By Marion Walton, Multimedia Education Group, University of Cape Town
I am grateful for having been given the opportunity to attend CHI 2002 and participate in the Development Consortium, with its focus on HCI in South Africa. This was an incredible opportunity, which I know will have a major impact on my research and development work in future. The formal Development Consortium sessions, and many informal discussions were an opportunity to engage with other people doing HCI work in the South African context. Interestingly, these sustained discussions in a foreign place brought me closer to home, in that they helped me see many opportunities for local collaboration that I had been unable to see previously.
On the whole, my rather narrow concept of how HCI could be useful in a developing context was well and truly stretched. Until now, my ideas have always been PC-based, and my perspective was broadened considerably by the inspiring examples of people working with wireless technologies, and (in the conference more broadly) of various kinds of virtual/augmented realities.
Nonetheless, while these models are technically impressive, in very few cases are they designed for the realities of a developing country. In fact, the keynote speaker of the conference, science fiction writer and futurologist David Brin, seemed blithely ignorant of the realities of the world economy. He argued that a libertarian approach to surveillance technologies would democratize their use, and thus help usher in his ‘optimistic’ version of the future, where all (a category which the audience was quick to point out did not extend beyond the U.S., let alone to the third world) would all live in “a world where everyone gets it”. Given this blindness to our issues, a great deal of work remains to be done, as we concluded in our two-day development consortium workshop.
HCI in South Africa should move beyond the relatively superficial focus on 'interface' and 'culture', since ‘surface’ interface issues, such as language and icons, are not the crucial aspect in software use in South Africa.
Instead, we should get down to the important and tricky task of using the methods and insights of user-centred design to acknowledge and investigate the local realities of IT use in our context. This will allow us to understand local user groups and provide technology or methodology to let them create content / tools to support which can be adapted to their needs. For example, under local conditions where computers are scarce and often shared resources might suggest the need for software adapted for communal use (“communitisation”) rather than the Western model of being adapted to individual preferences (“personalisation”).
In addition, we should focus on imagining, designing, and fine-tuning applications for those users and potential users who are currently not served, or only inadequately served by the global software and hardware industries. The realities of the global economy should bias us strongly in favour of developing software using the open source model, and cheaper local programmers. While in other contexts translation of language is probably the major issue when software designed for one group of users is transferred to a another socially similar, though culturally different group, the huge social and economic diversity within the South African context, means that schooling and literacy (in the broader sense) are crucial issues. The issue of literacy must be addressed not only for 'localisation' but also for democratising access to technology.
Our Development Consortium discussion was enriched in many ways by the addition of HCI voices from outside South Africa, particularly from Latin America. It certainly became became clear later in the conference that, if HCI in South Africa rises to this challenge, many of the applications we develop may be useful, or at least adaptable in other developing countries, and vice versa. Particularly important to my experience of the conference were the sessions that dealt with HCI in other developing contexts, such as the Latin American development consortium follow-up SIG, and the Equity SIG.
I hope that we'll be able to follow through on these informal connections in the proposed issue of Interactions - rather than a narrow focus on South Africa, this issue could look at the broader issues confronting IT in use in the developing world.
Several discussions during the Development Consortium as well as during the main conference highlighted the relevance of the theme of HCI for the developing world. The following following emerged from these discussions:
This session was a report-back on the Latin American development consortium, and their progress towards achieving their goals of advancing HCI publications, communication and collaborative research in Latin America.
Discussion between South African, Indian and Latin American delegates who attended this session indicated that there were many common issues for HCI in the developing world, particularly in countries where an advanced IT infrastructure co-exists with levels of extreme poverty. These connections deserve future discussion, possibly in a new issue of Interactions.
This was a more broadly focused SIG which looked at issues of IT use and equality globally, with the emphasis on “looking at the users at hand” and answering the somewhat strangely phrased question “How can we make computation available to them”.
Again, the localisation and usability of open source software was raised as a key issue (Mattheus Blumenfeld), while other discussions included the introduction of mandatory computer-based voting in Brazil, the development of local content in the developing world, and the skewed power-relations and the Western bias perspective of much of the information and applications. An interesting person to follow up with was Lucy Suchman from Lancaster University, whose work practice studies are ethnographic investigations of planning and situated action.
Development Consortium participants came to the conclusion that specific issues should be taken up with ACM SIGCHI:
In terms of the conference more broadly and my own personal interests, I most enjoyed attending SIGs and meeting up with a group of people who work on user experience issues and 'funology'! This also broadened my horizons, and helped me realise that my background in literary and media studies, narrative issues and gaming theory has direct relevance to some of the newer and more challenging issues in HCI.
This workshop focused on the role of pleasure and fun in understanding the whole user experience, rather than merely limiting HCI investigations to the more functional concept of usability. (Of course I think they left out the power of horror as a key feature of ‘enjoyment’.) These factors are particularly for computer applications such as games and entertainment - I would say education as well! “Fun and enjoyment are set to be major issues as information and communication technology moves out of the office and into the living room”. The workshop set out to challenge the simplistic division between tool and toy – and to look at ways of combining insights into both tools and toys, and deploying their features according to goal-oriented requirements.
“Even limited field data is better than no data”
This workshop seemed very interesting, although it was a closed session, the notes were detailed enough to provide a useful introduction. It surveyed a range of field methods, including contextual inquiry and work modeling.
Given the pressures on our development schedules and our limited time and resources for user testing, I was particularly interested in the ‘Streamlined fieldwork methods’ discussed by Stephanie Rosenbaum.
These methods involved “exploration of users’ behaviour in the context of their own work” and “intensive observation of users’ settings and artifacts”:
Colleen Page described “Discount user observations”, which also require a two-person team, and thus dispense with the need for transcription, relying instead on detailed note-taking, and the use of a digital still camera for detailed info about the context. Immediate review of these notes and images allows the creation of ‘user scenarios’.
This study compared cultural problems with the use of user evaluation methods developed in the U.S.A. Some methods seemed inappropriate in different cultural contexts, particularly the ‘think-aloud’ method of observation. “Observation sessions were particularly problematic for the Japanese. They felt uncomfortable speaking out loud about their thoughts, and seemed to feel insecure because they could not confer with others to reach a common opinion.” English users needed reassurance before feeling comfortable to talk out loud, while North Americans seemed under pressure trying to give the ‘right answer’, while Japanese users were freer with their criticism. As such, the authors recommend that well-matched pairs or focus groups should be considered instead of the individual approach.
Given our work on adaptations of bloatware, this seemed a promising area to investigate, as it recommends an adaptable front-end to MSWord 2000, which creates a personalized interface, which adapts the menus according to the needs of the user. This approach has obvious uses in relation to teaching computer literacy, and to context-specific computer literacy training.
Mutual trust is necessary in order to increase cooperative behaviour in online contexts. This paper found that "Using text-chat to get acquainted is nearly as good as meeting, and even just seeing a picture is better than nothing, while having a static photograph of the partner is as effective in establishing trust, whereas a text-based, static information sheet of personal information is not".
Zhenge explained that he suspected the motivation for the cooperative behavior may be accountability, rather than other forms of trust, in other words, the fear of retribution if you run into the other person again!
Online ‘bad behaviour’ is prevalent, and very expensive. Website owners who rely on community-oriented features lose users because of other users’ bad behaviour, and thus lose money. For example, it has been shown that 79.3% of people have left an online environment because of bad behaviour.
It has been shown in previous research that online bad behaviour is related to the lack of a sense of the other’s realness and presence. Solutions to the problem of ‘bad behaviour’ all relate to ways of increasing social presence. These methods include using voice communication and providing personal profiles (Personal profiles provide some of the information (age, gender, social class, nationality etc) which people get from voice communication.)
This particular experiment compared bad behaviour in three different chat interfaces. The first used text chat, the second text and personal profile, and finally, the third investigated a text to speech interface. Interestingly, personal profiles had no effect on reducing bad behaviour, while the use of the text-to-speech approach did help.
Other approaches to reducing anonymity without forcing people to show their faces need to be investigated, such as that of reputation-building.
In this study, pre-teens were given wireless beads, which light up and blink, according to their needs. A version of the ‘tamagotchi’ game, the children's goals are to keep the animals happy by trading them. It was found that trading physical objects encouraged social interaction, and that a game environment was a safe context in which to do this
The rules of object exchange in this subculture are implicit and complex, and need to be acknowledged by designers. The role of trade and gift-giving seems to be underestimated in many online environments, and we should be looking for socially appropriate ways of introducing these game-like social features into our online environments.
This study demonstrated a chat interface which foregrounds the articulation of concepts that may not be shared by a group of users, and thus has clear educational implications.
This student poster was a very interesting beginning of a discussion of the theoretical shortcomings in of the information foraging model (already pointed out by Mantovani) Information foraging is inadequate, because it does not consider social foraging (e.g. species which rely on communication to get them to the ‘goodies’ - such as ants, bees and people!). The imaginative part of the poster related to an idea of applying the insights of collaborative filtering to the information foraging model.
This workshop promised to focus on how communities work with shared resources – documents, discussions, conventions, roles, identity, senior members. Michael Muller is a fascinating person who attended our Development Consortium and made some very insightful contributions.
“There are three major questions facing designers of on-line communities: how to get users to behave well, how to get users to contribute quality content, and how to get users to return and contribute on an ongoing basis.”
This study discusses a set of custom tools used on community web sites to collect user data and feed it back to the users, as a way of differentiating users according to their status in the community, and by visual representations which show how hot or cold a topic is. The site’s admin also apply a profanity metric, which is used to verify the behavior of ‘problem members’.
Users are aware that profanity is being looked for and is not tolerated. The sites used various approaches, which all aimed to increase social consciousness among users, including using their real names – this is possible given that the music-interest sites are a context where people are proud of their real life identity, interests and contributions.
Interestingly, this social consciousness was linked strongly to the absence of the role-playing which goes with anonymity ‘The absence of a high degree of role playing and anonymity within the community is a very important factor in creating accountability, real social consciousness and strong behavioral norms.”
Users are awarded status ‘points’ for the contributions they make to the community, as in the following example:
Point Total for Larry Magri
20 points for creating a home page
20 points for 2 published lesson pages
44 points for 11 resources
68 points for 17 resource reviews
376 points for 188 FretBuzz messages
259 points for general site usage (logging in, emailing links, sending feedback)
“the system greatly aids the community in self-governing”
Members can only accumulate large amounts of status points by giving generously to the community, which usually also indicates that they wish to act responsibly within it.
These visible status differentials lead to a “constant dialogue between invested members and new ones” (394) This leads to optimal rates of attrition and retention.
This is a fascinating ethnographic study of teenage mobile phone users, which reads the messages sent by teenagers and their rituals of phone use as examples of “gift giving”. The study is a critique of the technologically deterministic rhetoric which dominates mobile computing - i.e. as soon as a new technology appears (e.g. WAP) it is simply assumed that users will find uses for it, rather than technological requirements being born through the complex interplay between technology and its users.
deterministic rhetoric oversimplifies the relationship we, as social beings, have with technology (439)
[This rhetoric] conceals the fact that we, and the immensely rich and varied social worlds in which we live, have a profound impact on the ways technologies are understood and used, and subsequently evolve across time. (439)
The study argues that "a range of teenagers’ phone mediated activities closely resembles ritual gift-giving and that their participation in these activities has a significant impact on the ways in which they see and understand the use of mobile phones".
"phone-mediated activities were a routine, taken-for-granted part of teenagers’ daily encounters" (440)
"teenagers use mobile phones to embody shared meanings, amongst other things" (440)
"These meanings, we will demonstrate, are taken on through forms of ritual exchange"
"The obligations of exchange: to give, accept and reciprocate".
"Text messages are made special, or valuable, because of the memories they evoke".
As a result of these observations, the researchers suggested a range of design features which would allow users to extend the support phones provide for these ritual exchanges and which would allow them to keep private, and thus treasure and preserve the shared meanings embodied in their text messages (removable memory cards, and lockable messages).
This study evaluated cancer patients’ use of an online community space designed to support them through their very isolated convalescence period. During this period, patients and their care-givers are isolated, since the patients’ immune systems are severely compromised, Hutchworld, the 3D, multi-user space, included a chatroom, a bulletin board, a gift shop, which allowed patients to exchange free gifts, a music garden, and an info desk, which provided information about the city.
The study explored Internet and Hutchworld usage, with one group of patients being given access to Hutchworld, and to high speed Internet access, while a control group were not, but had their own computers.
Used primarily for e-mail and surfing web, interacting with family and friends. All forms of usage were higher, in the experimental group, who used the computer once a week or more for social interaction, also seeking information, and wanted more social interaction. A popular activity was looking at and updating their own profile, which allowed them to document their own progress in the recovery period.
Hours spent on computer positively correlated with measures of stress, but life satisfaction decreased a lot more for those not in the control group. "Cancer is very isolating and the computer broke that isolation.” However, it was found that there was not the critical mass required for synchronous interactions, asynchronous interactions were most important including the use of gifts, notes and the bulletin board.
Hutchworld seemed to buffer patients and caregivers against the reductions in life satisfaction and social support following their transplant procedures.
This very useful application is designed to circumvent the problem that the page is the smallest unit of information on the web. Hunter Gatherer allows users rather than the system to determine the smallest units of value, depending on their research needs. Browser support is not well-integrated to help you create your own context from sea of web pages. Hunter Gatherer reduces the number of steps required when collecting bits of information from a web search, which currently involves 11 discrete steps. Since the process is so tedious, people often skip the stage of collecting the URLs. HG allows users to gather ‘collections’ of related data from the Internet, which are rendered real-time. I tried to use HG as it seemed a promising tool for student web use, but unfortunately it requires a proxy server, which does not work with UCT web access.)
This was a fascinating field study of knowledge workers’ web use practices over two consecutive days. The study used a combined diary and stimulated recall interview approach (using the browser history).
The study is particularly useful in that it reviews studies of web usage, finding that most are from the library and information sciences literature, and rely primarily on surveys, using questionnaires, rating scale or interview data. In addition, studies of click-stream data are mined to reveal emergent patterns in web navigational patterns.
[This survey excluded the vast marketing-related studies by, e.g. Nielson netratings]
The study found that people’s activities could be divided into six major categories
Knowledge workers consider the goal oriented categories most important (the first two) and that web use is mainly solitary, though the results of the search are often shared afterwards.
The study discusses all the above activities, but I will focus on info-gathering, as it is most relevant to university research.
Many of these activities involved a set of questions to be answered, and the searchers did not know in advance how much information there was to be gathered. The activities often involved comparing and contrasting information across sites or organizations, and the use of trusted sites was a key issue.
"Part of this knowledge worker’s job was to know and assess which sites provided the best quality information and which did not" (230)
"Whatever the purpose, information gathering activities could be quite time-consuming and complex. More often than not, such meta-level tasks encompassed navigating multiple links and sites. It was clear that this process was very reliant not only on complex navigational patterns, but also on scanning and skimreading large amounts of material to assess its relevance. Another characteristic of these activities was that they often unfolded over time, sometimes spread out over days or even weeks. We found that 40% of these activities were not completed in a single session, being interrupted either due to demands or because of the amount of time required to complete them". (231)
This study is a very clear indication of the extent to which terms like "surfing" or "browsing" the Web "gloss over and even misrepresent what people use the web for".
"There are many different reasons why knowledge workers use the Web and many different behaviors they engage in: some goal-oriented and some not; some centered around focused questions and some not; some short and self-contained and some spread over time. Not surprisingly, these activities involve different patterns of use and types of information. In doing so, they may draw on other resources (documents technologies, and people)".
The conclusions of the study are also very interesting, in that they relate these various types of activity to the nature of the interface. It found that "finding" tasks were in fact easily adaptable to mobile use, and would benefit from bookmarking to facilitate speed of use. On the other hand, the nature of the activities involved with ‘information gathering’ tasks imply that these are probably not at all suited to small screen contexts, given that they require scanning large sections of text, complex navigation, and complex information management and archiving. In addition, laptops are not ideal for this type of work, given that users would need to check their knowledge against other sources, and use paper folders for filing etc.
Some other ideas to improve support for this kind of searching: Tagging, search tools, web scrapbooks, better history
Popout Prism is a Web browser that aids navigation by providing an enhanced thumbnail overview of Web pages. This enhanced thumbnail contains attention-grabbing "popouts" which are generated dynamically based on user input of URLs and keywords. These popouts enable users to immediately locate relevant information in a page. Popout Prism also adds popouts to full Web pages so that users can recognize and locate keywords more easily and efficiently
This SIG brought together people interested in discussing the importance of HCI training for new media and design professionals.
HCI courses are often designed for Computer Science majors, and media and design students are excluded from them. However, in design curricula which do address HCI, there seem to be two existing patterns:
1) Interdisciplinary HCI programs offered by Computer Science (e.g. Georgia Tech, Carnegie Mellon, University of Glasgow, University College, London)
2) NAC Curricula which integrate HCI (Indiana, Vrije University School of Informatics)
Tony Faiola discussed a task from his Indiana HCI course for design students. This illustrated the integrated approach. The course revolves around a conventional web design project which, however, includes:
There seemed a strong consensus that the practical research skills which students learn from an HCI approach are important in teaching students to look at design empirically and objectively. This was considered a necessary skill for primarily aesthetically trained design students. In addition, students also needed to learn that there are key differences between marketing data (which is often the only data companies collect) and good HCI information.
The discussion became particularly interesting when someone commented that Computer Science HCI students would also benefit from taking graphic design courses. A lively debate ensued about the advisability of developing students’ expertise broadly rather than in a narrow-discipline related way. However, the consensus position appeared to be that people have to specialize, but that it is essential that they get experience in collaborating in multidisciplinary teams, as they will have to do this in the workplace. Students should know what is involved in the collaborating disciplines (i.e. as a graphic designer, you know about usability engineering, but you can’t be an expert in all areas.) One lecturer commented "We’re not trying to produce Renaissance men". Instead, students need to have the experience of design as an interdisciplinary process, (by working as part of an interdisciplinary team on a real project). Most importantly, the experience of teamwork across disciplines was felt to be the key issue - in this way the students have the experience of working with students and teachers from a wide range of collaborating disciplines (Computer Science, Psychology, Design and Media). This was seen to be very important in helping them to understand the limitations of their own disciplinary background, and also in teaching them when to go to others for assistance.
Additional skills required in business and often not taught in design and new media courses include the need to write requirements, prepare testing, communicate in multidisciplinary teams, and question primarily business-oriented decisions.
Brenda Laurel has said, as a critique of the masculine imagination which, she claims, dominates much high-tech culture, cyberpunk fiction, and most visions of virtual reality “in general boys have this fantasy about leaving their bodies”. Rather than simply erasing the body, Stelarc is an artist who explores the extreme edges of this fantasy, testing the limits of current possibilities for the human-machine interface. Stelarc is an Australian artist whose artworks probe new and sometimes involuntary interfaces with the body (usually his own!) by using medical instruments, prosthetics, robotics, Virtual Reality systems and the Internet.
"He has performed with a THIRD HAND, a VIRTUAL ARM, a VIRTUAL BODY and a STOMACH SCULPTURE. He has acoustically and visually probed the body- having amplified brainwaves, blood-flow and muscle signals and filmed the inside of his lungs, stomach and colon, approximately two metres of internal space. He has done twenty-five body SUSPENSIONS with insertions into the skin, in different positions and varying situations in remote locations." (From Stelarc’s bio on his website at http://www.stelarc.va.com.au/
His presentation at CHI 2002 left me nauseated, confronted, as if for the first time, with the materiality, and fleshly limitations of my own human body, and those of the people around me. In his speech to us, Stelarc strongly opposed the Cartesian dualism which, in Western thinking, separates the human mind and body. He claimed that there is ‘no mind at all in the traditional sense”, arguing that, suspended by large hooks through his skin, hanging naked over New York city he was aware only of himself as “all body” throbbing, with intense, all-consuming pain. I found this profoundly ironic – what mere body would allow itself to be subjected to such indignities in the name of art? Nonetheless, Stelarc regards the human body as obsolete in its current form, and his art explores the ways in which he imagines the body will co-evolve with human technology. Just as he objectifies himself by speaking of himself in the third person as ‘the body’, his art dehumanises and defamiliarises the body, exploring it “as sculptural and architectural space”.
I must admit that, while utterly fascinated by his presentation, I found Stelarc’s enthusiasm for his cyborgian project profoundly disturbing. He tried to reassure us “This is no mere Faustian option nor should there be any Frankensteinian fear in tampering with the body.” I’m afraid this sincerely cheerful assurance didn't really work for me. Although I was often seduced by the enthusiasm and clarity with which he explained his ideas, I was repulsed at a deeply visceral level by some exceptionally grotesque image, or by the shock of the maniacal laughter with which he punctuated the presentation.
He showed us the following projects:
SUSPENSION - He suspends ‘the body’ (in fact his own) by hooks through the flesh above cityscapes and landscapes. NY police pulled him through a window demanding to see his ID.
FRACTAL FLESH - This was a touch-screen interfaced Muscle Stimulation System, enabling remote access, actuation and choreography of the body, experimenting with external, extended and virtual nervous systems for the body using the Internet.
EXOSKELETON - a pneumatically powered 6-legged walking machine actuated by arm gestures.
EXTRA EAR - a surgically constructed ear as an additional facial feature that coupled with a modem and a wearable computer will act as an Internet antenna, able to hear RealAudio sounds. This is planned for the future, as Stelarc has not been able to find a doctor who will agree to perform the surgery. However, the vision is that the ear will also speak to people who come close.
MOVATAR is an intelligent avatar that will be able to perform in the real world by possessing a physical body. It will have a sound feedback loop from the body giving the virtual entity an ear in the world.
EXTENDED ARM - a manipulator with eleven degrees-of-freedom that extends his arm to primate proportions. He explained that to him, prosthetic augmentation was a symptom of excess, not of lack. As such, the arm was equipped with voice recognition – and anyone could tell the arm what to do! He showed us slides of himself using his three arms to write the words “EVOLUTION” and ‘DECADENCE”.