HCI is still a rather new subject for Scandinavian undergraduate level studies. Ten years ago only short courses at various computer science andsystem development educations were available. During the last two-three years HCI education has grown rather rapidly through the establishmentof minor programs at the smaller universities and with specific profiles at the bigger universities. In 1998 the first PhD in HCI graduated andnow an HMI graduate school has been started in Linköping and Stockholm. HCI has also been well established as a research activity in Sweden andis now more and more becoming regular departments at the major universities. Still, education in HCI is varied in content and direction, and thispaper describes a first Swedish workshop aimed at reviewing the contents of HCI courses in Sweden.

Introduction

A workshop at the Swedish annual conference STIMDI '98 (Swedish Interdisciplinary Interest Organization on Human Computer Interaction) gathered12 persons deeply involved in HCI education at 5 different universities in Sweden (Umeå, Stockholm, Uppsala, Linköping and Gothenburg). Thepurpose of the meeting was to get an overview of HCI education in Sweden today and to start a discussion on development and co-ordinationbenefits and improved teaching aids for the future. Initial questions could be; what is today and should be included in a basic HCI course at thedifferent universities? What is the relation between undergraduate and graduate level education? What are the needs from the industry?

Proposed discussion topics

The workshop was initiated through a session where everybody were asked about their expectation on the workshop in the forms of relevantquestions to discuss. The initial questions and discussion topics were, in order of appearance:

1. How do you create a balance between redesigning existing artifacts and promoting new and creative solutions?
2. Exchange experiences concerning different pedagogical approaches to HCI
3. New ways of examining HCI courses
4. What should a basic HCI course contain?
5. What should advanced HCI courses contain?
6. What basic knowledge should we require?
7. How can we introduce practical tasks into a HCI course?
   • Without it taking too much of the time?
   • Using simple and realistic problems?
8. How do you create motivation for the students to take HCI as a secondary subject?
9. Practical issues
   • How do we teach innovation?
   • Creativity?
10. What can we learn from the game industry?

Basic HCI courses

Table 1. The various subjects proposed for an introductory course on HCI

Subject	Vote
Design principles/process	10
Cognitive psychology	10
Interaction techniques	9
Usability	7
Task analysis	7
Evaluation methods	6
System development methods	6
Prototyping	5
Information visualization	3
New techniques	3
Interface stuff	2
CSCW	2
Method	2
HCI as a responsibility for programmers/developers	2
Web design	1
Design for all	1
Interdisciplinary	1
Emotion	1
Interaction principles	-
Hypertext	-
Linguistics	-
Real time systems	-
Graphical design	-
Complex systems	-
Social & organizational aspects	-
Ethics	-

Through a brainstorming session we attempted to condense a set of what we considered to be typical knowledge that a basic course on HCI shouldpresent. We performed an informal survey in which the workshop participants were to choose the five most important subjects and the summary ofthe votes went as shown in Table 1. The course should comprise 5 points (where 1 point corresponds to approximately one week of work). Thenumber of subjects turned out to be rather limited.

This informal survey indicates that cognitive psychology, design principles/process, interaction techniques, usability, task analysis, evaluationmethods, system development methods and prototyping could be regarded as the most important themes in a basic HCI course. In the discussions thatfollowed, there was a general agreement that it might be dangerous to go into to much detail on each of the separate topics and that theeducation should rather be guided towards looking at IT solutions as whole.

Supplementary HCI courses

We continued the session by looking at a set of potential supplementary courses which might be given to students (all based on 5 point courseblocks). The idea was to find course topics that would be:

1. Sufficient to form courses on their own
2. Central enough to defend their roles as separate courses

• Method, Evaluation methods and usability
• Design
• Functional design
• Screen presentation
• Cognitive design?
• Graphical design
• Prototyping & user centered design
• Creativity & Innovation
• Multimedia and multimodality
• Tools for UI design & Programming
• Social, moral and organizational aspects/effects of IT
• Documentation and supervision
• Pedagogical user interfaces
• "Theoretical views"
• Project administration, Co-operation

Course Literature

During the discussion on HCI courses, the issue of textbooks was raised. One intial observation is that the selection of central textbooksactually used on courses was relatively small (e.g., Preece et al. 1994, and Dix et al. 1998). Surprisingly, it seemed that the participants wasnot directly satisfied with the books. The lack of a good Swedish book was also mentioned.

The problem of selecting textbooks and reference literature seems to be a problem in many HCI educations. Also, whereas the basic books, asmentioned above, are at least potentially useful, there are very few books that can be applied on higher levels of education.

When asked for books that had been used in courses (at some point), the participants of the workshop collectively mentioned some 5 to 10 books,which were used for the basic education in HCI.

The general critique on the books was that they generally:

• are too shallow in the coverage of many topics topic
• impose a common structure on HCI, which is not necessarily the best from a pedagogical point of view
• try to cover too many topics within what is reasonable in 5 weeks

New forms of education

Since some of the issues mentioned new ways of teaching and examining HCI, a discussion grew in outbreak sessions about how new forms of HCIeducation could be developed. Since these discussions were spontaneous and unstructured, only a very small selection of the subjects covered ismentioned here.

One point that was made several times was the size of the educational groups (The participants reported experiences from 6-175 students).Generally a smaller number of students per group promotes better results, as well as a larger possibility to discuss interesting aspects of HCI.On the other hand, larger groups of students could be handled in project teams where comparative design solutions could be one way of developingthe students' awareness of alternative design solutions.

It was also considered beneficial to develop new forms for examination. The problem is to examine the student in a more holistic way, notfocussing on details. A starting point for the discussions was some experiments performed at Chalmers Institute of Technology in Gothenburg. Theexamination was performed as an oral examination on previously submitted mandatory tasks. By using this type of examination the throughput andthe frequency of examination in the organization increased substantially. In this case, it was also possible to let the students perform projectsthat lasted up to a year. One example assignment of this kind was to develop a system for handling medical records for gymnastic therapists.

As a related topic the use of problem based education in HCI was discussed in fairly large detail. This discussion also raised the question onlecturing versus supervising. Since HCI is an area, to a large extent based on understanding rather than actually learning of details, it wasconsidered interesting to increase the focus on supervising assignments, rather than using traditional lecturing in HCI education. This canprobably not be used as a general assumption, but it can be applied at least partially in most classes.

A fairly intense debate on textbooks and their quality also promoted one suggestion concerning course literature, namely to let the studentsthemselves find, analyze and use the various and often contradictory guidelines that can be found on the web as part of the course literature.The general idea in these discussions was to increase the students' awareness of that there are no simple solutions to HCI problems. One problemwith textbooks is that the students tend to regard them as the "solution" to all their problems.

Freedom in the choice of mandatory tasks also made a difference in the student's motivation. The treatment of real or realistic problems isadvantageous, such as computerization of bus stops or support for people with special needs, especially people with severe disabilities (á laStephen Hawkins).

Real user participation is also beneficiary. However, letting the students use real companies or users for their course assignments was by someof the participants considered a problem. However, at Chalmers the students were often asked to visit companies for their assignments andapparently they seemed to get a rather large number of supportive companies.

Industrial requirements on HCI Education

One special concern in HCI education is the promotion of HCI in the industry. One of the participants has practiced HCI education for industryfor several years. In this chapter his main points of view are summarized. HCI education is becoming increasingly important also from thecompanies' perspective. There is clearly a greater need for HCI experts than to promote usability increasing operations through methods andtools. Although this is also important an expert evaluation of a user interface is always more worthwhile when it comes to time and resources.The industry is also pragmatic. One can never achieve a perfect result, one should rather strive to achieve as good as possible within theavailable time. What the industry needs is both pragmatic knowledge and hands-on advice and guidelines. Therefore it is often better to do littlethan to do much (and never get anything done).

An important thing for HCI education is to train peoples ability to be sensitive for judging good and bad design, an ability that partly is thereby hart and partly can be trained. This could possibly be performed based on realistic problems and then applying HCI knowledge to be able toachieve better design. A practical design course could for example use Don Norman's book (The design of everyday things) and let the studentsapply the ideas in it to HCI.

HCI is best applied through "indoctrination" into projects, simply by showing that something works by applying it. "Learning on the job", to workwith your skills and expertise is important in education. Working in projects is one of the best ways to learn. Extra credibility can forinstance be achieved by inviting guest lecturers from companies that tell the "right" things. Usability should be a major goal in all developmentwithin a company, rather than being a goal to evaluate against in the end. Usability is not something that can be applied in the end.

Sunshine or disaster histories are useful in HCI education but difficult to acquire from the industry. There are many reasons for this:

• It can be difficult to actually tell whether a project has succeeded or failed.
• Evaluation is often performed by other than those who have done the work.
• The industry do not want to get an image of a company that failed or share a winning concept.

A good example of a small practical design task is to ask the students to design functional salt and pepper packages for airline passengers. Youcan arrive at several conceptual design solutions within 15 minutes and discussing the different design solutions is a good practical way to doparallel design. It is very useful for students to discuss each other's ideas. People could arrive at very different design solutions reflectingthe existing design space; e.g., color-coding, different granularity, transparent packages. Training the student's ability to figure aroundeveryday designs and their benefits and drawbacks is useful. Why can the specification of what is on the different floors only be read outsidethe elevator?

How can one teach or learn innovation? Is it even possible? Brainstorming sessions can be very useful to be able to achieve creativity.Observation is very useful to learn benefits/drawbacks with different design solutions. Showing system developers how the users actually used thesystem repeatedly actually made them understand and want to correct their designs rather than addressing an evaluation report.

The people that today are educated in HCI so far seem to have turned up as teachers, due to the heavy increase in the subject. Those who at lastreach the market usually ends up as consultants. There is a need for HCI education for people who finally end up in supervisory or strategicdecision making situations. People who study law, economics and business applications might eventually end up in supervisory positions in acompany.

We must promote an increase in HCI education at all levels. An English survey concluded that 6 % of the Englishmen physically abuse theircomputers and 2/3 of the Englishmen do not know how to program their VCR.

Discussion

Prior to the workshop an informal survey was performed in "ASKEN" the web service that support Swedish students with search possibilities oncourses and education programs that exist in Sweden. This survey shows that a large number and a great variety of different courses are offeredtoday, many of them not at all known by the participants. You could take HCI for 3 weeks or for one year. HCI courses were hidden underneathnon-descriptive names and in combination with very many different courses. Evidently there is a need for a Scandinavian curriculum for HCIeducation.

In addition to this, if you analyze job openings in the newspaper you'll see that HCI experts are sought under titles, such as usabilitymanagers, human factors specialists, IT-strategist, ergonomist, etc, lacking a commonly understood concept for the role needed. A commonly agreedHCI education curriculum could make it much easier to market the subject of HCI to the industry.

Conclusions

This workshop concluded that there is a need for (among many other things):

• a more thorough survey on the courses taught at the universities in Sweden.
• a forum for exchanging educational ideas in Sweden.
• an educational data bank where educational material, tips and text book evaluations (e.g.) can be found.

• to be supportive in need of inspiration
• to be a source of guest lecturers
• offer mentorship for new teachers

Acknowledgements

The workshop participants are all acknowledged for their input on the workshop. STIMDI and the University of Umeå are acknowledged for hostingthe workshop.

About the Authors

Jan Gulliksen, Ph.D. in Systems Analysis performs research on user centered design in several applied projects (e.g. with the Swedish NationalTax Board) at Uppsala University. Jan is also supervising the working group on user orientation at the Center for user oriented IT-design (CID),Royal Institute of Technology (KTH), Stockholm. Jan has been teaching HCI and Systems analysis for the last ten years

Lars Oestreicher, Ph.L., is a university lecturer at the department of Information Science, Uppsala University, and is working on his PhD at theHuman Machine Interaction Graduate School, Stockholm. He is the Swedish representative for the IFIP Technical Committee No 13 on Human-ComputerInteraction. Lars has been teaching on courses on Human-Computer Interaction, Software Engineering Cognitive Psychology for software engineers,Usability Engineering and Task Analysis since 1987.

Author's Address

Jan Gulliksen
Dept. of HCI, Uppsala University
Lägerhyddsvägen 18
SE-752 37 Uppsala
Sweden
email: Jan.Gulliksen@hci.uu.se
Tel: +46-18.471 2849

Lars Oestreicher
Unit of Computing Science
Dept. of Information Science, Uppsala University
P.O. Box 311
SE-751 05 Uppsala
Sweden
email: larsoe@csd.uu.se
Tel: +46-18.471 1019

References

Dix, A., Finlay, J., Abowd, G., & R., Beale (1998). Human-Computer Interaction (2^nd ed.). London, England: Prentice Hall.

Preece, J., Rogers, Y., Sharp, H., Benyon, D., Holland, S., & Carey, T. (1994). Human-Computer Interaction. Wokingham, England:Addison-Wesley.