HCI Bibliography Home | HCI Conferences | OZCHI Archive | Detailed Records | RefWorks | EndNote | Hide Abstracts
OZCHI Tables of Contents: 919293949596980102030405060708091011121314

Proceedings of OZCHI'92, the CHISIG Annual Conference on Human-Computer Interaction

Fullname:Proceedings of OZCHI'92, the CHISIG Annual Conference
Note:Interface Technology: Advancing Human-Computer Communications
Editors:Michael J. Rees; Renato Iannella
Location:Bond University, Gold Coast, Australia
Dates:1992-Nov-26 to 1992-Nov-27
Publisher:Computer-Human Interaction Special Interest Group Ergonomics Society of Australia, Inc. Canberra Business Centre Bradfield Street Downer, ACT 2602 AUSTRALIA
Standard No:ISBN 0-9595349-9-7; hcibib: OZCHI92
Links:Conference Series Home Page | Scanned Proceedings PDF 29 Mb
  1. Opening Keynote
  2. User Interface Design
  3. User Interface Case Studies
  4. Groupware
  5. Education
  6. Closing Keynote
  7. Poster Abstracts

Opening Keynote

A Vision of Universal Functionality for Tomorrow's User Interfaces BIBAKHTML 1-14
  Gary Perlman
In this paper, I describe universal functions -- functions that can be used in all applications -- and how they can be integrated in existing programs at the level of a common user interface. I consider reasons why some universal functions (e.g., search and sort) have failed to gain the same acceptance as others (e.g., print and copy). I describe a function library, the Semi-Structured Toolkit (SST), which provides universal functions based on abstractions for storage format- and data type-dependencies of semi-structured/frame-based information units. I explain how the SST overcomes obstacles to user acceptance of information technology, and end with a call for research on user interfaces for universal functions.
Keywords: Information system functionality, Application frameworks

User Interface Design

Personal Productivity Tools as User Interface Type Managers BIBAK 15-21
  Robert M. Colomb; Yongyuth Chongvilaiwan
We argue that an economical way to build high-quality user interfaces is to take advantage of the fact that most users interact with a computing environment consisting of a workstation equipped with personal productivity tools such as wordprocessors and spreadsheets. Our approach is to explicitly use these tools in the design of user interfaces. We describe a case study giving remote access to a database from a Macintosh computer. A critical component in this class of system is a dialog manager, which was implemented using Hypercard.
Keywords: Client-server, User interface management systems
Notation Systems in Software Engineering BIBA 22-28
  E. Chang; Douglas D. Grant
"Good design will lead to improved software quality and reduce the cost of producing software." [3]
   "The design process involves the exchange of information among a variety of people including customers, users, ..., the design team. One key to information exchange is the design notation used." [13]
   In this paper our concern is with the involvement of users in the development of computer systems, particularly in the design phase. We have studied the techniques which form an integral part of over 20 popular software development methodologies. In a significant number of these methodologies, it is claimed that notations incorporated may be of value in describing a system to its users. From the user point of view, we have explored the understandability and suitability of the representation techniques. Our studies make it clear that the claim of utility of notations for user understanding is spurious. Accordingly their use in communicating with users is likely to compromise the design process.
   We propose a differentiation between the visible part of a system presentation (that which is of relevance to users) and the invisible part of a system representation (which is primarily of relevance to developers) which is supported by system development methodologies.
   Finally we indicate the directions of our research into the establishment of notation systems for the visible part which should respect users' cognition, and therefore achieve communication requirements.
Gizmos: An OO UI Toolkit in Prolog BIBAK 29-35
  Neil Leslie; James Lothian
We describe the design and implementation of an object-oriented extension to the Prolog language for the construction of graphical user-interface components based on this extension and X-windows. The items in the toolkit are called 'Gizmos', by analogy with the 'widget' sets that are provided to interface C to X. General issues relating to the usefulness of an object-oriented extension to Prolog are discussed. A simple application built using these Gizmos is presented and evaluated.
Keywords: Prolog, Object-oriented programming, Graphical user interfaces, X-windows
Usability Testing of a Scheduling System: How Different Usability Testing Methods Support Redesign of the User Interface BIBAK 36-43
  Jukka Rantanen
Intelligent scheduling systems are decision support tools that assist in interactive scheduling of manufacturing plans. BHPR Research - Newcastle Laboratories and BHP Steel have been developing interactive scheduling assistants. The latest and largest research and development effort is the Steelmaking and Casting Scheduling System (SACSS) for BHP LPD.
   SACSS has quite a complex user interface, because of the extensive functionality of the system. Therefore it was necessary to put effort into making the user interface of the SACSS easy and effective to use. In April 1992, a usability test was carried out that aimed at
  • - identifying the main usability and performance problems of the user interface
  • - identifying the priorities of these problems to be fixed, and
  • - eliciting from users and designers suggestions for improvements. In the usability test, data was collected by video recording sessions, in which users did their typical scheduling tasks while "thinking aloud" what they were doing or trying to do. After the session they answered a questionnaire on the quality of the user interface. Moreover, one user annotated the screen dumps with problems and suggestions for improvements. Another follow-up study was made in September 1992 in order to see how users' reactions had changed.
       The results showed that users' reactions to and acceptance of the system was slightly positive. Several usability problems were identified by the different usability testing methods. This paper will consider how cost-effective the different usability testing methods were in identifying serious and minor usability problems that should be fixed by redesign of the user interface.
    Keywords: Usability testing, Cost-effectiveness of usability testing methods
  • Designing at a Distance: Experiments in Remote-Synchronous Design BIBAK 44-53
      Stephen A. R. Scrivener; Sean Clark; David Harris; Michael Smyth; Todd Rockoff
    In this paper we describe two studies of work between geographically separated designers in which verbal communication was mediated by an audio-link and visual interaction via a computer-based shared drawing system. The first study took place within a controlled laboratory setting, the second in more 'realistic' (i.e. less controlled) circumstances. Contrary to expectation, the users' opinion of the system did not differ significantly between conditions, although the latter was more impoverished. It is hypothesised that this lack of difference can be explained in terms of the extent to which users perceived the system as enabling or disabling scope for collaborative work.
    Keywords: Computer supported cooperative work (CSCW), Shared drawing surfaces, Product design, Integrated services digital networks (ISDN)
    User Interface Design and Multimedia BIBAK 54-61
      Sarah Bloomer; Fiona Ingram
    Much can be learned from user interfaces outside mainstream computer programs, such as interactive television or art, public kiosks, and games. This paper presents some innovative ideas emerging from a new breed of interface designers. These designers are creating multimedia and graphical applications with little or no reference to traditional application design. HCI practitioners can make use of the ideas emerging from these designs by exploring well-known issues from an alternative context. We'll look at issues such as alternative input and output devices, artists' approach to screen design and applications that use no screen at all.
    Keywords: User interface design, Multimedia, Graphic design
    How to Evaluate Prototyping Tools (Allowing Designers to Design) BIBAK 62-69
      Penny Collings; James Mackay
    For prototyping to be an effective part of the specification and design phase of a system's development, a good prototyping tool is needed. In this paper we define a method for evaluating prototyping tools particularly from a designer's point of view. The results could also be used to specify how to develop a prototyping tool.
    Keywords: Rapid prototyping, User interface design, Design process, Requirements specification, Evaluation method
    Supporting Up-Stream User Interface Design Activities BIBAK 70-77
      K. Mouzakis; S. Howard
    The early phases of user interface development (UID) have long been claimed to be the source of a significant proportion of the user dissatisfaction with interactive computer systems. This paper explores the utility of various technologies for supporting early phase design behaviour. From the literature and observational studles of a large scale UID project, an attempt ls made to distil some important features of these early stages of UID and then to explore their relationship with the features offered by advanced interactive technology. A conceptual architecture for supporting up-stream UID activities is proposed.
    Keywords: User interface development (UID), Task analysis, UIMS

    User Interface Case Studies

    Constructing New Front-Ends to Existing Software: Re-Use and Integration BIBAK 78-84
      Ernest Edmonds; Bryan Murray; Nick Rousseau
    Constructing bespoke software systems, or systems that address niche markets, increasingly involves the integration and packaging of existing systems. In particular, a frequent situation is where there exist one or more software applications, with proven track records, and with considerable potential to supporting a given user activity. One particular concern identified is that of increasing the extent to which such systems are user- and task-appropriate. The availability of many valuable resources via modern communications technology dictate that the integration must take place close to the user, rather than at some notional centre of the system. This paper briefly describes a coherent, substantial and validated technology package that employs user interface software to achieve this integration. The tools and methods will enable the efficient development of user- and task-appropriate IT systems in a wide variety of contexts. Case studies are presented that discuss the use of the technology in industrial settings.
    Keywords: Re-use, Integration, User interface, Separability, Front end, Knowledge-based
    A Visually Programmed Modeller for Workplace Design BIBAK 85-92
      Douglas Seeley; Riccardo Macri
    A discrete simulation system called SimView has been developed with a finely-tuned, direct manipulation interface, and a simple-to-learn but powerful modelling paradigm. As a workflow modeller and animator, it can be effectively used for systems planning, and the on-going design of productive and quality workplaces. Users can rapidly build and edit models in a totally visual manner. Its icon-based animation system is easily tailored, and it can be used as a training tool for workplace operations and for developing systems thinking skills. It can have a significant impact in supporting the emergence of more participative and systems-wise workplaces.
    Keywords: Visual programming, Direct manipulation, Participative design, Animation
    Grue: A Graphical User Environment BIBAK 93-101
      Andrew J. Herbert; Leslie M. Goldschlager
    The design and current state of implementation of Grue, a document and copy-oriented user interface system, is presented. The goal of Grue is to provide an environment that is easy to use for end-users, application programmers and designers. This is pursued by shifting functionality that traditionally resides in applications over to Grue itself, and by encouraging individual applications or 'gadgets', to be small and modular -- complex gadgets can be built by combining simpler ones. Grue is based on a prototype-instance object model which encourages customisation and experimentation by not enforcing a formal class hierarchy.
    Keywords: User interface management, Interactive design
    The Pictoid Package -- Extending HyperCard BIBAK 102-109
      Nigel Perry
    The Pictoid Package (Perry, 1992a) is a tool kit for stack designers consisting of scripts, fonts, and external commands. There are two main components in the package, Sticky Buttons and Pictoids. Sticky Buttons are "buttons" which may be inserted into the text of fields, and if the field is scrolled will move with the text. Pictoids are windoids which may be used to provide any-shape buttons, any-shape picture & QuickTime movie display, any-shaped buttons containing pictures or QuickTime movies, etc. This paper describes the implementation and use of Pictoids. A companion paper describes Sticky Buttons (Perry, 1992b).
    Keywords: HyperCard, Tool kit, Any-shape buttons, Pictures, QuickTime
    Directory Systems Interfaces BIBAK 110-117
      Renato Iannella
    Directory systems, as defined by CCITT's X.500 protocol, require the developer to define the user interface. This paper will give a brief overview of the X.500 Directory Systems and application areas in which it will used such as in addressing for Message Handling Systems.
       A current review of interfaces to the Directory Systems will also be presented. These interfaces use varying techniques from a simplistic flat textual view to graphical tree structures that mimic the X.500 structure.
    Keywords: Graphical user interface, X.500 directory services
    Developing FormsDesigner: A New Interface Technology for Advanced Human Computer Communication BIBAK 118-125
      David Sless
    FormsDesigner, a new type of 'smart graphic' software for the Apple Macintosh, developed in Australia by the CRIA National Centre for Human Computer Interface Design, and released in May 1992 to widespread critical acclaim.
       This paper gives an account of the range of skills used to develop FormsDesigner; how we created and tested the user interface, dialog boxes, menus, manuals, quick reference card, and on line context sensitive help system.
       Our experience suggests why leading edge HCI research is unlikely to appear first in journals or at conferences but rather in commercial products.
    Keywords: Interface, Intelligent, Forms, Graphic communication


    Computer-Supported Cooperative Work Design Reviews: Face-to-Face Groupwork Factors and a Groupwork Example BIBAK 126-133
      Mark Andrew
    Groupwork makes sense quickly; consider the potential of unifying individual expertise and insights and applying them to everyday problems. When computing support assists with the elicitation, organisation and prioritisation of information, groupwork can become popular. Although group facilitation seeks to explicate design values, over-enthusiasm within groups can result in non-adherence to key principles, and the resultant highjacking of the process. This paper recognises and discusses selected face-to-face groupwork factors, and reviews an exercise conducted by ALARA where simple computing support provided clear benefits.
    Keywords: Groupwork, CSCW, Value analysis
    Designing Computer Systems to Support Collaborating Authors BIBAK 134-141
      Steve Jones
    Authors who collaborate via a computer are faced with a variety of problems concerning both the processes of writing and collaboration. This paper is concerned with reducing the scale and number of those problems. It discusses the applicability of models of writing to the design of collaborative authoring tools, and highlights the difficulties authors currently face with the use of such tools. New principles for the design of groupware are proposed and a novel collaborative authoring tool, MILO, is described. Conclusions are drawn about the development of such tools.
    Keywords: Computer supported collaborative work (CSCW), Computers and writing, Collaborative writing
    Automatic Conversational Context: Avoiding Dependency on User Effort in Groupware BIBAK 142-149
      Andy Cockburn; Harold Thimbleby
    Relating individual messages to their on-going conversations enhances the value of electronic mail as a medium for collaborative and coordinated work. Some groupware systems have offered these facilities, but their ability to determine conversational context is dependent on explicit user actions -- being told -- and the use of specific systems by all users involved.
       This paper describes Mona, an email system that provides an automatic hypertext representation of conversational context. Mona is novel in that conversation facilities are provided without requiring any user effort or the use of particular systems by other collaborators. This lack of requirements and independence is made possible by inferring conversational context with heuristics from information inherent in all email communications.
       Mona's heuristics are described, together with its central design motivation: that the cost/benefit disparity resulting from dependency on user actions is liable to cause system rejection.
    Keywords: Email, Free guidance, System uptake, Conversational context, Heuristics, Mona (a CSCW system)
    Extending Electronic Mail for Coordination in Groupware BIBAK 150-157
      Martin Leadbeater; Douglas Seeley
    This paper discusses the extension of electronic mail programs to support coordination of actions between participants by including features for tracking key conversational events. Users of the program still use messages as the basis for their communication. However, they are encouraged to tag messages appropriately when they alter the stage of a conversation. This makes conversations and relationships clear to themselves and to recipients. Personal message agents record the flow of these tagged messages, providing a small database of conversations in progress. At any time, users can inquire of this database, viewing conversations and outstanding requests and commitments. This enables easy access to "outstanding work" lists and opens up applications to distributed calendars and other office organisational tools.
    Keywords: CSCW, Coordination, E-mail, Conversations for action, Emergence


    Computer-Based Education: Does On-Line Help Help? BIBA 158-165
      Bernie Winter
    Computer-based education (CBE) is seen as an attractive alternative to traditional teaching formats. Commonly, CBE effectiveness is measured in terms of trainee attitude, achievement and knowledge retention.
       The model presented in this study proposes that the effectiveness of CBE learning can be related to how effectively trainee responses are reinforced through the computer-based session. Accordingly, the study examines how trainees behave in accessing and using on-line help provided in the treatment CBE material.
       Main findings largely support the model proposed. Significant differences are recorded in the frequency that trainees initiate on-line help, determined by whether trainee responses are correct, or in error. Also, trainees responding incorrectly to CBE material, initiate on-line help significantly faster than trainees who respond correctly. Results indicate a significant effect of previous computing experience on help access times.
    Usefulness of Online Help to Novice Users in a Supportive Learning Environment BIBAK 166-173
      Charmaine Ryan
    Research has indicated that computer users, particularly novice users, can find information about the commands of an application software package more quickly using printed documentation or with the assistance of a tutor than with online help. Currently many novice users begin using computers within the educational system where the two more favoured media, human tutors and printed documentation, are readily available. In such an environment, it might be argued that the requirement for online help may be considerably reduced or even be eliminated. However, information about this group of novice users' requirements appears incomplete. This study aims to redress that situation by discovering if these users found it necessary to use the online help systems and how useful they found each of the help systems. Overall results indicate that students tended not to differentiate between the help systems of the three application packages. A relationship was found between the students' perceptions of usefulness of the online help and the amount of previous computer use.
    Keywords: Online help systems, Novice computer users, Online documentation
    Designing and Evaluating User Interfaces for Electronic Examinations BIBAK 174-181
      Michael J. Rees
    BRUIT{Delta}Exam allows a Macintosh personal computer to replace the traditional written examination. Students read the examination questions on the screen and type their answers below the question text. Their completed script is saved for marking by the assessor. Other modules of BRUIT{Delta}Exam allow the assessor to create electronic examination papers and mark the completed scripts.
       The user interface of BRUIT{Delta}Exam has been developed in actual user trials over a period of two years and has been used for several courses at Bond University. The students and assessors experiences of the interface designs are discussed and suggestions for extensions presented.
    Keywords: Electronic examinations, Interface design, Interface evolution

    Closing Keynote

    Getting HCI on the Agenda: What's the Message? BIBAK 182-189
      Gitte Lindgaard
    HCI is not an integral part of systems design yet. This paper claims that some reasons for this are attributable to the HCI community itself neglecting to practise what it preaches, namely to use the language of its users in communicating HCI cum Human Factors messages to developers. We need to rethink the purpose of HCI guidelines, to think through the implications of findings from tests before these are conducted, to communicate these in a language that make sense to developers, and to calculate the value of such findings in monetary terms for senior management. Possible ways to integrate HCI into design are outlined.
    Keywords: Guidelines, Test results, User's language, Monetary values

    Poster Abstracts

    Warning: Computers are a Health Hazard BIBAK 190
      V. Coomb; K. Cooper
    In this day and age it is not unusual to see students and staff within tertiary institutions using computers continuously throughout the day without adequate breaks. The Commerce Faculty at the University of Wollongong have developed a Computer Based Training (CBT) program entitled ERGONOMICS, which demonstrates the appropriate techniques to apply to reduce the effects caused by incorrect computer use. The software package warns the user against excessive use and provides an extensive exercise program to take during daily breaks. Additionally, the user is presented with a series of procedures and rules to apply when using the keyboard and mouse. It is hoped by introducing all new staff and students to the software, the instance of eye strain, repetitive strain injury and lower-back problems will be a thing of the past. This paper covers the entire software development process beginning with the original idea and finishing with the introduction of the completed product.
    Keywords: Computer based training (CBT), Ergonomics
    The Pictoid Package -- Extending HyperCard -- Sticky Buttons BIBAK 190
      Nigel Perry
    The Pictoid Package (Perry, 1992a) is a tool kit for stack designers consisting of scripts, fonts, and external commands. There are two main components in the package, Sticky Buttons and Pictoids. Sticky Buttons are "buttons" which may be inserted into the text of fields, and if the field is scrolled will move with the text. Pictoids are windoids which may be used to provide any-shape buttons, any-shape pictures & QuickTime movie display, any-shaped buttons containing pictures or QuickTime movies, etc. This paper describes the implementation and use of Sticky Buttons. A companion paper describes Pictoids (Perry, 1992b).
    Keywords: HyperCard, Tool kit, HyperText, Active text, Sticky buttons
    Design and Implementation of a Postgraduate Computer Human Interaction Subject BIBAK 191
      Ron Balsys
    The University of Central Queensland is running a Honours course in Computing in both on campus and off campus modes. Subjects within this stream will include a two semester course in Computer Human Interaction (CHI). The design of this subject has been influenced by both a review of what has been done elsewhere and the requirement that it be doable by students who do not attend campus due to work or other commitments. It is expected that this subject will be offered in a course work masters sometime in the near future.
       A description of the contents of the course, included text and reference material. A set of practical tasks that prospective students must do are described. The combination of theory and practise is designed to give the student a high level of proficiency in the mainstream areas of CHI as well as giving the broader view of the range of issues to be examined in CHI.
       Training in HCI is important to the success of the software industry in Australia. Offering of advanced courses in CHI in distance mode should allow those people who are currently working in the field to gain valuable experience in this area. This will be reflected in both increased productivity and competitiveness of the software industry in this country.
    Keywords: Human computer interaction, Training, Education, Honours course
    Developing Guidelines and Standards for Usability BIBA 191
      Judy Hammond
    There is considerable debate amongst practitioners about rules, guidelines and standards relating to usability. Many organisations would like to be able to provide guidance for their systems developers and designers about how to ensure usability is incorporated into their systems development life cycle, but are unsure how this might be done. This paper discusses rules, guidelines and standards for usability in general terms, their importance, their usefulness and their pitfalls. The paper discusses a process for developing guidelines, based on practical experience in developing usability guidelines within the framework of a large organisation.
    Participatory Research for Knowledge Support Systems Development BIBA 192-193
      Linda Candy
    Human-Computer Interaction (HCI) is a technology-centred field that employs well established scientific disciplines such as Psychology alongside Engineering concerns. Thus the research methods employed do not form a set of unified scientific approaches. HCI research is directed towards methods and tools that aid the construction of well designed artefacts or products. The goal of deriving generally applicable scientific knowledge is often subordinated to the drive for well designed and engineered products that incorporate sound principles. In reality, the rapid time scales of technology driven research is in conflict with the needs of basic science. The selection of lines of scientific investigation and the research methods employed, can be seen as one dependent to a large extent upon the set of the "beliefs" that prevail in the community of researchers in any given discipline (Barnes, 1974). Lakatos (1970) argues that scientific progress is not about the successive overthrow of dominant theories but a situation of "progressive and degenerative problemshifts" and the outcome of rival research programs in contention. In the development of computer technology particularly in the field of interactive computing, there is a wide diversity of approaches and a steady shift away from traditional "mainstream" scientific methods.
       This paper will survey the research methods current in HCI and place our research strategy proposed in that context. Based on empirical studies, we have derived a strategy for participatory research for knowledge support systems development. This encompasses the intertwining roles of the personnel involved, whether domain specialist or designer or investigator, and different techniques according to the goal of the exercise at any given point in the cycle of studies. In that context, the centre of concern is the role of a knowledge support system in that process and the reflexive relationship between research problem and research methods (Woolgar, 1988). From an initial emphasis upon the development of a knowledge-based system for Speech Recognition (Connolly et al, 1986) a quest for appropriate research methods for studying the "co-operative" interaction between domain specialist and knowledge-based system emerged (Candy and O'Brien, 1992).
       A study of the approach to knowledge development we have adopted and the implications for system design are described. A key element is knowledge representation at the user interface represented to the user in domain familiar language, abstractions and images which are then mapped to the machine representation. We found that not only did the system enable the specialist to capture expertise in an effective manner, more significantly, it opened new avenues of research and supported the evolution of new understanding about the domain (Candy and Edmonds, 1992).
       This form of support to scientific investigation is an exciting and ongoing research question. Two issues are of particular interest: developing machine tractable representation of domain specific knowledge and discovering design principles for user interfaces that actively support that process. In the course of this work a research strategy has been developed and used. The paper proposes that the research strategy that has been developed is more generally applicable.
    Explanations and Communication of Knowledge from Computer to Human BIBAK 193-194
      Shirley Gregor
    Knowledge-based systems can include an explanation facility which gives information such as why certain questions are asked, what terms mean, how conclusions are reached, why other conclusions aren't reached and so on. The explanation facility overlaps to some extent with help and reporting functions but goes beyond these in that different and deeper types of knowledge can be communicated. Explanation facilities are generally held to be desirable but there is little research investigating their use in practice or laboratory studies. The capacity to explain reasoning is also regarded as important by those who expect that knowledge-based systems can be used in training. Again little research work has been done to examine learning as a side-effect of the use of knowledge-based systems.
       This paper describes an investigation into human learning and the use of explanations when a knowledge-based (expert) system is employed as an aid in problem-solving. Theory from cognitive psychology led to the propositions that (i) learning would occur when an individual used a computer program which was not specifically designed for tutoring, but did give advice and explanations in the course of problem-solving activities, and (ii) individuals who were at a lower stage of skill acquisition before a period of problem-solving would make greater use of explanations than individuals at a higher stage prior to the activity. Twenty students studying taxation law were matched on prior grades and then randomly assigned to one of two experimental groups. One group completed a problem-solving session in which they had an expert system as a decision aid. The other group completed the same activities with reference material available but without the expert system. Subjects were given a pre-test and post-test to measure the extent of learning. Use of various explanation functions by the expert system using subjects was automatically recorded by the expert system.
       Results were contrary to expectations though learning occurred with both groups. The non-expert system using group showed greater improvement though the difference in improvement between the two groups was not significant. Scores on the pre-test were positively, rather than inversely related to the number of times explanation functions were accessed. Analysis of the usage of explanation functions showed that a function giving a worksheet-type screen with details of arithmetic calculations was used more than functions giving narrative descriptions of procedures.
       The experiment involved a relatively short period of problem-solving and it was possible that the time and effort involved in gaining familiarity with the computer program may have been a confounding factor. Further investigation with longer trial periods and larger groups was suggested. The positive relationship between degree of mastery and use of explanations was interesting as it suggested that help information was not used most by those to whom it would have appeared to be most useful. Reasons were suggested for this finding, including the idea that less cognitive effort was involved in understanding explanations when an individual had already a well-developed cognitive structure for relevant knowledge.
    Keywords: Expert system, Computer assisted teaching, Explanation facility
    Towards a Design for HyperCompass -- An Intelligent Software Agent BIBA 194
      Daniel Jitnah; John Lenarcic
    The concept of hypermedia offers an integrated computer-based approach to information management and knowledge organisation. Hypermedia environments have the potential to provide users with interactive access, via associative indexing, to vast quantities of multimedia data connected by non-linear links.
       The progress of hypermedia has been hampered due to technical problems inherent in manipulating massive amounts of computer-based data, one such hindrance being user disorientation in information spaces.
       This poster will focus on an ongoing project to improve user navigation in hypermedia documents. Our aim is to create an intelligent software agent -- HyperCompass -- serving to augment the human-computer interface to hypermedia by providing users with navigation and other rudimentary assistance when required.
       An embryonic design for HyperCompass, embedded within a prototype hypermedia system, will be outlined. The initial design focuses on using the logic programming paradigm for the actual development of HyperCompass. The limitations of existing software prototyping tools for authoring "intelligent" hypermedia documents will be discussed. In mapping out future directions for research in this area, speculation will also be offered as to whether the new technology of Virtual Reality can improve the user interface to hypermedia even further.
    Quasi-Experimental Studies of E-Mail Usage and Usability BIBA 195
      Perry Morrison
    E-mail usage of volunteer subjects at two Australian universities was monitored in this two phase study. Phase one observed computer science faculty and first year computer science students over a twelve month period. Results showed that both groups used less than one third of the functionality provided by the unix mail interface, although experts did exhibit more sophisticated usage.
       Phase two trained post-graduate e-mail novices on either unix mail, or the menu based mailer known as Elm. A three month monitoring period showed that Elm users sent significantly more messages, used a greater array of functions, generated lower error rates and needed less help access. For the first four weeks however, the number of different features accessed in the two interfaces was indistinguishable and after 6 weeks, error rates and help access were indistinguishable, although Elm users continued to send significantly more messages.
       In conclusion, users can tolerate poor mail interfaces and eventually achieve performance levels comparable to better mail interfaces. However, the time for performance to converge may be extreme in certain applications and commercially unjustified. In addition, more frequent usage and use of a wider array of functions were long term benefits that the better interface encouraged.
    Executive-Computer Interfaces -- A Case Study BIBA 195-196
      Helen Hasan
    The introduction of computer technology into organisations began in the 1950's and 1960's with transaction processing systems. These were followed in the 1970's by Management Information Systems (MIS) and Decision Support Systems (DSS) which were intended to service the needs of middle and upper level management respectively. By the early 1980's it was clear that few senior managers were making use of computers in their jobs and a commonly held view was that computer systems were not appropriate for the vast majority of managerial tasks. However a number of researchers proposed that the roles of senior management were sufficiently different from their subordinates to justify a separate class of systems.
       As a set of users, managers have particular characteristics and needs. They nominally hold positions of power within the organisation but often feel locked out of sections of the organisation where the computer, and its operators, have taken over. They are varied in their work styles and computer awareness, but are often not very computer literate. They will demand a system that is easy to learn and use, that is reliable, accurate, current and fast. In this context the term Executive Information Support (EIS) was introduced, followed by Executive Support Systems (ESS).
       Reports on the successful adoption of EIS systems by organisations, seem to indicate that there is a small but growing trend among executives to make use of this type of system. This study deals with an investigation of the progress of a medium sized public organisation as it is going through the initial stages of integrating the needs of senior management into it Information Systems. The author acted as an independent (unpaid) consultant, interviewing and advising employees at all levels of the organisation. The two areas of particular importance appeared to be the design of the dialogue between the system and the executive user, and the method by which the executive users become involved in the system's development and evaluation.
       Many of the findings in this study confirm those of previous work although several additional issues were identified. In particular this study focuses on the problems encountered at the interface between executives, as a group of users, and an EIS system that is evolving from existing Information and Data Processing Systems.
    Teaching the Environment of Systems Design BIBA 196-197
      Penny Collings; Anne McMahon; David Walker
    In teaching systems analysis and design, three main aspects must be covered:
  • 1 analysis tools (e.g. system modelling tools);
  • 2 analysis and design as a mutual learning process involving both user and
       designer; and
  • 3 the understanding of the environment from which the need arises and into
       which the resulting system must integrate. The first of these can be addressed through traditional written assignment specifications. We have attempted to address the second of these through the use of role play within an iterative design framework with considerable success. The third of these, however, presents problems because of the complexity of the environment, and in particular the need for any system to satisfy the political, social and technical requirements of the organisation.
       One way of addressing this problem is by simulating the environment in which information technology development takes place. The simulation should address a significant number of the major issues in systems development, e.g.
  • 1 determination of requirements for systems, both at a strategic level (what
       systems are required) and at the level of an individual system;
  • 2 system design, both at a system architecture level, e.g. centralised vs
       decentralised, mainframe vs PC, packages vs in-house development and at an
       application level, e.g. user-interface design, prototyping;
  • 3 strategies for system development, e.g. "big bang" approaches involving
       development of large, monolithic systems vs evolutionary development of
       smaller by potentially interacting systems, including ideas about open
       systems and standards to ensure compatibility;
  • 4 the management of the development process, including project planning and
       control, quality control, outsourcing;
  • 5 evaluation and acceptance of proposals and completed systems. Participants in the game should be involved in all aspects of this environment, i.e. their roles should include IT professionals (e.g. analysts/designers, data administrators, DP managers), users, and different levels of management.
       We are currently developing a simulation of an Australian public sector organisation which attempts to incorporate these features. The game will also incorporate an IT infrastructure which allows participants to perform some tasks in a cooperative manner with computer support. The idea is to enable a holistic appreciation in which human behaviour is emphasised in relation to technology, and both are seen to have a central place in systems design. The completed simulation will be flexible enough to work with a range of IT infrastructures and will be able to be played in any networked behavioural laboratory. It will provide a frame work for further simulations.