 2002. This computer responds to user frustration: Theory, design and results. Interacting with Computers 14, May: 242–243. https://doi.org/10.1145/632716.632866
  3. Lucy A. Suchman. 2007. Human-machine reconfigurations: Plans and situated actions. Cambridge University Press, New York.
  4. Christine Satchell and Paul Dourish. 2009. Beyond the user: Use And Non-Use in HCI. Proceedings of the 21st Annual Conference of the Australian Computer-Human Interaction Special Interest Group on Design: 9. https://doi.org/10.1145/1738826.1738829
  5. Liam J. Bannon. 1991. From Human Factors to Human Actors. The Role of Psychology and Human-Computer Interaction Studies in Systems Design. In Design at work: Cooperative Design of Computer Systems., J Greenbaum and M Kyng (eds.). Lawrence Erlbaum Associates, Hillsdale, 25–44. https://doi.org/10.1016/B978-0-08-051574-8.50024-8
  6. John McCarthy and Peter C. Wright. 2006. Dialogical Approach to Experience: Uncovering Critical Potential. Proceedings of The Virtual Conference 2006: Designing Digital Experience.
  7. Peter Wright. 2011. Reconsidering the H , the C , and the I. Interactions 18, 5: 28–31. https://doi.org/10.1145/2008176.2008185
  8. Susanne Bødker. 2006. When second wave HCI meets third wave challenges. Proceedings of the 4th Nordic conference on Human-computer interaction changing roles – NordiCHI ’06, October: 1–8. https://doi.org/10.1145/1182475.1182476
  9. Chadia Abras, Diane Maloney-krichmar, and Jenny Preece. 2004. User Centered Design. In Encyclopedia of Human-Computer Interaction, W Bainbridge (ed.). Sage Publications, Thousand Oaks.
  10. Donald A Norman and SW Draper. 1986. User-Centred System Design: New Perspectives on Human-Computer Interaction. Lawrence Earlbaum Associates, New Jersey.
  11. Paul Dourish and Scott D Mainwaring. 2012. Ubicomp’s Colonial Impulse. UbiComp ’12: 133–142.
  12. Jeffrey Bardzell and Shaowen Bardzell. 2016. Humanistic HCI. Interactions, 20–29. https://doi.org/https://doi.org/10.1145/2888576
  13. OECD. 2016. Skills Matter: Further Results from the Survey of Adult Skills. Paris.
  14. HW Rittel and MM Webber. 1973. Wicked problems. Man-made Futures. Policy Sciences 4, 2: 155–169. https://doi.org/https://doi.org/10.1007/BF01405730
  15. Antti Oulasvirta and Kasper Hornbæk. 2016. HCI Research as Problem-Solving. Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems – CHI ’16: 4956–4967. https://doi.org/10.1145/2858036.2858283
  16. Rick Muir and Imogen Parker. 2014. Many To Many: How the relational state will transform public services. IPPR. London.
  17. Jenny Morris. 1993. Independent Lives? Community Care and Disabled People. The Macmillan Press Ltd, London. https://doi.org/10.1007/978-1-349-23136-2
  18. Graeme Cooke and Rick Muir. 2012. The Relational State. IPPR. London.
  19. Wendy Brown. 2015. Undoing the Demos. Neoliberalism’s Stealth Revolution. Zone Books, New York.
  20. Maren Batalden, Paul Batalden, Peter Margolis, Michael Seid, Gail Armstrong, Lisa Opipari-Arrigan, and Hans Hartung. 2016. Coproduction of healthcare service. BMJ Quality and Safety 25, 7: 509–517. https://doi.org/10.1136/bmjqs-2015-004315
  21. Atau Tanaka, Lalya Gaye, and Ranald Richardson. 2010. Co-production and Co-creation: Creative Practice in Social Inclusion. Cultural Computing: 169. https://doi.org/10.1007/978-3-642-15214-6_17
  22. David Boyle and Michael Harris. 2009. The Challenge of Co-Production. How equal partnerships between professional and the public are crucial to improving public services. Public Administration: 1–28. https://doi.org/10.1111/etap.12092
  23. Rachel Perkins and Mike Slade. 2012. Recovery in England: Transforming statutory services? International Review of Psychiatry 24, 1: 29–39. https://doi.org/10.3109/09540261.2011.645025
  24. Caroline Glendinning. 2008. Increasing Choice and Control for Older and Disabled People: A Critical Review of New Developments in England. Social Policy & Administration 42, 5: 455. Retrieved from http://onlinelibrary.wiley.com/doi/10.1111/j.1467-9515.2008.00617.x/abstract%5C
  25. Josephine Ocloo and Rachel Matthews. 2016. From tokenism to empowerment: Progressing patient and public involvement in healthcare improvement. BMJ Quality and Safety 25, 8: 626–632. https://doi.org/10.1136/bmjqs-2015-004839
  26. Manzini. 2015. Design, when everybody designs. MIT Press, Cambridge, Massachusetts.
  27. Luisa Fernanda Barrera, Jaime A Pavlich-Mariscal, Angela Carrillo-Ramos, and Leonardo Flórez-Valencia. 2002. Runa-Kamachiy: Conceptual Integration Model Between HCI and Adaptation Oriented to User Interface Usability. Dyna 137.
  28. 2014. NHS: Five year forward view. Retrieved from https://www.england.nhs.uk/wp-content/uploads/2014/10/5yfv-web.pdf
  29. Mark Weiser. 1991. The Computer for the 21st Century. Scientific American 265, 94–104. Retrieved from http://www.nature.com/doifinder/10.1038/scientificamerican0991-94

 


Author: Peter Glick

Leave a Reply

Your email address will not be published. Required fields are marked *