Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/unpaywall/usability-is-the-best-policy-public-policy-and-the-lived-experience-of-kyf2b2Z2dg
References
References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.
- Publisher
- Unpaywall
- ISSN
- 1477-9358
- DOI
- 10.14236/ewic/hci2007.4
- Publisher site
- See Article on Publisher Site
Abstract
Usability is the Best Policy: Public Policy and the Lived Experience of Transport Systems in London Philip Inglesant M. Angela Sasse Department of Computer Science Department of Computer Science University College London University College London Gower Street, London WC1E 6BT Gower Street, London WC1E 6BT +44 (0) 20 7679 3039 +44 (0) 20 7679 7212 [email protected] [email protected] and computer power incorporated into smaller and cheaper ABSTRACT personal devices, are coming together, so that computers are This paper explores the relations between public policy and more and more integrated into daily experience [45]. So, usability in lived experience, drawing on 3 case studies in one increasingly, e-Government meets the public not only in important area of urban policy, transport. For these studies, increased efficiency and better information for existing discourse from interviews and focus groups with a total of 120 government services, but in new services provided in new participants, and a written corpus of over 80 documents, were ways. collected and analyzed, together with interviews with 25 key staff and observations of user interactions both in the laboratory With these extensions of computer power comes a widening of and in situ. The resulting rich dataset presents a new the boundaries for Human-computer interaction research. perspective on e-government systems in use. The results show Computers are now everywhere, often unseen. People that usability must be prioritised at the policy design stage; it encounter them in many different ways in different contexts of cannot be left to implementation. Failure to do so is use, not only in the workplace; interactions with computers are experienced by users in systems which fail to work together to no longer restricted to the keyboard, mouse, and monitor. meet their needs. Negative experiences, in turn, may lead to In software development, the role of design in creating a loss of trust and legitimacy, and detract from public value and positive user experience has been recognised at least since community well-being. These findings, therefore, provide Kapor [27], in the early 1990s, put forward the case for lessons from HCI insights for both public policy-makers and software design distinct from user interface design and from implementers of e-government systems. The paper concludes software engineering, even if this has not always been honoured by suggesting some HCI methods for pre-venting usability in practice. Yet in public policy, design starts long before the problems in e-government systems, by involving users in implementation of systems, with decisions made to balance the design in order to understand their lived experiences around the interests of divergent stakeholders [3]. ecology of the systems. 1.1 Why this Matters in e-Government Categories and Subject Descriptors In the private sector, services must be appealing to customers as H.5.2 [Information Interfaces and Presentation (e.g., HCI)]: well as free of basic usability problems to succeed. But quality User Interfaces - user-centered design; C.3 [Special-Purpose services have additional urgency in e-government, for two and Application-Based Systems]: Smartcards. reasons. Firstly, good public services are important to strengthen political legitimacy as sources of public value [28]. General Terms Secondly, government provides or enables public services in Design, Human Factors, Theory. pursuit of desired policy outcomes. These outcomes could include, for example, the encouragement of behaviour change (e.g., to use public transport, rather than the car). But beyond Keywords these immediate outcomes are wider goals such as social Ubiquitous computing, embodied interaction, public policy, inclusion, environmental sustainability, and community well- e-government, lived experience, usability, design. being; this has implications for the evaluation of systems in support of public policy [20]. 1. INTRODUCTION Poor usability might be expected to reduce take-up of The use of computers by government is not new. However, e-government initiatives. Take-up is important if the larger ‘computers’ are changing: the Internet, mobile communications, public policy aim is to free resources, perhaps towards the provision of better services in other ways. Margetts and Dunleavy [31] examine demand-side barriers to e-government. In usability terms, innovations must be capable of domestication; they must fit into everyday, personal routines. © Philip Inglesant, M. Angela Sasse, 2007 Published by the British Computer Society The transaction costs, the effort and time required in learning People and Computers XXI – HCI… but not as we know it: new ways of doing things, are a strong initial barrier to Proceedings of HCI 2007 adoption; it has been repeatedly shown that even small up-front Linden J. Ball, M. Angela Sasse, Corina Sas, Thomas C. Ormerod, Alan costs deter people from making a change, even if this would Dix, Peter Bagnall, and Tom McEwan (Editors) lead to a saving in the long run. It is well-established, however, that usability is not the only the other way round; this implies the need for policy-making determinant of take-up [10]. There must be clear benefits from which is citizen-centred, in the sense that the citizens situated using electronic rather than traditional channels [31]; and this is needs are placed foremost and evaluated at all stages of policy in the context in which a more consumerist public has high design. expectations of quality, influenced by their experiences of private-sector services [28]. 2.1 Usability Beyond the Display Terminal The most widely accepted definition of usability is the ISO Government, like the private sector, could give financial 9241 Part 11 standard [25], with its three dimensions of incentives (or impose financial dis-incentives) to encourage effectiveness, efficiency, and satisfaction. A strength of this behaviour change by citizens, or simply to increase take-up, but standard is its emphasis on context and on the type of user and government, unlike commerce, can also draw on its authority [23] to use more coercive measures. However, deeper the type of task. ISO 9241-11 is clear that usability depends on the context of use, which includes the users in their specific behaviour change can only be on the basis of intrinsic circumstances. motivation, based on factors such as interest and enjoyment [9]. However, reflecting its age, the standard is limited in its scope: Usability is therefore an important, yet often under-considered, usability is presented as a subset of ergonomics for office work aspect of e-government. Citizens’ satisfaction with public with display terminals, although the document actually services is an essential contribution to public value in its own discusses usability beyond this narrow specification. right; but as well as this, in achieving policy outcomes, trust, legitimacy, and co-operation are preferable to an antagonistic The definition of usability as detailed in ISO 9241-11 is too relation between citizen and government. restrictive. This is partly in its narrow focus and outdated hardware context; the ways in which computers are encountered Unfortunately, as the findings in this paper illustrate, usability today include emergent interactive devices such RFID, mobile, is currently often overlooked in public policy design – there is a misconception of usability as ‘something the system and specialized interfaces, as well as Web interfaces. Non- electronic interactions, as well as interactions between people, implementer will take care of’. But the origins of usability work alongside computers in experiences of computer use [22]. problems may lie with public policy decisions which cannot be overcome at the implementation stage. The ISO 9241-11 definition of usability is too restrictive, more fundamentally in its too-narrow conception of usability in [11]. 1.2 Meeting This Need in HCI Research ISO 9241-11 is concerned with the usability of a piece of The findings of this paper are drawn from 3 case studies of equipment or software for a defined set of goals in office work. e-government in a large urban area (London) in an application But goals are wider than this; it is not simply a question of area (transport) which is widely considered to be one of the being able to use, but of emotional reasons for wanting to use a most pressing concerns of urban government. Transport touches system [11]. In the messy world outside the workplace, the the daily lives of most Londoners and visitors. At the same goals of a user, their intentionality [12], will be framed in larger time, use of transport is outside traditional locations of HCI terms - getting to work on time, for example - which are not concern, being inherently mobile and outside the home or necessarily the same as the goals of the service provider. workplace, and so extends the HCI understanding of usability into the lived experiences of service users. 2.2 Encounters in Passing For service providers, systems and their related interactions are These issues are addressed in this paper as follows. Section 2 central to their work. For the service user, on the other hand, builds on existing research in interactions between people and these interactions are peripheral. As computers are increasingly computers to develop such an understanding. Based on this pervasive, accepted as ordinary tools for everyday life [4], so framework, usability issues in each of the case studies are the focus of users is on the objects of their activities rather than identified, in Section 3; common themes which give coherence on the computer artefacts themselves. From the point of view of to these issues are joined together in Section 4. service users, the interaction is merely a part of their daily This is the basis on which, in Section 5, the implications of routine: “In fact, users don’t think of themselves as primarily these issues for public policy are analysed. It is argued that the having anything to do with the computer at all” [29]. insights of HCI research have lessons for policy-makers, not These interactions are often semi-transparent [5]; the user only in the implementation, but in the making of public policy presses the button or types on a keyboard without consideration at all stages. Beyond the interface, this implies the need to for the action itself. Only when there is some problem does the understand the real-world, situated actions of service users. The equipment in itself become the focus of attention, an paper ends with some suggestions from the HCI tradition for interruption in the flow of action, which Winograd and Flores carrying this out in practice. [46] call a breakdown. This is action engaged with the world; its basis is tacit, pre-reflective knowledge; action is always 2. BACKGROUND carried out in a particular place and context. In systems design, there is no longer the lack of consideration for usability; the former secret shame [27] is now seen as a 2.3 Meeting Users in the World crucial component, even by the engineering mainstream. Yet a The need for e-government to be capable of domestication [31] usability engineering approach can fail to see the complexity of gives a new meaning to Nielsen’s well-known heuristic: “match the social and situational context of interaction [11]. between the system and the real world” [36]. Beyond using Intrinsic motivation and a relation of public trust and words and conventions which are familiar to users, the co-operation in government suggest the need for e-government interactions, no longer limited to what happens on a screen, systems to emphasise softer, hard-to-measure aspects of must conform to users’ expectations and fit with their daily usability, beyond the absence of problems, to pleasure and even routines. enjoyment. E-Government should meet the needs of users, not The analysis of affordances, popularised by Norman [37], has not only on discourse but on observation and analyses of the proven a powerful tool for HCI research and design. In meeting affordances [15, 34, 37] of the systems, across cases which the needs of users, perceived affordances provide not only contrast in their uses of technology but are united by a single clues, but possibilities for action. As originally conceived by application area. Gibson [15], affordances are essentially in the bodily relation The cases studied here are all examples of electronically- between an actor and the environment; an animal perceives that enabled services which go beyond putting existing services a particular surface affords walking on, for example. online, and indeed go beyond online in the sense of interaction Software and computer interfaces also afford action [34]; the through the Internet alone. All three of the implementations perception of affordances at an interface may be made available make use of Web interfaces; but in each of them, this is just one by a visual symbol whose interpretation may depend on some part of an ecology of interfaces and different media types [35, prior learning, but this does not deny the real possibilities for 39]. action in these virtual spaces. The user perceives that by The studies consisted of interviews and focus groups with users performing a certain action, touching a smartcard against a card of e-government services (120 participants in total), interviews reader for example, a physical effect, such as opening a ticket with 25 staff in the government bodies and companies gate, is achieved. Thus, perceiving the possibilities for action is implementing the systems (at all levels including strategy and important if a system is to be useful as well as usable. senior management), and observations of systems in use. The corpus also included over 80 documents on transport policy and 2.4 Interactions in Lived Experience its implementation. The case-specific data are detailed in the The most frequent encounters with ICTs have - until recently - case study reports that follow; the total given here includes a been in workplaces. Once the computer leaves the desk, number of staff interviews and documents relating to overall simplifications which rely on tacit assumptions about the users transport and e-government issues. are no longer applicable: beyond the workplace, the user is not The average duration of interviews was approximately 40 necessarily well-trained and may not have skills from long-term minutes. Interviews and focus groups were voice-recorded, use. The user cannot be compelled to use the system in certain transcribed, and analysed using discourse analysis [38] and ways. Users outside the workplaces are heterogeneous, and Grounded Theory [41]. This combination of analytical methods encounter systems in situations which are varied, contingent, was suggested by the need to not only identify key causes and and unpredictable. consequences in an inductive way, but to go beyond the There is increasing awareness that human-computer interaction findings of Grounded Theory to produce a rich, yet rigorous, needs to be seen as part of the complex interplay between social reflexive analysis from different perspectives [1]. actors and technology in many contexts and through various interfaces. Meeting the real world, computers are no longer 3.1 Background to the Case Studies only concerned with processing ‘facts’, but are used in diverse The findings are based on three case studies of systems in a contexts which may be very different from those envisaged by single area of public policy, urban transport, implemented by a the designers, and raise higher-order political and emotional single authority, Transport for London (TfL). TfL, established issues [40]. in 2000 as part of the newly established Greater London These are the human needs of the lived experience [33] with Authority (GLA), has a remit covering all transport modes technology. The lived experience goes beyond user experience (currently with the significant exception of ‘overground’ Train with a computer to include many other technological and Operating Company – TOC - services). The GLA, which human objects, as well as the social user's own experience, includes a powerful mayor and an assembly, has overall capabilities, and values. responsibility for transport policy in London. Along with this devolution to a new, unified authority has come an integrated Lived experience emphasises the ways in which difficulties at transport strategy, in which the implementations studied here, the interface can lead to serious disruptions away from the while diverse in their applications, support coherent overall interface in the lives of users. Thimbleby et al. [44] give an policy objectives. analysis of a ticket machine, in which poor interface design can have especially serious consequences. The extremely high It is perhaps not immediately obvious that transport is a major volumes of use lead to important cost implications, as well as area of public policy, and that information systems in support of frustration and inconvenience for large numbers of passengers. transport policy can be categorized as e-government systems. But transport as an issue area touches on central public policy questions: the problems of urban traffic congestion and urban 3. UNDERSTANDING USERS’ mass transport are high priorities for city government. EXPERIENCES OF E-GOVERNMENT: THREE CASE STUDIES 3.2 The TfL Oystercard The increasing integration of computer power into mundane The TfL Oystercard is a transport smartcard for use in London. artefacts gives new significance to the question raised by The card itself is an almost blank, blue, contactless Radio- Bannon and Bødker: how is it possible to understand an artefact Frequency Identification (RFID) card which stores period which “reveals it self to u s fully o n ly in u se?” [4]. How is it tickets as well as Pay-as-you-go value which can be used to pay possible to grasp usability in the lived experience of socially 1 for individual fares . Oyster is one of a number of similar situated users? implementations worldwide. From a research tradition going back at least to Boland [6], this is a hermeneutic process, iteratively interpreting qualitative data Pay-as-you-go usually gives a discount over the equivalent from a close engagement with users. The aim is to understand cash fare. A daily ‘cap’ applies to such fares. This cap is not only what people do at the interface, but how this fits within guaranteed to be no more than the cost of the equivalent Day their wider needs. This is a case study approach [47], drawing Travelcard, considering modes, zones, and off-peak travel. For TfL, the drivers behind the Oyster are to encourage public I don’t really realise that it was charging me, ... so I went to transport use through reducing barriers to access, to improve customer service, I talked to them, and said, “oh, it’s not bus journey times by removing cash from bus boarding, and to working”, and they said, “oh, you know you have, like, a limit, free staff from the ticket office for other customer service you haven’t paid your prepay [Pay-as-you-go]”. (Focus group activities. member). But in keeping track of the value of spending on fares, the 3.2.1 The Oystercard in Practice blankness of the Oystercard contrasts with the embodiment of The Oystercard is touched onto touch-pads at ticket gates, on traditional tickets epitomised by single-use “Carnet” or “Bus open DLR stations, and on board buses. Period tickets are Saver” tickets, where the link between a paper ticket and the checked, or, for Pay-as-you-go, the cost of the fare is deducted. fare value that it represents could not be more direct: Value - either period tickets or Pay-as-you-go money value - …the exact opposite of the, the way the Oyster card works with can be added at ticket machines, at some local shops, or by pre- these, all these secret sort of amounts that go off or on you ordering online and then collecting (downloading) the value don’t actually see, is the bus tickets with the little octagonal when passing through a ticket gate at a nominated station. Auto tabs that you tear off, you know exactly how many you’ve used, Top-up, introduced between September 2005 and June 2006, you know exactly how many you’ve got left, it’s very, very automatically adds a fixed amount whenever the value falls simple. (Focus group member). below £5. This avoids the recurring need to pre-order and to pass through a station (it works on buses as well), but requires 3.2.4 Discussion pre-registration with TfL and still needs an initial pass through Collection of fares is clearly an important part of transport a nominated station to complete the configuration. service provision. Oyster therefore represents the intervention The TfL Oyster implementation centres on an artefact, the of electronic information systems into a traditional application Oystercard; but this is only a part of the larger Prestige project area. As such, it is an example of the mundane, daily activities to upgrade all London Underground and bus ticketing systems. through which people encounter computers in their routine In practical use, other artefacts - readers on ticket gates and lives, usually in passing and peripherally to their central machines, access to Oyster accounts online - are essential parts experience and volition. of the Oyster system as a whole. The provision of unified ticket availability across all transport modes is particularly significant for transport integration, a 3.2.2 The Case Study stated transport priority [16]. It is all the more unfortunate that For this study, 13 transport users took part in focus groups and Oyster Pay-as-you-go is not available, for political and 10 Oystercard users were interviewed individually. This organisational reasons, on many TOC services; in London, research took place between January and December, 2004, thus many inner-suburban areas are served only by TOC lines. In covering several phases in the introduction of the card. Ten addition to the inconvenience of the need for TOC users to buy senior management staff and 3 representatives of transport one-off single tickets, this adds a layer of complexity to an advocacy groups were interviewed. An observation of the already complex fare structure; it adds to a users’ mental load, Oystercard in use at a very busy Underground station was also and reduces confidence in the system. the opportunity for informal interviews with front-line staff. The written corpus included detailed analysis of relevant 3.3 The Central London Congestion Charge sections of the GLA Transport Strategy [16], explanatory The Central London Congestion Charge (CLCC) is a road leaflets, press reports, committee minutes, Web pages, and charging scheme covering a single zone in central London. reports relating to areas of special interest: rail services and Charging is based on camera recording of number plates smartcards (a total of 25 documents). (Automatic Number Plate Recognition, ANPR) at entry and exit from the zone and while driving in the zone. As part of the 3.2.3 Experiences of Oyster in Use integrated transport strategy, the CLCC is aimed primarily at For TfL, the Oystercard provides revenue protection and new reducing congestion, and at encouraging the use of forms of fare options, it enables cashless travel services, and possibly transport other than private vehicles. savings in the cost of sales. For a transport user, the Oystercard affords storage of Pay-as-you-go value or season tickets, it 3.3.1 The CLCC in Practice affords opening of ticket gates, and it affords use without Because the system does not work using electronic tags or other removal from a wallet. However, unlike paper tickets, it does vehicle modification, vehicles entering the charging zone are not afford visibility of the stored value without the mediation of recorded without any need for modification or pre-registration some third device, such as a ticket machine at a station. on the part of the driver. This presents problems for the user, since ticket machines are It is only in payment of the charge that drivers interact directly not always ready-to-hand; card-readers on buses and ticket with the CLCC system. The charge can be paid through a gates provide additional visibility of the stored value, but these number of payment channels: at some local shops and petrol displays are small, and the time available to read them is very stations, online, by phone to a call centre, at machines in some short. car-parks, and, if pre-registered, by formatted SMS text Provided he or she trusts the system, an Oystercard user does message, or by post. need to track the value on their card; if using a period ticket, then, on expiry of the ticket, or if travelling beyond its zones, Pay-as-you-go value may be deducted automatically. However, In January, 2007, the Association of Train Operating this may conflict with the user’s understanding of the system: Companies accepted TfL’s proposals to enable Pay-as-you-go When I first used it, I don’t really realise, I [went] to visit my by all London train operators [2]; some train services are friend, in High Barnet, which is like zone 5 and, “Oh, my God”, already available for Pay-as-you-go, or soon will be. There are high penalties, initially £80 (reduced to £40 if paid 3.3.4 Discussion within 2 weeks) for failure to pay the charge by the deadline . The CLCC is the most high-profile of the implementations As originally implemented and as in force at the time of this studied here, not only in terms of public awareness, but also as research, the charge had to be paid before midnight on the day a political initiative which forms one of the main policies of the of travel; if paid after 10:00pm, there was an additional £5 mayor of London. As such, it raised particular issues for surcharge. usability; problems at this level could have led to political failure. 3.3.2 The Case Study The charge was introduced in February 2003, that is, during the For this study, 50 CLCC payers were interviewed between mid- first term of office of the mayor of London, who had included a January and mid-February 2005 using a structured interview, promise to introduce a congestion charge in his election followed by 10 in-depth interviews from a mixed sample manifesto and argued strongly for it. Timing of the including drivers and non-drivers. There were also interviews implementation was therefore crucial; this was constrained not with staff responsible for implementation, particularly for only by the political timetable, but also by the need to enhance improving the customer experience, and with high-profile public transport provision in readiness [3]. As a further campaigners both for and against the CLCC, a total of 7 consideration to make the CLCC more acceptable to the public, management and policy personnel. revenues for the scheme are used to fund public transport. The Road Charging Options for London report [18] and the These balances and changes to details of policy are continuing. relevant sections of the Greater London Authority Act 1999 For example, when the charging zone was extended in [19] were analysed in depth, together with policy documents February, 2007, as part of the discussion around the extension such as party manifestos, committee minutes and scrutiny of the zone, the deadline for payment was extended to midnight reports, TfL and GLA press releases and other press comment, on the day following the day of travel (this change was actually and impacts assessments (a total of over 30 documents). introduced in June, 2006) [17]. 3.3.3 Experiences in Payment of the CLCC 3.4 The TfL Journey Planner In the case of the CLCC, users are involved with conscious The final study focuses on the TfL Journey Planner. This is effort only in interactions in the process of payment. The similar to a number of other electronic journey planners in that process of payment, however, imposes on drivers the need to it suggests preferred routes between a start and an end point, but pay by the deadline, and to remember to pay. In this, it it specifically gives journey information for multi-modal, resembles the payment of transport fares, but, unlike fares, there sustainable or public transport in London. It is real-time, in the are no physical barriers and no warning signs to act as sense that it gives warnings of known current problems and reminders. Payment of the CLCC is peripheral to people’s takes the time of travel into account. ordinary awareness; it is “not at the top of their mind”: By providing information and route-planning on public …you get home at nine-thirty, you've got half an hour to pay it, transport, walking, and cycling, the TfL Planner (JP) aims to but you come in .. I mean, sometimes people have been here, or encourage use of sustainable modes of transport (public something like that, you think, you know, you're not, it's not on transport, walking, cycling) over unsustainable modes such as top of my mind, really, ... but the worst one comes when you single-occupancy driving. It also aims to support socially forget altogether, 'cos you've relaxed and had a glass of wine, inclusive empowerment of service users through accurate, clear, or something (laugh), it's often happened, and then it's forty and real-time information. Real-time information could quid. (Individual interview). potentially give to public transport travel some of the flexibility of the private car, by enabling travellers to bypass disruptions It is perhaps unsurprising, given the consequences of failure to or to reschedule travel plans dynamically [26]. pay, that drivers are concerned to have tangible evidence of payment. Although payment of the CLCC is a licence in legal terms, no physical license is issued. Having heard reports from 3.4.1 The TfL Journey Planner in Practice colleagues of payment problems during the early days of the As considered by TfL, the JP across various channels is part of CLCC, this payer has devised ways to ensure something 6 key information touchpoints, which also include the TfL slightly more concrete: information call centre, staff at stations, and Travel Information Centres, and TfL Web, mobile, and interactive TV portals. I always print out a copy of the receipt, just in case, and I email it to myself and make a note in my diary (laugh), well, because However, there are many other information resources available there were so many problems initially. (Individual interview). to London transport users, including written information in printed maps, leaflets, public displays at stations and bus-stops, For this charge-payer, electronic information does not feel personal knowledge, information from fellow travellers, and sufficiently secure. As well as these subjective needs, there are ‘home-made’ genres such as print-outs of street maps or hand- sometimes practical needs for a proof of payment, for example written notes. in order to claim expenses. For such needs, the fluidity and uncertainty of information on the screen is insufficient; a The laboratory observations identified some detailed usability printed copy, although originating in electronic data, provides problems at the Web interface. Some of these were simple greater certainty. issues: difficulties locating the desired information, information off the bottom of the screen, information not what had been expected. More subtle problems were found in the quality of the suggested routes and in searching for a location; sometimes the participant could identify a better route, or very short bus trips were suggested, for example. Automated payment schemes exist for fleet vehicles, but not for private cars. 3.4.2 The Case Study 3.4.4 Discussion For this study, in-depth interviews were conducted with 20 Information from the JP centres on a Web site, but is also individual JP users, and focus groups with a total of 15 available over other channels: as a text message by sending a participants, between July and September, 2005. Data collection specially-formatted SMS message, at kiosks in stations and on also included field observations of conventional and electronic the street, using PDAs with WiFi access, using WAP from a sources of information in use. Documentary sources included mobile phone. business plans showing the importance TfL places on the JP, However, it is in working together with other information reports on customer information provision by transport resources that the JP it most useful. For users, there is not a organizations, and reports of the JP and other journey planners, clear distinction between the JP and other sources of a total of 10 documents. The study also included information; these resources are electronic, traditional, or social laboratory-based observations of a sample of 11 users as people ask friends or staff for advice, taking information as performing scenario-based tasks; these observations were they find it and need it. voice- and screen-recorded for later detailed analysis. 4. RE-CONSIDERING USABILITY IN 3.4.3 Experiences in Urban Navigation with the JP THE LIGHT OF USERS’ EXPERIENCES The Journey Planner is an essentially mobile application, in the sense that travel information is potentially most useful when the It is now possible to apply the understanding of usability user is away from home or desk [26]; it might be expected that developed in Section 2 to the case studies described in the mobile information channels would be widely used. However, preceding section, to illuminate the ways in which these e- the evidence from these interviews and observations refutes this government systems meet, or fail to meet, the needs of their expectation, at least at the current state of development of users. mobile information resources. It is not simply that there are usability problems. A common Analysis of in-depth interviews shows, rather, that transport thread running through the cases is the ways in which mundane users navigating the system make use of a constellation of contingencies of daily life combine with technological systems unconsidered resources, working with the JP, forming an in the lived experiences of transport users. These form what ecology of information; objects found serendipitously to hand have been called ecologies, human, social spaces alongside provide information and resources for action. In the choice of more or less sophisticated technologies as well as simpler these information sources, the physical properties of media artefacts [35]. come to the fore. Printed cycle maps, available free of charge But these grounded experiences of service users have from TfL, for example, were noted by several interviewees as implications for meeting policy aims, not only in the convenient to carry even for non-cyclists, smaller and lighter rationalistic, direct ways proposed by policy-makers, but in than a standard street atlas. Conversely, printouts from the JP more complex ways, as the systems are interwoven into the can run to several pages and are unwieldy; the JP has a section positive or negative experiences of service users. dedicated to cycle routes, but interviewees reported that these maps are too large to be useful while actually cycling. 4.1 Meeting the Users’ Needs in Action A particularly cogent example of the failure of information Finding that an Oystercard has expired, being unsure as to resources to work together was mentioned in several focus whether a CLCC payment has succeeded, or not having a map groups. As a mode of transport, buses present special to hand when travelling round London, are simple, almost challenges to a traveller to learn the complex and frequently- trivial, everyday problems. However, in the experiences of changing routes. Having identified a bus route using the JP, the service users, these apparently small problems can lead to a passenger then has the problem of knowing when the great deal of inconvenience and annoyance. destination has been reached. It is not always possible to see the name of the bus-stop from inside a bus, and, unlike in some Rushing to a shop or a computer to pay the CLCC, or other cities, buses in London do not indicate either visually or re-organising a journey in order to top-up an Oyster card, are audibly the names of the upcoming bus-stops for the benefit of deeply negative experiences, directly arising from a failure of passengers. Thus, the JP information alone is insufficient; design to take full account of users’ real needs. Such negativity service users must find other sources of information, such as can lead to antagonism. In interviews, a number of people who street signs, to complete their journey successfully. were in general supportive of the CLCC recounted strongly negative experiences with payment, particularly around This combining of different information resources as needed is forgetting to pay; some who already held negative views of the also found at the central interface of the JP, the Web site. To CLCC had those views reinforced by these usability problems. overcome some of the shortcomings of the JP interface, observation participants were seen to make use of online maps These usability shortcomings have their origins failures to to provide additional information in searches where the JP was match interactions to the context. The literature of ubiquitous unable to identify a route from the known data: computing is rich in considerations of context-aware computing; but, as Dourish [12] elaborates, context is not I never had this much fun. I used to put in a [postcode] and just something separate from interaction, but rather context is a get it and keep going. (Laboratory observation). mutual achievement by parties to an interaction as part of their everyday, common-sense understandings of the social world. In this case, it is clear that this user is familiar with the JP but usually searches using personal knowledge, in this case the To take another example from the studies, Section 3.2.3 showed postcode; that is, personal knowledge is also an information the importance, for the user, of the knowing the value stored on resource, to be drawn on as necessary and in a way which the an Oystercard. If an Oystercard affords gate-opening or bus- user would not normally remark on. boarding, then all is well, but unexpectedly running out of value on the card could lead to the very physical experience of being refused entry to public transport. Overcoming this breakdown, to read the Oystercard or to download value to it, involves an Underlying these issues, though, is a more fundamental ecology of artefacts, which must work together; but since ticket question: many of these interactions place responsibility on the machines, card readers, and so on, are not readily available in users to perform functions which might once have been many situations, this overcoming of the problem is only partial, performed by public servants. Enabled by e-government, the and contingent on circumstances. service user performs the interaction while public servants, if present at all, merely monitor and assist [21]. This is not only for economic reasons, but might also be preferred by a user, 4.2 Transparency and Abstraction being perhaps faster or more convenient. Many of the interactions encountered in these studies – passing through a ticket barrier, glancing at a Tube map, driving past an However, passing responsibility down to the user is also a ANPR camera while hardly being aware of it - are at the usability issue; in Nielsen’s terms, placing this onus onto users periphery of consciousness; they are transparent [24], hardly adds to their memory load [36]. The user has to grasp the need noticed by the transport user. This is how it should be; without to perform the interaction, often without any prompting - this, there would be breakdowns in the smooth flow of action. contrary to Nielsen’s ‘recognition rather than recall’ heuristic. However, breakdowns can also be positive; sometimes, For example, in addition to the need to ensure that they are in examining the unconsidered flow beneficial, leading to new possession of a current ticket, common to traditional tickets as ways of seeing things. well, the user of an Oystercard also has to ensure that they Transparency in this sense is a form of abstraction, hiding from “touch in and touch out” against the card reader at the start and the user the underlying details of implementation. Abstraction end of every journey. As from November, 2006, failure to do so is the fundamental way to manage complexity in information results in a higher fare being deducted from Pay-as-you-go systems, “the very stuff of system design” [13]. Yet, as Dourish users. Thus, where once a transport user would have paid cash and Button also point out, abstraction can also lead to failures in exchange for a simple ticket, now there is an additional because “information hiding” may make important information responsibility to ensure that the fare is paid ‘correctly’. Thus, unavailable to the user. although abstraction masks the need to calculate the correct fare, a new and different responsibility is placed on the user to The Oyster system provides some clear examples of the ensure that the collection is completed, in circumstances in contingent need to expose some of the underlying details of the which this is not always easy. system. The abilities of the information systems to mask the complexities of the transport system and to provide a seamless This ‘downloading’ of responsibility is even clearer in the case journey are compromised by real-world realities. The fares of the CLCC, in which the licence and enforcement model scheme is complex, and has some arcane exceptions; perhaps places on the charge-payer the risk of incurring a large Penalty the most cogent example is the non-availability of Oyster Pay- Charge. In response, in the practices discussed in Section 3.3.3, as-you-go on some train services. To avoid running out of charge-payers are anxious to avoid payment problems, for value, the user does have to be aware of some details of the fare example, by printing out electronic receipts. structure and of a complex set of associated rules. In the case of the CLCC, it is possible to envisage systems In a similar way, the TfL Journey Planner provides an which could, conversely, make use of technology to alleviate abstracted view of the transport system, without which it would some of the responsibility from the user; some trials of location not be usable or useful. Yet users do need to push and prod [13] identification technologies and of ‘tag and beacon’ have been the system to obtain the information they need, and, in some undertaken by TfL [43]. However, even if such a scheme were cases, become skilled in doing so; an interviewee described to be used, not all vehicles entering the zone would be fitted searching for a route in two separate stages, having found that with the necessary on-board units or other equipment. only in this way would it give the information that they wanted. Perhaps suggesting a counterbalance to this general impetus, A strong finding from the interviews, which builds on some Oyster offers an example of the use of technology to remove foundational HCI research [42], is the ability of users to some of the responsibility from the user, even while adding new understand and to overcome problems with the systems. Service responsibilities in the form of the need to “touch in and touch users developed narratives to make sense of the system and are out”, since the Auto Top-up system removes the need for a user able to perceive and comprehend what Gaver [14] calls to remember to add value to their card - once they have complex sequential affordances and hidden affordances. It is registered it. This is only a partial solution, since this implies a this working understanding which enables service users to new need to trust TfL to calculate and deduct fares correctly. overcome situations in which the systems fail to work together to match their everyday needs. 4.4 Analysis: Technology in Support of The point for policy-makers, and HCI designers advising them, Policy Aims is that interactions at the interface eventually lead to some real- The basic premise which opened this paper is that world practices, in which details such as distance to the next e-government is implemented with the aim of supporting public bus-stop become immediately relevant. These details, directly policy outcomes in various ways. Each of the implementations observable in the physical world, are sometimes hard to link to described in this paper is broadly in support of a rather simple their corresponding representation in the virtual JP world. policy aim: to encourage the use of sustainable modes of urban Systems must make the resources available to users; failure to transport, and discourage the use of the private car. do so could have very non-virtual, practical consequences. Section 1 suggested some at a theoretical level some ways in which policy might be supported or threatened by good or poor 4.3 Placing the Onus on the User usability. The findings in Section 3 and the analysis in the Up to this point, this section has considered failures to meet the subsections above give concreteness to the theory, in the practical needs of transport users, and abstractions which mask positive or negative experiences of service users. necessary information from service users, and it has shown some of the strategies available to overcome these problems. The continuing response by e-government service providers, for 5.1 Tensions Between Policy and Practice example in the extension to the CLCC payment deadline, A traditional approach to public policy-making would envisage suggests that there is a concern for usability by policy-makers, professional analysts proposing rational strategies to achieve but a lack of understanding that implementation cannot be political objectives set by decision-makers [3]. separated from the fundamental design of policy. This reflects a This traditional view is challenged both by political realities systems rationalism which considers the embodied experiences and experience in practice. On the level of interaction, what of service users to be peripheral; in contrast, the view taken in actually happens locally in mundane encounters is local, this paper places experience in use of the systems at the centre intimate, and flexible [40]; on the other hand, public policies, and, therefore, brings it into all stages of policy-making. providing baseline standards and continuity, are global and So, for example, the Oystercard is not simply an large-scale. Policies made in debate and in policy documents implementation of a policy to upgrade ticketing systems and to are experienced and interpreted locally. Local breakdowns, make public transport more attractive, but is part of the policy. perhaps unconsidered or hard to see at the design level, are The feelings of uncertainty around the need for users to “touch immediately perceived by service users; the larger breakdown, in and touch out”, mis-understandings about some of the harder to see and to deal with, might be a lack of co-ordination charges made, the “secret sort of amounts that go off or on” the between service providers and the policy level. card, and the need to trust the correct calculation of fares, are One of the mayor of London’s statements on changes to the inseparable from the larger policy. The uncertainty is charging scheme shows the balancing of differing realities compounded by the lack of visibility of the stored value and the particularly clearly [17]. The extension of the payment practical difficulties arising from breakdowns in the ecology, deadline, on the recommendation of TfL, is an implicit for example the need to download value at a pre-nominated acknowledgement that paying within the day does not station. necessarily “match the real world” [36]. But, as outlined in In the JP, the challenge is rather different: policy objectives practical terms in Section 3.3.4, this change was not made as a may fail if they are based on a rationalist assumption that plans one-off revision to the scheme, but as part of same variation for sustainable travel will in some way of themselves encourage order that extended the charging zone into west London; that is, the use of sustainable transport modes. The study found that the as part of a package of political decisions, carefully balanced to JP is used as just one of many information resources, and in be acceptable to a range of stakeholders. ways which do not conform to a simple planning model. For policy-makers aiming to encourage the use of sustainable 5.2 Towards Usable e-Government transport, the lesson is that, rather than concentrating on the JP However, timescale, cost, and other competing pressures, as alone to provide travel plans, it is more appropriate to ensure well as the scale of implementation, are constraints on policy that navigational resources are coherent, thorough, accurate, and its implementation. Meanwhile, various strategies are and timely, and work together with other information resources applied to manage usability shortcomings: phased introduction to support navigation. of services, using the authority of government to oblige the use of particular services, or providing a choice of channels. 5. USABILITY IN POLICY DESIGN Each of these strategies is problematic. Phased introduction E-government provides a privileged position from which to may allow for post-implementation changes to overcome observe usability in the mundane practices of heterogeneous usability issues, but, whilst a continuing, dialogical users. E-Government may provide useful services for the development of systems is welcome, it would be preferable to citizen, in which interaction flows smoothly, peripheral to the prevent usability problems prior to implementation. users’ main concerns. On the other hand, citizens may be A variety of channels is an appropriate way of addressing the obliged to interact with services to undertake a legal duty, such different situations in which services may be encountered, but as renewing a licence or paying a charge, which for most people choice alone does not overcome fundamental problems if the are not so much a service as a nuisance [30]. choice is among channels each of which is unsatisfactory. The Usability is a fundamental in both of these types of interaction, analysis of the ways in which systems sometimes fail to meet whether to enable pre-reflective, spontaneous performance of users’ needs (Section 4.1) demonstrates that, in everyday routine actions, or to reduce as far as possible the situations, often no single channel can provide adequate, usable inconvenience of citizens’ obligations. Yet policy-makers have support for the desired action. not often considered their analyses in HCI terms, and What is needed is systematic analysis of usability concerns at conversely, HCI practitioners do not in general consider the all stages, from the start of the design of public policy. This implications of their work for public policy. leads to a didactic message, familiar to HCI research [7] but Underlying much e-government policy is an implicit often neglected in public policy: usability design must be assumption that it is sufficient for services to be provided in incorporated in, not added on to, systems design at all stages; it ways which are more cost-effective, information more readily cannot be left to implementation. available and more timely, and access to services faster and easier. For reasons outlined in Section 1, this ignores the 6. THE WAY FORWARD essential usability of services, and the lived experiences with How can policy-makers actualise this need? Drawing on the those services. It is in a thorough understanding of everyday findings of these case studies, this paper makes its final details, and of what Suchman [42] calls the somehow which contribution with some suggestions of practical ways in which links daily practice to the ability make use of the systems, and this might be done in public policy design. links the use and usefulness of systems to public policy goals, that HCI can make its strongest contribution. Designers can start by developing an involved understanding of the situatedness [42] of interactions. More than the context of use, interacting with systems is a continual process of making sense of them and of their possibilities for action, in order to [3] Banister, D. Critical pragmatism and congestion charging achieve practical goals. in London. International Social Science Journal, 55, 2 (2003), 249-264. These are the goals of the users, rather than those of the service [4] Bannon, L. J., and Bødker, S. Beyond the interface: provider. E-Government is implemented to support policy aims; Encountering artifacts in use. In J. M. Carroll (ed.), these are the ‘drivers’ of the systems, built into strategy Designing interaction: Psychology at the human-computer documents and business plans. But the goals of the users are in interface. Cambridge University Press, Cambridge, UK, their lived experiences, in their personal desires and practical 1991, 227-253. needs; the systems as such are of little interest to them. Cooper [8] makes clear the distinction between goals and tasks; design [5] Bødker, S. Through the Interface: A Human Activity should be at the level of goals, rather than tasks. For a user, Approach to User Interface Design. Lawrence Erlbaum payment of the CLCC is not a goal, but is rightly called a task; Associates, Hillsdale, NJ, 1991. their goal is to drive in central London, and this, in turn, is part [6] Boland, R. J. J. Information systems use as a hermeneutic of whatever larger goal requires them to drive. process. In Proceedings of IFIP TC8 WG8.2. (Copenhagen, Denmark, 1990). 1990, 439-458. The sometimes antagonistic responses of users to local breakdowns reflect the tension between users’ goals and those [7] Chrusch, M. Seven great myths of usability. Interactions, of policy-makers, and between policy and its realisation. Local (September & October, 2000), 13-16. breakdowns are immediately apparent to users, but this [8] Cooper, A. The Inmates are Running the Asylum: Why immediacy and familiarity offers a potentially useful resource High-Tech Products Drive Us Crazy and How to Restore for policy-makers and designers. If there is resistance or the Sanity. SAMS, Indianapolis, IN, 1999. antagonism on the part of the user, citizen or customer, the [9] Deci, E. L., and Ryan, R. M. The support of autonomy and point is not to ‘solve’ the resistance, but to use it as a clue to the control of behavior. Journal of Personality and Social what went wrong and for suggestions for improvement [32]. Psychology, 53, 6 (1987), 1024-37. Involving the users at the earliest stages makes such local [10] Dillon, A. Group dynamics meet cognition: Combining breakdowns visible to designers, overcoming mis-matches socio-technical concepts and usability engineering in the between systems and the world of users. design of information systems. In E. Coakes, D. Willis, Thinking about ‘the user’ as an ill-defined ‘factor’ is to ignore and R. Lloyd-Jones (eds.), The new SocioTech: Graffiti on the users as real, varied human beings, with needs and desires. the long wall. Springer-Verlag, London, 2000, 119-125. McCarthy and Wright [33] have developed a profound [11] Dillon, A. Beyond Usability: Process, Outcome, and Affect understanding that, as technology is more and more involved in in Human Computer Interactions. Paper presented as the daily life, so computer power is not simply used, but lived with. Lazerow Lecture at the Faculty of Information Studies, Designing technology and living with it are both emotional and University of Toronto, 2001. aesthetic experiences; bringing users into design overcomes the [12] Dourish, P. Where the Action Is: The Foundations of separation of design and use, restoring continuity between Embodied Interaction. MIT Press, Cambridge, MA, 2001. design and lived experience [33]. [13] Dourish, P., and Button, G. On 'technomethodology': This paper has been concerned to emphasise that the physical, Foundational relationships between ethnomethodology and situated world is not something separate from abstract reason, system design. Human-Computer Interaction, 13, 4, but that both design and use occur in interaction with the (1998), 395-432. concrete world. Seen in this way, the material properties of [14] Gaver, W. W. Technology affordances. In Proceedings of artefacts provide affordances, or resources, for social action. the SIGCHI conference on Human factors in computing This suggests that designers should be asking: in what ways systems (CHI 1991). ACM Press, New York, NY, 1991, does the new ecology of affordances support the policy, and 79-84. what is lost between the old and the new? For example, unconsidered features of a simple payment system, the [15] Gibson, J. J. The Ecological Approach to Visual tangibility of a paper slip, or the personal knowledge of a Perception. Houghton Mifflin Company, Boston, MA, regular customer, are all part of the ecology of the system. [16] GLA. The Mayor's Transport Strategy: Executive The systems studied in this paper show numerous examples of policy-makers responding to usability problems following summary. GLA, 2001. At http://www.london.gov.uk/mayor/strategies/transport/ resistance or negative responses from the public. While these pdf/final_execsumm.pdf, Last retrieved: 06/06/2007. improvements may be seen as positive to the extent that they reflect dialogue between service users and politicians, this does [17] GLA. Statement by the Mayor Concerning his Decision to not overcome already negative experiences of users resulting Confirm the Variation Order for the Western Extension of from a failure to recognise the need for usability to be part of the Central London Congestion Charging Zone with the design of the system prior to implementation. Modifications. GLA, 2005. At http://www.london.gov.uk/mayor/congest/docs/mayor- statement-092005.pdf, Last retrieved: 06/06/2007. 7. REFERENCES [1] Alvesson, M., and Sköldberg, K. Reflexive Methodology: [18] Government Office for London. The Road Charging New Vistas for Qualitative Research. Sage, London, UK, Options for London (ROCOL) Report. GOL, 2000. At http://www.gos.gov.uk/gol/transport/ 161558/228862/228869/, Last retrieved: 06/06/2007. [2] Association of Train Operating Companies. ATOC's Position on Oyster Pay-as-you-go and the Offer by [19] Greater London Authority Act, 1999: Schedule 23: Road Transport for London. Press Release, January, 2007. At User Charging. http://www.atoc-comms.org/dynamic/atocpress- [20] Grimsley, M., and Meehan, A. e-Go vernmen t informatio n story.php?atoc=997793, Last retrieved: 06/06/2007. sy s tems: E valuatio n-led desig n for pu b lic value an d clien t tru st. Electronic Journal of Information Systems, [34] McGrenere, J., and Ho, W. Affordances: Clarifying and 16, 2 (2007), 134-148. evolving a concept. In Proceedings of Graphics Interface 2000. (Montréal, Québec, Canada, 2000). 2000, 179-186. [21] Grönlund, Å. E-democracy and e-government: State of the art. In Proceedings of 68th IFLA Council and General [35] Nardi, B. A., and O'Day, V. L. Information Ecologies: Conference (Glasgow, UK, 2002). 2002. Using Technology with Heart. MIT Press, Cambridge, MA, 1999. [22] Grudin, J., Interface: An evolving concept. Communications of the ACM, 36, 4 (1993), 110-119. [36] Nielsen, J. Ten Usability Heuristics. Useit.com, 2005. At http://www.useit.com/papers/heuristic/heuristic_list.html, [23] Hood, C. C. The Tools of Government (First American Last retrieved: 06/06/2007 Edition). Chatham House Publishers, Chatham, NJ, 1983. [37] Norman, D. A. The Design of Everyday Things. Basic [24] Ihde, D. Technics and Praxis. D. Reidel Publishing Books, New York, NY, 1988. Company, Dordrecht, Netherlands, 1979. [25] ISO 9241-11 Ergonomic Requirements for Office Work [38] Potter, J., and Wetherell, M. Discourse and Social with Visual Display Terminals (VDTs), Part 11: Guidance Psychology: Beyond Attitudes and Behaviour. Sage, London, UK, 1987. on Usability. International Organization for Standardization, 1998. [39] Spinuzzi, C. I. Describing Assemblages: Genre Sets, [26] Jain, J. Bypassing and WAPing: Reconfiguring timetables Systems, Repertoires, and Ecologies. Computer Writing for 'real-time' mobility. In M. Sheller, and J. Urry (eds.), and Research Lab White Paper, University of Texas, Mobile technologies of the city. Routledge, Abingdon, UK, Austin, TX, 2004. At 2006, 79-101. http://www.cwrl.utexas.edu/files/040505-2.pdf, Last retrieved: 06/06/2007 [27] Kapor, M. A software design manifesto. In T. Winograd (ed.), Bringing design to software. Addison-Wesley [40] Star, S. L., and Ruhleder, K. Steps toward an ecology Professional, Reading, MA, 1996. infrastructure: Design and access for large information spaces. Information Systems Research, 7, 1 (1996), 111- [28] Kelly, G., Mulgan, G., and Muers, S. Creating Public Value: An Analytic Framework for Public Service Reform. Strategy Unit Discussion Paper, Cabinet Office, London, [41] Strauss, A. L., and Corbin, J. M. Basics of Qualitative UK, 2002. At Research: Techniques and Procedures for Developing http://www.cabinetoffice.gov.uk/strategy/downloads/files/ Grounded Theory. Sage, London, UK, 1998. public_value2.pdf, Last retrieved: 06/06/2007. [42] Suchman, L. A. Plans and Situated Actions: The Problem [29] Lamb, R., and Kling, R. Reconceptualising users as social of Human-Machine Communication. Cambridge actors in information systems research. MIS Quarterly, 27, University Press, Cambridge, UK, 1987. 2 (2003), 197-235. [43] TfL. London Congestion Charging Technology Trials: [30] Lenk, K., and Traunmüller, R. Electronic government: Stage 1 Report: February 2005, at Where are we heading? In Proceedings of Electronic http://www.tfl.gov.uk/assets/downloads/technology- government: First international conference (EGOV 2002). trials.pdf, Last retrieved: 06/06/2007 (Aix-en-Provence, France, 2002). 2002. [44] Thimbleby, H., Blandford, A., Cairns, P., Curzon, P., and [31] Margetts, H., and Dunleavy, P. Cultural Barriers to e- Jones, M. User interface design as systems design. In Government: Better Public Services through e- People and Computers XVI (Proceedings of HCI 2002). Government. Academic Article in support of Better Public Springer, London, UK, 2002, 281-301. Services through e-government, NAO, 2002. At http://www.governmentontheweb.org/downloads/papers/ [45] Weiser, M. The computer for the 21st Century. Scientific Cultural_Barriers.pdf, Last retrieved: 06/06/2007 American, 265, 3 (1991), 66-75. [32] Markus, M. L. Power, politics, and MIS implementation. [46] Winograd, T., and Flores, F. Understanding Computers Communications of the ACM, 26, 6 (1983), 430-444. and Cognition: A New Foundation for Design. Ablex, Norwood, NJ, 1986. [33] McCarthy, J., and Wright, P. C. Technology as Experience. MIT Press, Cambridge, MA, 2004. [47] Yin, R. K. Case Study Research: Design and Methods (Second Edition). Sage, London, UK, 1994.
Journal
Electronic Workshops in Computing – Unpaywall
Published: Jan 1, 2007