Full Paper

Irina Verenikina

irina@uow.edu.au

Russian Academy of Sciences

Edward Gould

e_gould@uow.edu.au

Department of Business Systems

University of Wollongong

The role of psychology in fields such as human computer interaction (HCI) has long been recognised but the practice of applying it to meaningful design techniques has been a frustrating undertaking at best. Given that HCI is a study of the relationship between humans and computing machinery it would seem that a fundamental basis of HCI design should be psychology. Despite this, evidence exists (Card, Moran and Newell, 1983; Polson and Kieras, 1985) that even when sophisticated attempts have been made to extend cognitive psychology in the direction of HCI they have had no effect on design systems. Alternatively, some of the most successful user interface design work of the past twenty-five years has made no explicit use of psychology (Carroll, 1991). Landauer (1991) has also criticised the inadequacy of cognitive science, the ëwesterní branch of psychology, when it comes to providing a blueprint for HCI design. He says that in spite of 100 years of investigation into human behaviour ënothing remotely resembling what one would hope for as a basis for HCI, nothing with substantial generality, power, and detail at the required level of cognition has ever materialisedí. In a study of British software houses Bellotti (1988) found that designers were not using any of the methods recommended by HCI research and in fact, most were unaware of their existence. Other criticism can be found in Suchman (1987), Ehn (1988), Bannon (1991), Bodker (1991), Kuutti (1992), Draper (1992), Kaptelinin (1992). Naturally, in the face of such frustration and criticism, researchers began looking for alternatives to the traditional Western cognitive psychology as a possible basis for the study of such areas as computer mediated communication, computer supported cooperative work, computer interface design, information systems analysis, computed supported decision systems and so on.

This paper elaborates an approach to interface design based on the Russian developed Activity Theory which provides a more complete analysis of human nature and which avoids the problems inherent in the view of humans as exclusively information processors.

Attempts to apply the traditional cognitive psychology have had limited success due to the narrow focus of this discipline. It has concerned itself more with the cognitive process and projected this into a philosophy of human-computer interaction. The reason for this is the compatibility between this branch of psychology and the computer both in concept and vocabulary. There are compatible input devices such as the eyes and the keyboard, compatible storage devices such as the brain and computer memory and compatible output devices such as the fingers and VDU. This has led to the concept of the equality of the computer and its human operator. The logical extension of this is that ultimately it would be possible to build a computer with the ability to outstrip the human mind. Fields dedicated to this quest such as the study of artificial intelligence and knowledge based systems have sprung up and have had limited success most recently in their ability to produce a machine capable of humiliating the world chess grandmaster. Philosophically the dreams of the technophile were realised in this chess contest, the role of the machine has been reversed from that of a tool which enhances life to that of a machine which seeks to dominate it. The practical outcome of this general approach is that any experiment to examine the compatibility of a computer with a human is restricted to an interaction between these two components alone to the exclusion of the context of the real life interaction of human work activity.

This general approach has led HCI experimenters to a frustrating lack of design tools suitable for predicting human reaction to information technology. A branch of psychology developed by Lev Vygotsky in the beginning of this century provides a different philosophical approach and opens new perspectives for constructing experimental research into this problem. Later the approach was elaborated by A.N.Leontiev and others into a structure known as activity theory.

Vygotskian psychology and activity theory take a much broader view of human psyche as being the product of cultural and social forces, we are what we are precisely because we interact with the society in which we live (Vygotsky, 1978). We are not just a collection of cognitive processes but a complex interaction of activities, motives, actions, goals, operations and tasks in a hierarchical arrangement all directed to the process of living. The hierarchical structure of activity is one of its most distinguishing characteristics (Leontiev, 1978). In this process we make use of tools of which the computer is just one of a long line stretching from stone knives, through the wheel to modern technologies such as the aeroplane. In order to understand how tools have affected human development it is necessary to study the methods by which tools have extended our ability to achieve goals which satisfy some needs. Physical tools extend our physical abilities (shovel, saw, pen) while psychological tools such as language, mathematics and other symbolic systems extend our mental abilities. The computer is confusing because it appears to exhibit peculiarities of both a physical tool directed at external activity and a psychological tool directed at internal activity. Activity theory examines the relationship between internal and external activity. It treats the two types of tools similarly, both as means of satisfying real needs and achieving corresponding goals. This leads us to a different experimental scheme, one with many more elements which emphasise the role of the computer as a tool embedded in human activity, both mental and physical.

Kaptelinin (1992) has identified four points which are of central importance to activity theory and the special role that tools in general play within this framework.

• the identification of the specific psychological functions of computer tools and how they differ from other kinds of artefacts,
• consideration of the psychological nature of the functionally equivalent processes involved. More specifically whether they are activities, actions or operations since this can affect the design and training strategies,
• an analysis of the development of users from novice to expert as they develop new skills and abilities with experience,
• the relationship between computer tools and cultural influences which can render successful systems in one social setting ineffective in another

The diverse nature of the many current techniques used in HCI design and evaluation makes the job of classifying and identifying them as models, methods or architectures difficult. Over the last fifteen years a large number HCI design tools and techniques have been proposed. Howes (1994) has identified twenty three different approaches, each oriented towards a small and manageable part of the overall issue of the way humans interact with computers.

Activity theory provides a paradigm for the description and understanding of the way humans interact with computers within the context of the user's environment. This makes it ideally suited to the study of human-computer interaction. It stresses the vertical integration of psychological analysis and stresses the following points usually missed by the cognitive approach:

• the goal which is determined by the nature of human actions, their structure and performance,
• the role of mechanisms underlying the integration of a computer tool into the structure of activity,
• the importance of developmental aspects of human computer interaction,
• the significance of social interactions surrounding computer use (Kaptelinin, 1992).

We describe two practical examples of the application of this approach. One is to designing an interface for demographers to assist in the onerous task of school population modeling and the other to executive information systems interface design.

A population prediction model has been developed (Gould, 1993) to project school population figures for a period of five years into the future. The model is part of an information system for educational administration which can be used for both middle management decision making and strategic planning by executives.

The aim was to design an interactive visual interface for this model to cater for its unique characteristics of population projection in small geographical areas and to compare amalgamated figures with projections at a national level. The overwhelming mass of relevant data available simultaneously makes it difficult for the demographer to see patterns of change within the area of interest when output results of the mathematical models are presented in the form of tables.

When describing the activity of demographers we can see that the mathematical models are embedded in the activity as a computer tool on an operational level. The next step was to find computer support on a level of actions.

We have assumed in this case that different saturations of coloured sections on a map representing different levels of school populations in the geographical areas would be the proper tool. Thus, the intuition needed by experienced demographers to devise the correct proportion of population is supplemented by materialised computer support. It enables demographers to make the necessary comparisons between local and national figures without the problems associated with using tables of figures. Incorporated into this design is the ability to use the saturation of colour in each section of the map to vary the population and thus reach an equilibrium point where local projected populations are in "balance" with the more accurate national figures. The aim is to mediate the users motive of predicting the population with the goal of achieving the balance as well as reinforcing motivation by putting an element of play into the design.

As an integral part of the design process an experiment was set up to evaluate the association between the commonsense notions of increasing, decreasing and stable areas of population as well as the role of luminance as it related to the two most popular colours. Varying the colour saturation can have a profound effect on the users perception of actions needed to achieve the goal which is to balance the varying totals of the smaller areas according to colour saturation without exceeding the fixed total.

The development of executive information systems (EIS) present particular problems for the designer. Not only are the clients for whom the system is being designed not known, but are of a disparate group with different ways of doing business, making decisions, using data, arriving at conclusions, etc. For an executive to make use of an information system to improve management efficiency an interface compatible with their management style must be developed.

Software developers have been producing systems for many years which purported to be executive information systems. Unfortunately very few of these systems have been successful. The problem cited in the past has always been that the spreadsheet software on which the systems were usually based was not sufficiently sophisticated to deal with the problem. This, however, is no longer the case. Very sophisticated multidimensional databases are now capable of extracting complex combinations of the most obscure of data and presenting it in every imaginable way, all at blinding speed. The problem has changed from one of software engineering to one relating to the psychology of executive users. It has long been known that executives manage in their own particular style. It is also well known that these different styles need different types of information presented in different ways. The question is how do we go about finding out the best mesh of information and business management style so that an EIS is useful to an individual executive.

Existing research which studied personalities and tried to apply it to interface design were unsuccessful because of the difficulty in applying psychological studies of personality to this particular case - they were too far removed from the problem. We suggest an approach based on the study of a personal profile of an executive. This profile will consist of an ensemble of the executives business activities relating to the use of the EIS as a tool embedded in the activities. This will allow designers to devise ways that an interface can be designed to extract, amalgamate and present information in a way that suits the business management style of the executive. The question is: what type psychological profile is needed to find this out? What lessons are there for the designers of multidimensional database shells to allow for easy incorporation of an individual style in the design of an interface.

We are now conducting a case study of the application of this idea to the creation of the first prototype of an EIS for a head of a department at a university using a university finance system and its derived faculty/department budgets as an example. This is an example of one particular activity of a department head in a university, that of financial planning. Others could relate to student information or subject information.

The head of each department is responsible for a certain amount of money which is allocated to them at the beginning of each year. In order to spend this money wisely a budget is prepared with shows how much is to be spent on certain items such as salaries, conference travel, purchase of computers and so on. Each department head needs to be able to keep track of this money so that at the end of the year it is neither overspent or underspent. The money allocated is not physical money but exists electronically and by looking at the universityís finance system each head can see what money has been spent and how much is left of the amount allocated. The problem is that unless a good interface has been installed to filter the massive amount of data it is almost impossible to see trends or even identify areas which are likely to be potential problems such as ëblow outsí some area which is not being spent due to a new policy.

An aim of the research is to create an individual business profile of the head of the department, analysing those work activities which relate to the financial information system as a tool for performing the activities. The analysis of the structure of the activities is based on an Activity Check List created recently by Kaptelinin and Nardi (1997). It should be noted that the list is a good example of operationalisation of activity theory concepts for practical use. The next step is to find the relevant structure and organisation for a prototype the personal interface design. The long term aim is to develop a questionnaire for identifying personal executives business profiles which would serve as a guide for executives and/or their technical assistants for developing a personal interface design for an executive information system.

Activity theory is becoming more widely recognised as a promising alternative to HCI problems particularly now that a new book ìContext and Consciousness: Activity Theory and Human Computer Interactionî edited by Bonnie Nardi appeared last year (1996). It suggests that the activity theory approach may allow us to move HCI research out of the laboratory and into the field and may also be used as a means of guiding field studies.

(c) Irina Verenikina and Edward Gould

The author(s) assign to ASCILITE and educational and non-profit institutions a non-exclusive licence to use this document for personal use and in courses of instruction provided that the article is used in full and this copyright statement is reproduced. The author(s) also grant a non-exclusive licence to ASCILITE to publish this document in full on the World Wide Web and on CD-ROM and in printed form with the ASCILITE 97 conference papers, and for the documents to be published on mirrors on the World Wide Web. Any other usage is prohibited without the express permission of the authors.