The specialist and the non-barking dog

I have participated in a project where dr. Brenda Dervin has interviewed a large number of academics and practitioners about their co-operation and about co-operation between experts in HCI, information systems and medias. I was first interviewed and then asked together with a large number of other to write an essay based on some or all of the transcribed interviews. This is the result.

"Is there any point to which you would wish to draw my attention?"
"To the curious incident of the dog in the night-time."
"The dog did nothing in the nighttime."
"That was the curious incident," remarked Sherlock Holmes. (Doyle 1894)

My own area is HCI: Human Computer Interaction. I have worked as a product manager in a large electronics company, as usability specialist in another company and I am today associate professor in computer science at the University of Copenhagen. This means that my primary goal is to do research that makes it possible to design better technology and to use the technology in a better manner. I have also found that the most interesting sometimes is not what the different disciplines focus on, but rather what they more or less consciously leave out.

Usability testing

One of the participants in this study told what a wonderful service a usability lab was and how a test made it possible rapidly to learn what problems users had with an interface (040). In contrast another criticizes how "...people come in and they do the usability study of searching for some stuff on the web and the researchers aren't getting at all of the things that would really go into play if the tasks were being done in the real world with the users." Many usability tests are probably done in that manner. However, it is possible to do a usability test in a realistic setting, for instance to test an interface for a library system in an actual library and with tasks that users realistically will want to do with the system.

One participant describes the problems he experienced when deciding how to design catalogue cards and states that the lab approach has outlived its usefulness (037), even though the design problem that he describes is of a type that often easily are solved by doing a usability test.

It appears to be difficult to provide good arguments for or against usability tests in general. In contrast, it is much easier to discuss whether a test that is designed in a certain manner and with a specific group of participants is suitable for identifying problems experienced by the actual users.

Interface design

One of the participants noted that users of electronic media’s tend to read small fragmented chunks and that they have less experience reading what he calls a narrative or a sustained narrative (012). This can be explained as a consequence of reading being more difficult on a screen and the reading speed being only 80 to 90 percent of the speed on paper. This reduction in reading speed and ease of reading appears to be a serious disadvantage of electronic documents compared to paper-based documents, but I did not find it mentioned by any participants.

When the reading is more difficult, it may be considered important that users easily can scroll or skim through a large document and get an overview of it. Partly because of that, there have been a number of studies of scrolling and HCI tools. I only found one participant who mentioned scrolling (017), and he did not discuss scrolling when using a large text, but only scrolling on web pages.

One of the participants sees a need for visualizations that make it possible to get a quicker overview of large amounts of data. He seems to include formatting on the screen ­ what normally is described as graphical design ­ as part of visualization (042). That is interesting, because graphical design and visualization normally are seen as two separate areas in HCI. Another participants sees a demand for better visualization techniques to reduce data overload (018). However, a third complains that even though the techniques are interesting, users often cannot make sense of them (019). This suggests that there is a gab between the visualizations often presented in HCI literature and actual user needs.

Within HCI there is a large number of scientific studies of reading, scrolling, visualization and the best ways to design specific types of interfaces. What I find interesting is that participants in this study do not mention the need of such studies more often, even though a large proportion of the participants probably know about them. It appears that even participants who know such studies find that their value is limited.

Users versus situations

A substantial number of interaction designers have started to use personas ­ a description of a single possible and demanding user as basis for their design. However, I have not heard of anyone who tried to make a design based on "situationas".

In contrast, most participants focus on the situation or use or discuss users in the context of the situation of use. They talk about a situation or event-based approach (007, 031, 074), the complicated environments that users operate in (008), how we shall realize that information providers also may be information users (031) so the difference is more situation than person specific. One of the participants describes how it makes a difference how much the user knows or has worked with the information or whether it is used as a citizen, an artist or a teacher (032), and two of the participants who discuss mental models and cognitive psychology notes that it is necessary to study how the user reacts in specific situations (032, 033).

The first approach may be similar to when we talk about a person in daily life, whereas the second may be similar to cognitive psychology focusing on peoples’ perceptions and reactions in specific situations.

Research of user needs

Two participants describe that it is critical to know the motivations, needs and orientation of users and what they are trying to accomplish (001, 004). In reality that is also what marketing tries to do, only much faster and in a much more purposeful manner. As an example one of the participants describe how his group every year does a research of the goals of their customers and how they ­ the customers - plan to use their discretionary income (009).

Two participants complain that researchers may be too superficial and not spend enough time to understand another culture (008, 068). I have done culture studies even though I have a technical background, and I believe some of the problems may arise when people with a technical or natural science background does what essentially are social science studies. In natural science it may take a long time to figure out how to solve a problem ­ which formula to use and what to put into it, but then you have learned how to solve it. In social science it is fairly easy to feel that you understand how a question shall be answered ­ similar to finding the right formula ­ but from then on it takes a lot of effort until you understand the question and have even a tentative answer. This means that when you come with a natural science or technical background you may easily feel that you have solved a social science problem when in reality you have just started to study it.

Another aspect is that studies of users of software or electronic equipment often have to be completed faster than traditional anthropological or sociological studies. Two participants note that if research shall be used as input for a development project, the results must be available when needed in the project (001, 004). One of them describes how it might be necessary to develop methods such that a researcher who is asked a question in the morning may go out and do research in the afternoon and have an approximate answer ready the next morning (004). As one of the participants described it is necessary to be satisfizing ­ in contrast to optimizing ­ to deliver something that is good enough but not perfect (004) and to deliver it on time.

It is my impression that even academic studies of use of electronic equipment and media must be completed faster than traditional anthropological or sociological studies. In many areas the development is going so fast, that a study that takes several years to complete may be outdated soon after it is published. One example is a study of telephone use in a rural area that I completed in 2000 (Strøm 2002). When I visited the area three years later, mobile phones had become widespread and the use of telephones had changed completely.

Collaboration between disciplines

A number of participants stress the need for collaboration between disciplines (003, 004, 005. 010, 032), and one participant mentions how it is necessary with a number of different viewpoints if we want to understand a cross-disciplinary field as user experience (040) ­ a field that spans both cognitive psychology, conventional HCI and interface design and broader issues about culture, satisfaction and aesthetics.

Two participants note that it causes problems when people do not really study what already has been made in another area than their own (005), but instead invent their own vocabulary (038). A third participant notes that people tend to be grounded in one discipline and then pick one or a few items from another discipline and use them (032). I must confess that I on several occasions have been guilty of that crime. Once again, part of the reason may be my technical background. In a technical education you learn quickly to acquire an overview of a new area and only to read the minimum of theory that is necessary to solve the actual problem. In contrast, I have the impression that people from social sciences through their education have learned that they shall master most of the relevant theory before proposing a solution.

Relations between academics and practitioners

According to the interviews a number of academics have a fairly low opinion of both their colleagues and ongoing research. In particular they state that researchers tend to work on artificial laboratory-type studies, even though the results of such studies are not valid in a realistic setting (004, 035, 072). One of them notes that academics should spend a couple of years bringing out products before going back to the research lab (004). Other participants describe how practitioners find it difficult to get access to research results and use them, and a number of participants stress the need for people who can translate the research results into a format where they are accessible and usable for practitioners (004, 033, 074).

Design and technical development

One view of technological development is, that it is done by copying what appear to be successful solutions, while improving the points in the design where the largest gains are possible with the least effort (Law 1987). When one of the participants describe how he likes to search on Amazon, but also complain about the limitations of keyword searching (037), it is likely that he will demand that a library system have some of the search features of Amazon, but that it includes semantic web functions that makes it possible to search on all materials on a given topic no matter their language or format.

Another example of a gradual improvement is the "Did you mean..." function in Google mentioned by one of the participants (033). This function is fairly simple and helps users who used to get stuck because they could not spell what they were looking for. It is likely that the success of Google, Amazon and Microsoft Windows can partly be attributed to such improvements. This means that I disagree with the participant who does not want his students to know too much about the work that already has been done, because they then cannot be original (004). In contrast, I will like to immerse students in knowledge about existing solutions, how they actually are used and the deficiencies experienced by the users. Only then is it possible to avoid making the same mistakes and doing the same work all over as earlier designers.

Technology changes the needs of its users

One of the participants notes that even though some or our habits are fairly stable, new technology will affect the way we react (003). Another participant state that information technology is different from industrial technology because it is so flexible (032). However, I believe that the similarities are larger than the differences, and that we can use the experiences from earlier technologies that have changed peoples lives. At least it may be easier for us to accept changes, when we realize how previous generations have had to live through similar disruptions of their lives and habits.

One participant describes problems with proper behavior when using e-mail (072), and two participants note how technology changes our expectations to media, information and communication in general (001, 006). Other participants describe how technological development creates new demands and new challenges for system designers. Two participants mention that users are multitasking more (032), that we experience more disruptions and more often need to recall something (004). A third describe that we have to cope with more information channels with information overload and more fragmented information and that people in their everyday life only looks for information when they need it to accomplish something. (035). One consequence of these developments is that we tend to become more impatient and expect to get information that fit our needs in the exact moment (001, 010). It appears that our expectations are increasing at least as fast as our access to information.

Another issue is the design of search and retrieval systems. One participant describe how user’s feedback and collaborative filters as used in Amazon can guide users to information that in particular are useful for them (071). Another asks for systems that are more democratic, and he believes such a system would require different categories and keywords for the same concept (005). These requests seem to be very similar. However, I see a risk that both a system as Amazon’s and a system that adapt to keywords entered by users when searching may tend to promote the most widely used and accepted concepts and suppress minority views.

Breaking down divisions

One of the advantages of digital information systems is that they make it possible to combine private information with what already is published; essentially what many people have wanted to do with a ball point pen in a printed book, only now it can be done without disfiguring the public material. Users may construct their own portal (038): their own personal electronic library.

Users may combine learning and entertainment as one participant notes that children do (008), blogs may at the same time be an individualistic and a broadcast media (068) as personal web pages to some extent are today, and information may be delivered through a combination of web pages, ipods and pdas (061).

One participant suggests that course management systems and library services at a university were integrated (061), so students attending a course could get easier access to library materials related to the course. That may in particular be interesting because students may use a course as an entrance to find the most relevant materials about a specific topic.

Another participant suggested that instead of the present system of scientific publishing where an article is reviewed and send back and forth a number of times, it is possible to set up a system where temporary versions of an article are posted and everybody invited to comment on it. When these comments later are incorporated, the result may be a final article that then goes into the archive (048). We may even end up with something like the wikipedia where every article with comments is the result of a collaborative effort where participants contribute and correct each other.


A few days ago I attended a meeting with participants from Danish universities and private companies and also from different disciplines ­ in particular HCI, communication and marketing. A private company described how they had established collaboration between different professions, and we had a lively discussion between both academics and people from private companies. The general agreement was that collaboration bwtween different disciplines was valuable and fairly easy, and that the only major problem was to get everyone to accept that their own profession could not run the show. In contrast, the participants in this study give me the impression that collaboration is difficult and almost impossible. There may be a cultural difference ­ in Denmark there is a strong tradition for project work and collaboration while studying ­ or it may be that collaboration feels much easier when you are together with people from other disciplines and have time to talk with them in a friendly atmosphere.

How I wrote this essay

In order to write this essay I read about 25 interviews selected more or less randomly from the international study, and I used the search function to look for specific keywords. While reading I extracted quotes that I thought looked interesting from the interviews, and while doing that I found some recurring themes that I wanted to focus on. Finally, I sorted most of the quotes according to my themes and used the quotes as a basis for writing the essay.


Doyle, Arthur Conan (1894): Silver Blaze. In Arthur Conan Doyle, The Memoirs of Sherlock Holmes. UK

Law J. (1987). Technology and heterogeneous engineering; the case of portuguese expansion. In W.E. Bijker, T. P. Hughes & T. Pinch (eds.), The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology. US: The MIT Press

Strøm, Georg (2002): The telephone comes to a Filipino village. In James E. Katz & Mark Aakhus (eds.), Perpetual Contact. UK: Cambridge University Press