Editorials

back  To Overview of Editorials

Editorials 2009

Editorials 2008

Editorials 2007

Editorials 2006

Editorials 2005

Editorials 2004

Editorials 2003

Editorials 2002

Editorials 2001

Editorials 2000

 

The Times They Are A-Changin' – So Are The HCI Paradigms

By Gerd Waloszek, SAP AG, SAP User Experience – July 24, 2003

Human Computer Interaction (HCI) and its "parent" cognitive psychology have long been dominated by the information-processing paradigm. Currently, however, there are signs that we are experiencing a shift to new HCI paradigms. This editorial offers a glimpse into the future of HCI.

 

Looking Back...

About ten to twenty years ago, research and thinking about human cognition was dominated by the information-processing paradigm. This paradigm borrowed many of its concepts from computer science, especially from artificial intelligence (AI). It regarded the human brain as a symbol manipulating system in the spirit of digital computers, having a central processor and different memory systems, such as short-term and long-term memory. There was no place for emotions in such a system – or if there was, then even emotions were seen as processes that manipulate symbols.

On the empirical side, this theoretical orientation was reflected by the dominance of controlled experiments: Hypotheses were stated, variables manipulated, and the resulting behavioral changes measured. The results were then statistically evaluated, allowing scientists to extend to entire populations the results of experiments carried out on individuals. All in all, and despite fundamental differences, cognitive psychology tried to imitate the natural sciences, not the social sciences.

Model of the human brain

Figure 1: Model of the human brain inspired by the digital computer (from R. L. Solso, Cognitive Psychology, modified; click image for larger version)

Due to the fact that HCI inherited a large part of its theoretical foundation from cognitive psychology (or cognitive science), the information-processing approach and the preference for controlled experiments also dominated this field for a long time. Here the most influential book was Card, Moran, and Newell's "The Psychology of Human-Computer Interaction," which presented the GOMS model (goals, operators, methods, and strategies) and the Human Information Processor. This approach was similar to that of production systems, that is, collections of IF-THEN rules, which were used to simulate human cognition and behavior. The GOMS model, for example, allows scientists to predict error-free human performance in a wide range of computer-based tasks. More recent paradigms, such as neural networks, that literally overran cognitive psychology in the 1980s and 1990s, did not have much impact on HCI. The models and methods that they offered were not as useful and practical as those provided by the information-processing approach.

Human information processor

Figure 2: The Human Information Processor by Card, Moran and Newell (1983)

Nevertheless, there were always researchers who pointed to the limitations of the information-processing approach and its inability to deal with more complex, or social systems. Critics pointed out that lab experiments are not embedded in real work contexts, and are therefore of limited use. Field studies, however, were discredited for a long time due to the fact that they are not controlled, and thus allow the use of descriptive evaluation methods only.

The recent spread of "guruism" in the field is probably a result of the deficiencies of the above theoretical approaches. Both practitioners and the more theoretically-oriented experts make comments, such as "users do not want to scroll," or "users do not want to look stupid," or "base your design on user data, not on assumptions." I do not want list names here – most readers will be familiar with at least a few of the gurus. Sometimes, I feel I've been transported back to the psychology school wars of the 1920s, when people cited a guru to carry their point in a discussion.

But "the times they are a-changin'." While mainstream HCI research may still be carried out in the spirit of the information-processing paradigm, new theoretical, empirical, and methodological approaches are emerging and making themselves known at HCI conferences. And some of the gurus contributed to this change.

 

What's Coming?

In this editorial, I will be fairly brief about what's in the store for the field of HCI and will provide only an overview of the trends and paradigm shifts that I have recently noticed. I expect articles on the new trends to appear in the SAP Design Guild in the near future. There is already a CHI 2003 report that comments on some of the new trends.

Ethnography, Cultural Differences

At the CHI 2003 conference, some researchers cited ethnography as a "new HCI paradigm." Whether this statement actually reflects reality, remains to be seen. At present, there is notable interest in cross-cultural issues, comparative studies, and field studies that observe users within their work context. This is one of several focusses spurred on by companies that supply global markets and need to adapt their products to specific cultural or social groups.

Design (More Important than Engineering)

For a couple of years now, there has been a trend in HCI away from the computer science community toward the design community. Such a move, of course, has consequences for the methods and self-image of the usability profession as a whole. Whether the design advocates who claim that "design is more important than engineering" are right, remains an open question. But as the next trend illustrates, this is the direction in which we are moving.

Emotions, Affect, Pleasurable Products

The interest in emotions, or rather, the relationship between emotions and products, is, amongst others, nourished by the belief that the software industry has reached a certain level of maturity: Customers today take functionality for granted. It is on the emotional side that products must qualify themselves: They must be both desirable and pleasurable.

New theories of emotion, such as Don Norman's, abandon the symbol processing view. Norman emphasizes the role of the chemical system and the influence of emotions on thinking and behavior (Norman manages to reverse the traditional roles of thinking and emotions). Even AI pioneer Marvin Minsky has modified his "Society of Mind" theory to include emotions – he is currently working on a new book, "The Emotion Machine."

Commercialization, Business Models

Business issues typically play a more important role for mature products and industries than they do for younger products and industries. The Web is also a major factor that calls for new business models.

Patterns

There is currently much interest in user interface patterns. The widespread use of patterns streamlines design and makes development much more efficient: Patterns would allow a more "industry-like" software development process.

Communities, Collaboration, Social Systems, E-Learning, Media (Web)

While personal computers lead to a "singular" view, networked computers, especially since Internet use has become so widespread, call for a focus on communities and collaborative activities. This change also requires new theoretical approaches that allow experts to deal with the related issues in a systematic manner.

Distributed Cognition and Other New Theoretical Approaches

And finally, there are new theoretical approaches, such as distributed cognition, or activity theory, which shift the focus of the research perspective from single users to collaborative systems, including both humans and technical artifacts that help users to accomplish their tasks. In the attached references you find a paper by Hollan, Hutchins, and Kirsch that provides an overview of the theory of distributed cognition.

On the more practical side, "Contextual Inquiry" by Karen Holtzblatt and Hugh Beyer also incorporates work context into design considerations and observes users in real-world environments – perhaps even helping to pioneer this approach. It is hardly surprising that Karen Holtzblatt has a background in social sciences and applied psychology. At CHI 2002, this methodology was named as one of the most influential current HCI theories.

 

Conclusion

It is always easier to report on paradigm shift with the advantage of hindsight, as trends can be fragile and are certainly subject to change. Only time will tell whether I have spotted the trends that will survive.

 

References

Hugh Beyer & Karen Holtzblatt (1997). Contextual Design: A Customer-Centered Approach to Systems Designs. Academic Press.

Stuart Card, Thomas Moran, & Alan Newell (1983). The Psychology of Human-Computer Interaction. Lawrence Erlbaum.

James Hollan, Edwin Hutchins, & David Kirsch (2000). Distributed Cognition: Toward a New Foundation for Human-Computer Interaction Research. ACM Transactions on Computer-Human Interaction, 7(2), pp. 174-196.

Marvin Minsky (1988). The Society of Mind. Touchstone Books.
Marvin Minsky's homepage at MIT: web.media.mit.edu/%7Eminsky
There you will find a link to the draft version of his forthcoming book, "The Emotion Machine."

Don Norman's homepage: www.jnd.org
Don Norman: Emotion and Affect (Ubiquity interview)

 

top top