HCI Bibliography Home | HCI Conferences | INT Archive | Detailed Records | RefWorks | EndNote | Hide Abstracts
INT Tables of Contents: 84879095979901030507-107-2

Proceedings of IFIP INTERACT'84: Human-Computer Interaction 1984-09-04

Fullname:Proceedings of IFIP INTERACT'84: Human-Computer Interaction
Editors:Brian Shackel
Location:London, England
Dates:1984-Sep-04 to 1984-Sep-07
Standard No:ISBN 0-444-87773-8; hcibib: INT84
  1. Keynote Addresses
  2. User Aspects
  3. Visual and Display Characteristics
  4. Workstations and Workplace Issues
  5. Input and Output -- Some New Approaches
  6. Input Methods and Comparisons
  7. Speech Input and Output
  8. Dialogue Interaction / Analysing Interactive Dialogues
  9. Dialogue Interaction / Graphical Interaction
  10. Dialogue Interfaces / Command Interfaces
  11. Dialogue Interfaces / Menu Interfaces
  12. Dialogue Interfaces / Adaptive and Flexible Interfaces
  13. Tools to Aid Interface Design and Programming
  14. Language Design and Comprehension
  15. Knowledge Based Techniques
  16. Modelling Users and User Interactions
  17. Design -- Approaches and Methods
  18. Design -- Guidelines
  19. Evaluation -- Approaches and Methods
  20. Learning and Training
  21. Aids for the Disabled
  22. Organisation and Social Issues
  23. Behavioural Issues in the System Development Cycle / Introduction and Overview
  24. Behavioural Issues in the System Development Cycle / System Examples and Deductions Therefrom
  25. Behavioural Issues in the System Development Cycle / Evaluation Issues
  26. Usage Issues in Electronic Mail, Conferencing and Journal Systems / Introduction
  27. Usage Issues in Electronic Mail, Conferencing and Journal Systems / Overview of Systems
  28. Usage Issues in Electronic Mail, Conferencing and Journal Systems / User Experience and Usage Results
  29. Usage Issues in Electronic Mail, Conferencing and Journal Systems / Problems and Issues of Present and Future Systems

Keynote Addresses

From Ergonomics to the Fifth Generation: 30 Years of Human-Computer Interaction Studies BIBA 3-7
  Brian R. Gaines
From the earliest days of computer systems it was realized that computers were essentially tools to be used by people and that good human factors were an essential part of system design. In the early years the struggle to generate reliable, low-cost hardware and software dominated the industry and human factors played a minor role in computer science and technology. In the past decade, however, advances in vlsi and software engineering have made advanced computer systems increasingly widely available and manufacturers have turned to human factors for product differentiation. In particular, human-computer interface requirements have a dominant role in fifth generation computer specifications. This paper surveys the development of human-computer interaction research from the 1940's through to the present day. The goals and expectations of the research, trends in the literature, the status of the results, and the directions of future development, are discussed.
Designing for People in the Age of Information BIBA 9-18
  B. Shackel
Some characteristics of the Information Age and the importance of human factors issues are outlined. Immediate questions for the next 7 years or so are discussed, including 9 substantive areas needing research (from a recent survey) and the development and better implementation of design procedures. Longer term questions discussed are -- the passing of paper, the reduction of writing, the victory of voice, the wired society and the expert in the system. Finally, some of the important broader issues are mentioned and the need for synergy by human and information engineers is emphasised.

User Aspects

Users in the Real World BIBA 21-26
  David Owen
Based on the premise that people demonstrate a considerable degree of competence at formulating and achieving goals in the world, this paper seeks to identify and examine the relationship between the crucial characteristics of the real world and inherent or acquired human skills that support this competence, in order to improve the human computer interface. Aspects examined include a "naive physics" of computing and the reconstruction of propositionally held information.
Choice of Interface Modes by Empirical Groupings of Computer Users BIBA 27-32
  K. M. Potosnak
A method for classifying computer users on a large number of variables was developed and tested. The method was based on cluster analysis and yielded seven groups of users and non-users. Volunteers from each group used a computer program which allowed them to choose among three modes of interaction: prompts, commands and form-filling.
   Overall, people preferred the prompts and the form, but used the prompts the most. People with the most use and knowledge of computers used the commands more than the other groups. Discussion focuses on the hypothesis that naive users prefer a system-driven mode of interaction, while experienced users prefer a user-driven mode.
The Social Psychology of Computer Conversations BIBA 33-38
  Dianne Murray; Nigel Bevan
Human conversations are complex interactions motivated by both task-related and social goals. It is proposed that the optimal form of conversational interaction with a computer is a computer conversation which closely emulates the nature of a human conversation in similar circumstances. The cultural differences between computer conversations and human conversations are identified, and it is suggested how the crucial social factors in human conversation could be integrated into a conversational model of computer interaction, structured by the need to achieve mutual goals.
Minimalist Design for Active Users BIBA 39-44
  John M. Carroll
Recent studies in computer human factors indicate that novice learners of office systems are "active", preferring self-initiated problem solving to rote drill and practice as a learning strategy. But in high-function systems users may need help in exploring basic functions without being distracted by advanced material. Two experimental approaches to this problem are outlined: The MINIMAL MANUAL attempts to support active learning by providing concise instruction focussed on easy-to-understand goals. The TRAINING WHEELS WORD PROCESSOR encourages exploration of basic functions by disabling the more advanced functions that can distract and confuse novices.
Note: Reprinted in Baecker & Buxton, 1987, p. 621
The Role of Experience in Editing BIBA 45-50
  Mary Beth Rosson
An important question for designers of text-editing systems is the use to which the systems are put by experienced users. Most systems provide a range of function from basic to very advanced, yet their designers typically do not know whether users ultimately make use of the full range of function, or indeed whether they develop effective use of even the most basic function. In the present work, survey and automatic monitoring methodologies were combined to study experienced editor users. The two methods provided converging evidence that not all users learn to exploit a system's facilities simply through continued experience with the system. Many time-saving strategies (e.g., assignment and use of program function keys) were associated with job type and with prior experience on other editing systems. The implications of the results for the design of editors are discussed.
On-Line Composition of Text BIBA 51-56
  S. K. Card; J. M. Robert; L. N. Keenan
The use of text-editors for writing original text has been little studied despite the importance of the task. A study conducted by Gould found that the composition rate of writers using a text editor was more than 50% slower than the same writers writing by hand. We show that the source of the slowness is the design of the text editor and that using a display-oriented editor writers can write as fast and good typists faster than by hand. Like Gould, there was no quality difference based on the source of the letters and users of the text editor made many more modifications. Fewer than half of the modifications users made actually improved the text.

Visual and Display Characteristics

Legibility of the VDT Required for High Performance Task BIBA 59-63
  Hiroshi Tamura; Noboru Takematsu
This paper proposes an experimental method to evaluate legibility of the display and efficacy of the pointing device, to be applied to compare human performance both at VDT and paper-pencil system. The result confirmed the performance is still considerably low compared to the paper-pencil system. The difference is larger for the task of the higher work load. The measure of performance defined in this paper may be used as the objective of improving devices.
An Investigation of Visual Discomfort amongst Clerical Workers with Specific Reference to the VDU as a Potential Causal Factor BIBA 65-73
  Howell Istance; Peter Howarth
This paper reports the results of a field survey of visual discomfort amongst 192 subjects divided into 4 occupational groups. Each subject was studied over the 5 consecutive days of a working week. Pre-work and post-work recordings of subjective symptoms were made on each day and of objective measures of visual function on 2 days. It was intended that the four groups comprise two matched pairs of 'VDU-users' and 'non-VDU users'. Post hoc analysis showed that one of the two pair of groups could be considered thus. No significant differences were found between the 'VDU user' group and the 'non-VDU user' group in terms of subjectively reported visual discomfort or objective measures of visual function.
Visual Discomfort and Cathode Ray Tube Displays BIBA 75-79
  Arnold Wilkins
Visual stimuli can induce a variety of noxious effects, including unpleasant illusions, tired eyes, headaches and seizures. The stimulation responsible for these effects is quite specific and includes that from televisions and text. According to a general theory of visual discomfort reviewed here, the effects have a common physiological basis.
Quantitative Measures of the Spatial Properties of Screen Designs BIBA 81-89
  Dennis J. Streveler; Anthony I. Wasserman
This paper proposes certain quantitative measures for analyzing the spatial properties of screen designs and for designing experiments aimed at determining optimal values for the measures. The measures result from the application of three analytic techniques termed boxing analysis, hot-spot analysis, and alignment analysis.
VDT Screen Resolution and Operator Performance BIBA 91-95
  Jeffrey L. Harpster; Andris Freivalds
The study compared operator performance on a visual search task with the resolution of the display used. The four displays consisted of low, high, high resolution simulating low resolution and hard copy print as a control. Performance was significantly worst on the low resolution, better for the high resolution modes displays and was best for hard copy print. An explanation based on the spatial frequency characteristics of the display stimulating the accommodative system is given.

Workstations and Workplace Issues

'Environment' Problems in Introducing Office Information Systems BIBA 99-101
  J. J. Florentin
At the present time many organisations are changing from centralised time-sharing computing to distributed office information workstations. There are technological advantages, but there is also felt to be a very significant advantage in placing computers under local and personal control. Experience with three such situations unfortunately shows that in organisations with shared information there is a need for central data control which can make such local autonomy infeasible. The difficulties, and a possible future resolution, are discussed.
A Universal Workstation Concept for Air Command & Control Application BIBA 103-108
  K. Camm
The shortcomings of air defence and management information system terminals and workstations in an integrated air command and control system are discussed and the requirements for a universal workstation "kit" based on "off-the-shelf" commercial products are introduced. A design outline for such a workstation is presented with several examples of its capabilities in a tactical air system to support the concept, based on work in SHAPE Technical Centre's ACCS Test Bed.
Human Factors Aspects of Terminals for Non-Expert Users BIBA 109-113
  W. Noe
Continuing advances in office automation are putting computer power on the desks of more and more non-expert users. An example of this is interactive videotex, which is intended to give every white-collar worker low-cost access to databases in the office. This calls for a new class of terminals, which should be easy to operate and suit the non-expert users' needs.
Comprehensive Development for People in STA; Efforts and Results BIBAK 115-119
  Per Abrahamsson
The coming and growing use of video display terminals (VDT's) in Operators' Workstations in STA has been accompanied by rising worries for negative health effects from the operators and their Trade Unions.
   This paper reports efforts made within STA in order to eliminate possible problems, such as physical discomfort of various kind, reduced job satisfaction etc.
   Also reached and expected results are reported.
Keywords: Ergonomics, Human factors, Operators, Operator services, Telematics, Video display terminals (VDT's), Workstations and environment
The Impairment of Concentration as a Consequence of Computer Work BIBA 121-125
  Kazuo Saito; Toshiyuki Hosokawa; Kunihiko Nakai
In this presentation, we set out to describe the impact of working a computer and the impairment of computer-workers from the work physiological view point. Attentional phenomena will be classified and discussed from the standpoint of higher nervous functions. We analyse the physiological basis of visual attention required for the operation of computers by the electrophysiological methods of brain waves, and the auditory evoked response of the electroencephalogram.
   The VDT work-load as an example of computer-work will be made clear by an analysis of a physiological performance test and of the neurochemical substances in the urine of the workers.
Dialogue Handling with User Workstations BIBA 127-131
  Tom Carey
This paper discusses the requirements for a user workstation in an open network environment. The workstation implements the dialogue component of an interactive application, while the functional or computational component may be implemented on a network host machine. Distributing processing in this way is influenced by ideas from three areas: software ergonomics, software engineering, and computer networks. An initial prototype workstation is reviewed, and research issues shaping a new implementation are described.

Input and Output -- Some New Approaches

New Techniques for Gesture-Based Dialogue BIBA 135-138
  Martin Lamb; Veronica Buckley
A music editor is described which requires no typing and which has been used successfully by children as young as three years of age. No syntax errors are possible, and all input methods are intuitive -- stroking, pointing, dragging and circling, for example. Commands (such as move, delete or copy) are not specified by words, but are conveyed by gesture; (for instance, in order to delete an item, the user merely circles it and pushes it off the screen)
   Throughout the programs, the auditory, visual and kinesthetic modalities are employed either simultaneously or in immediate proximity, thereby reinforcing the intuitive nature of the interface. Alternative user-models of interaction are supported. The techniques presented form the basis of a computerised Learning Environment for music and a User-Interface Management System.
On-Line Cursive Script Recognition BIBA 139-143
  C. A. Higgins; R. Whitrow
A method is proposed for the automatic recognition of handwritten words using a hierarchical description. Possible letters and segmentation shapes are suggested by matching the script to stored templates. Ambiguities are resolved using higher level context in the form of letter quadgrams and a dictionary lookup.
On-Line Acquisition of Pitman's Handwritten Shorthand as a Means of Rapid Data Entry BIBA 145-150
  C. G. Leedham; A. C. Downton; C. P. Brooks; A. F. Newell
In this paper we discuss the use of Pitmans shorthand as a means of converting dictation speed speech (up to 120 wpm) directly into readable text for computer entry or direct output. The Pitman shorthand notation is compared to a machinography or machine compatible script and the recognition problems associated with handwritten shorthand are discussed. The requirements of a writing tablet and instrumented pen for on-line acquisition of Pitman shorthand are described and the preprocessing techniques which have been usefully applied to the raw data are outlined.
Automated Machine Shorthand Transcription in Commercial Applications BIBA 151-156
  A. C. Downton; C. P. Brooks
This paper describes the development and evaluation of a Palantype machine shorthand transcription system for court reporting and commercial applications. The system is based upon a transcription computer which provides special purpose software for transcription from Palantype to English, for efficient editing of transcripts, and for maintaining and optimising the Palantype to English dictionary. The design of the user interface to the system and the dictionary structure are described in detail, to illustrate how these aspects influence the overall efficiency and commercial viability of the system.
An Adaptive Editor for Shorthand Transcription Systems BIBA 157-161
  R. Dye; A. F. Newell; J. L. Arnott
An automatic transcription system for machine shorthand takes the output from a shorthand machine, and converts it into a Draft Transcript. This draft transcript needs to be edited to perfection using word processing techniques. A suite of programmes has been written which takes advantage of the particular characteristics of Palantype transcripts to provide a very efficient editing environment. The editor adapts to the user in a way which improves his efficiency without an overhead of the necessity to learn complex control structures. Some of these facilities are also appropriate to a standard word processing environment.
Interactive Stereoscopic Computer Graphic Display Systems BIBA 163-168
  Neil Storey; J. Ffynlo Craine
Conventional CRT displays give only a two dimensional representation of three dimensional objects which means that depth information can be presented only indirectly, using such techniques as hidden line removal, shading and object rotation. Improved perception of 3D images is possible by presenting to the viewer a stereoscopic pair of two-dimensional images. To be truly effective, the image presented by a stereoscopic display should change as the viewing point moves. Such a system requires a knowledge of the user's head position and orientation in order to compute the view to be presented to each eye. The paper describes an approach to a system of this type.

Input Methods and Comparisons

Input Devices for Public Videotex Services BIBA 171-175
  M. Francas; D. Goodman; J. Dickinson
A number of experiments aimed at developing effective key-press input devices for public videotex services are reported. Results generally indicated that performance differences were minimal across a variety of devices for different types of users (novice and experienced). There were, however, some distinct preference differences exhibited by the two user groups. Novices strongly favoured a simple minimal-function keypad, whereas subjects trained to a high level of Telidon proficiency preferred the perceived functionality of a full-sized keyboard, even though it did not lead to improved performance.
Comparison of Input Devices for Correction of Typing Errors in Office Systems BIBA 177-182
  R. Haller; H. Mutschler; M. Voss
In an experiment several devices for correction of typing errors in office systems were compared. They were chosen with regard to an increasing degree of compatibility between input operation and system's response. The task was the correction of typing errors in a letter already prepared on a word processor. After positioning the cursor on the erroneous character with one of the six locators light pen, graphic tablet, mouse, tracking ball, cursor keys, and speech recognizer, it had to be replaced by the right character with one of the two correctors a-keyboard and speech recognizer. Concerning the locators the results corresponded roughly to the degree of compatibility with voice input as the slowest and light pen as the fastest device. There were no essential differences between correctors.
A Cerebral View of Task Optimality in Japanese Text Typing BIBA 183-188
  Takeshi Okadome; Hisao Yamada; Hiroshi Watanabe; Kenji Ikeda; Masao Saito
We discuss an activity of typists' brains during touch typing, based on knowledge of experimental psychology and cerebral physiology, especially in connection with the functional lateralization of cerebral hemispheres. Our hypothesis is that a skilled copy typist depends heavily on the cortical reflex in response to direct visual information from a manuscript, but depends less on the linguistic facility of the left hemisphere. As an indication of the validity of the hypothesis, we report our results of the measurements of typists' electroencephalograms (EEG) during Japanese touch typing. Our findings are preliminary but they are compatible with the hypothesis.
A Comparison of Selection Techniques: Touch Panel, Mouse and Keyboard BIBA 189-193
  John Karat; James E. McDonald; Matt Anderson
A study was conducted testing user performance and attitudes for three types of selection devices. The subjects were tested on target selection practice tasks, and in typical computer applications using menu selection and keyboard typing. The study showed an advantage for on-screen touch panel over keyboard selection, and for keyboard selection over mouse entry. Differences between this result and those reporting an advantage of mouse selection are discussed.
The Effects of Input Medium and Task Allocation Strategy on Performance of a Human-Computer System BIBA 195-200
  Siu-Tong Lam; Joel S. Greenstein
The allocation of tasks between human and computer, and the merits of a dynamic approach to this allocation are discussed. Dynamic task allocation requires efficient human-computer communication. This communication may be accomplished in an implicit or explicit manner. A conceptual framework for the study of explicit human-computer communication is introduced and a study exemplifying the use of the framework is presented. This study investigated the effects of two input media and four task allocation strategies on the performance of a human-computer system. The task environment represented a simplified version of an air traffic control scenario wherein computer aid could be evoked by the human to accomplish task sharing between the human and the computer.

Speech Input and Output

Speech Input as an Adjunct to Keyboard Entry in Television Subtitling BIBA 203-208
  R. I. Damper; A. D. Lambourne; D. P. Guy
This paper describes an investigation of the potential benefits of using automatic speech recognition in television subtitle preparation. Analysis of the subtitling task indicated that the principal benefits were likely to be obtained in spoken entry of subtitle 'style' commands. However, off-line subtitling trials showed that use of speech recognition increased preparation time by 9%. Speech input significantly reduced the time spent transferring between text and style entry but increased the time spent in other activities. The error rate for style entry by keypad was consistent and averaged 5% whereas speech recognition errors ranged from 5% to 15%. A valid comparison of the two modes of entry can, however, only be made with subjects having as much experience of voice entry of data as they have of keying.
The Use of Simple Speech Recognisers in Industrial Applications BIBA 209-213
  David Visick; Peter Johnson; John Long
This paper points out, and attempts to deal with, some of the problems that may be encountered when using simple speech recognition systems in industrial applications. An experiment compared a voice recogniser with a keyboard, as the destination input device in a parcel sorting task. The task was represented first by a simple laboratory simulation of the coding sub-task, and then by an authentic simulation using real parcels on a sorting rig. Results showed that voice input may be quite unsuitable for tasks having little or no manual content. Also, for tasks requiring precise sequencing of operations, voice may offer inadequate intrinsic timing feedback. Finally, a practical means of empirical vocabulary optimisation is described.
How Should People and Computers Speak to Each Other? BIBA 215-218
  M. A. Richards; K. M. Underwood
Voice is becoming a more common mode of communication between man and machine. We have previously shown that computer-naive users naturally adapt the way that they ask questions in an information retrieval task when they think that they are addressing a computer system, as opposed to a human (3). The present study investigates how the content of the messages with which the computer system initially addresses the user can further encourage behaviour that would be potentially useful for enabling automatic speech recognition and analysis. Elements of 'politeness' and 'explicitness' in the system's initial message were varied, and the regularity and conciseness with which users responded was studied. Several aspects of users' behaviour were found to be affected, and age and learning effects were also observed.
Human Factors and Synthetic Speech BIBA 219-224
  John C. Thomas; Mary Beth Rosson; Martin Chodorow
Recent advances in linguistics, speech science, psychology, and especially in computers have made unlimited text-to-speech conversion systems a practical reality. However, the use of audio output from a computer poses special problems in ergonomics, most of which have not been dealt with in the literature. In this paper, we review relevant findings in the literature and recent work in our own laboratory. We then provide some guidelines for good human factors in applications that use speech synthesis. These guidelines address both the process of development and suggestions for the end-product. The latter must be considered highly tentative due to the nascent nature of this research area.
Experiments in Speech Interaction with Conventional Data Services BIBA 225-229
  Christopher Labrador; Dinesh Pai
Access to conventional data services typically involves the use of a standard keyboard terminal over a data network. This paper discusses our experiments at BNR in using speech as an alternative. We used a text messaging service as the vehicle for our experiments to investigate the interaction between humans and computers over the telephone, voice network using the telephone as a terminal. In our experiments, presentation to the user is accomplished using speech synthesis and user input is effected with DTMF (Dual Tone Multi-Frequency) signaling.
Speech -- The Natural Modality for Man-Machine Interaction? BIBA 231-235
  A. F. Newell
In recent years the use of speech as a man-machine interface has received considerable prominence. A number of systems have been developed which use speech output from machines, and speech input has been introduced in a small number of cases with rather less effectiveness. In those cases where it is impracticable to look at a display, or the hands are fully occupied, speech has obvious advantages, but often a major justification for the use of speech has been that is the 'natural' method of communication for man, and therefore must be the optimum solution. This contention, however, is a simplification of the situation, and, in general, much greater thought must be given to the choice of modality of input-output means than is implicit in justifications of this nature.

Dialogue Interaction / Analysing Interactive Dialogues

Window-Based Computer Dialogues BIBA 239-243
  S. K. Card; M. Pavel; J. E. Farrell
In recent years a number of systems have used windows as the basis for advanced user interfaces. Yet how exactly users benefit from windows or what features of windows are important for design is neither understood nor has it been studied. Current window designs are given a simple classification. Seven functional uses that have been identified for windows are described. The Window Working Set concept based on operating system theory is introduced for the analysis of space constraints on window use.
Note: Reprinted in Baecker & Buxton, 1987, p. 456
Generic Commands BIBA 245-249
  Jarrett K. Rosenberg; Thomas P. Moran
A generic command is one which is recognized in all contexts of a computer system; examples from the Xerox 8010 Star system are move, copy, and delete. They may be viewed as extremely general actions which make minimal assumptions about their objects, the particular interpretation of the commands depending on the contexts in which they are issued and the nature of the objects to which they are applied. Of the several tradeoffs involved in using generic commands, the primary one concerns having to design the objects in the system so as to efficiently use them.
Transaction Processing Using Videotex or: Shopping on PRESTEL BIBA 251-255
  John Long; Paul Buckley
The suitability of videotex (VT) for supporting transactional services (including shopping) was assessed, firstly by relating VT technology to a model of transaction processing. Expectations regarding suitability were generally confirmed by an observational study in which naive subjects ordered goods via two different systems. Further assessment involved analysis of user difficulties and errors. These were used to identify variables affecting performance and were modelled in terms of a mismatch between a naive user's incomplete representation of the system and inappropriate representations which interfere with task performance.
People Can Retrieve More Objects with Enriched Key-Word Vocabularies. But is There a Human Performance Cost? BIBA 257-261
  Louis M. Gomez; Carol C. Lochbaum
Recent research using statistical data and models has suggested that information systems with many different and not necessarily unique names for each system object will dramatically increase the likelihood of finding a target object over more standard retrieval systems in which each object has only one or a few names. In the current experiment this conclusion was supported by observations of people using actual information systems to retrieve target objects. In addition, those people using systems with richly indexed vocabularies needed slightly fewer, rather than more, key-word entries to find a given target object.
Accessing Large Data Bases: The Relationship between Data Entry Time and Output Evaluation Time BIBA 263-267
  Carla J. Springer; James F. Sorce
The total time required for retrieving information from large data bases has two components: (1) the time spent entering the retrieval request and (2) the time spent evaluating all of the output that matches the request. There is an inverse relationship between these two components. This study assessed the particulars of this data-entry/output-evaluation trade off in the context of Directory Assistance. Results argue strongly that total retrieval time is reduced when output evaluation time is reduced even at the expense of increased data entry time.
Constructive Interaction: A Method for Studying Human-Computer-Human Interaction BIBA 269-274
  C. E. O'Malley; S. W. Draper; M. S. Riley
In this paper we describe a promising technique for studying human-machine interaction (HMI) called Constructive Interaction. We describe two pilot studies from a set which explore its application to HMI. Constructive Interaction was developed by Naomi Miyake (Miyake, 1982). It consists essentially of recording sessions with two participants who are discussing some topic which they do not fully understand, in the hope of sharing their knowledge and arriving at a fuller understanding. Miyake was interested in what was revealed about the underlying schemas of the participants and how new schemas can originate in an interaction between two people. We are interested in what this basic situation can offer for the study of HMI.

Dialogue Interaction / Graphical Interaction

Saying What You Want with Words & Pictures BIBA 275-280
  Aart Bijl; Peter Szalapaj
A logic modelling system for modelling descriptions of any kind of thing is described. The system employs a symbolic representational structure that permits integration of graphical and textual parts of descriptions, offering these as alternative means for depicting the same information. The goal is to represent knowledge about drawings and words in computers in a way that will enable a computer to interpret intended meaning from human interactions, to understand what people want it to know and do.
Concept Refinement in Social Planning through the Graphical Representation of Large Data Sets BIBA 281-286
  M. Visvalingam
Social planning is part of a process of arbitration. Empiricism is sought as a means of reaching a decision even if not consensus. This paper argues that the prevailing rule-based approach to the production of statistical evidence can confuse as much as clarify social issues. A graphical information system, based on Advanced Information Technology, can expedite concept refinement in complex problems, wherein the underlying assumptions remain implicit and vague rather than explicitly articulated. However, this calls for changes in both the framework for and form of human-computer interaction in computer cartography.

Dialogue Interfaces / Command Interfaces

User Representations of Ordered Sequences of Command Operations BIBA 289-293
  Phil Barnard; Allan MacLean; Nick Hammond
An experiment is reported in which users learned how to operate a "laboratory" system for handling electronic mail. Two variables were manipulated. Users were asked to learn one of two task structures involving eight operations. In one form the task was structured into a sequence of four pairs of semantically related operations (4x2). In the other, operations were structured into two groups of four on the basis of their abstract class. Two sets of command names were employed one being less discriminable than the other. Both variables were found to influence the ways in which users learned the system. The results suggested that users of the 4x2 structure were constructing mental representations in which individual operations were more semantically integrated than users of the 2x4 grouped structure.
The Role of Prior Task Experience in Command Name Abbreviation BIBA 295-299
  Jonathan Grudin; Phil Barnard
An experiment is reported in which subjects previously naive to text editing are asked to generate abbreviations for a set of editing commands. We manipulated the degree of the subjects' experience with the editing task prior to the point at which they were asked to produce the abbreviations. We found effects of experience on both the length and the form of the abbreviations produced, with more experienced subjects inclined toward shorter abbreviations and, independently, toward truncation as an abbreviation scheme. We conclude that experimental paradigms previously used to investigate naming and abbreviation may have encouraged subjects to construe their task falsely as one in which they would be using abbreviations to reconstruct referent names, whereas the actual task involved recalling the abbreviation given recall of the referent object or its name.
Command Interfaces and the Hierarchy of Needs BIBA 301-309
  Julian Newman
Effective office automation must support integration between different areas of application, enhance the users' understanding of the system and its potential, and encourage them to develop their own applications on the basis of that understanding. Thus the design of the interface cannot assume a fixed prior classification of users and of applications with certain users being adequately served by hiding most of the facilities behind a restricted menu.
   The degree and kind of protection to be afforded to users requires careful attention, if the motivation for user growth is be sustained. Human beings have security needs, implying requirements for protection, but also needs to explore and to realise their potential, implying requirements for freedom. At different times, users will require a greater or lesser degree of protection.
   The Unix concept of a command language as a set of software tools, provides the elements from which can be developed an interface that allows for user growth albeit that Unix as it stands is perceived as user-friendly only by professional programmers. This paper outlines the development of a set of Unix commands for office applications, designed to allow the user to gain confidence in the use of the system. Illustrative examples are given of problems in the original Unix commands and of the solutions implemented.
Recall as an Indicant of Performance in Interactive Systems BIBA 311-315
  Allan MacLean; Phil Barnard; Nick Hammond
Recall measures are often used in the area of human computer communication as a quick means of obtaining an index of the 'goodness' of alternative command sets. However there is a rich assortment of additional information available to mediate use of an on-line system, which is absent in conditions under which recall is typically elicited. The present paper reviews a number of experiments in which both on-line performance and recall measures are available, with a view to determining the extent to which recall can be used to explore the user's representation of the computer system in interactive performance. In addition, it relates the phenomena observed to established findings from the psychological study of memory.

Dialogue Interfaces / Menu Interfaces

Making the Right Choices with Menus BIBA 317-321
  Gary Perlman
Menus provide an effective way to present a limited set of options to users. System designers have to decide how many options to present in what format, and how users will indicate their choices. Two experiments are reported that manipulate (1) menu size, (2) option ordering, (3) option selector type, and (4) selector/option compatibility. The results show (a) people use simple search strategies for ordinary menu sizes, (b) people are sensitive to menu length, (c) sorted menus are easier to search, and (d) letter selectors can produce the best or worst performance depending on compatibility. Some guidelines for menu design and suggestions for further research are discussed.
A Comparative Evaluation of Menu-Based Interactive Human-Computer Dialogue Techniques BIBA 323-328
  M. D. Apperley; G. E. Field
Menu selection is an often used type of human-computer dialogue. However, there is little data on the effectiveness, efficiency or merits of this technique. This paper describes an experiment designed to compare the utility of recently published intuitively derived techniques relating to the syntax of menu interaction with more conventional menu techniques.
   This experiment is a complex problem solving task which involves retrieving a number of related items from a data-base using a menu dialogue. Subjects will be presented with a task for which they must access, interpret and relate information from several different pages of a Viewdata-like database. Traversal paths and time taken to achieve the goal, are monitored, to provide data with which to assess the effectiveness of the menu syntax.
   It is anticipated that these results will be of significant interest to the designers of Viewdata systems, and to all people interested in human-computer interaction.
Graphical Support for Dialogue Transparency BIBA 329-333
  J. Kaster; H. Widdel
The development of a dialogue interface is described and general problems of a design process are discussed. Principles of cognitive ergonomy lead to the graphical representation of the menu hierarchy of the dialogue as realisation of its formal transparency. The experimental set-up for investigating the influence of graphically displayed dialogue structure on human-computer interaction of unexperienced users is presented.

Dialogue Interfaces / Adaptive and Flexible Interfaces

MONITOR. A Self-Adaptive User Interface BIBA 335-341
  David Benyon
The design of a human-computer dialogue is widely recognised as being difficult, as it includes principles of graphics and information presentation underpinned by psychological factors such as closure and control over the system. This realisation has prompted the call for adaptive or self-adaptive user interfaces. Such systems need to maintain a model of the user and the dialogue so that the dialogue can be altered as users develop their skills.
   MONITOR has been designed to provide a self-adaptive user interface. The system is flexible enough to cater for any dialogue, and a prototype system in the area of computer aided learning has been implemented. The prototype is seen as a contribution to the collection of research tools which are needed if a useable system is to be developed. The system is still unsophisticated in much of its operation but the design has proved itself to be a suitable and flexible representation of any problem which can be analysed along the user-task dimensions, and the feasibility of the system has been established.
Adaptive Interfaces for Naive Users -- An Experimental Study BIBA 343-349
  H. S. Maskery
An experimental study is described in which eighteen naive users used a developmental adaptive interface. The subjects were divided into three groups who initially used a statistical tool, either daily, weekly, or with a six weeks interval. A number of the subjects progressed to using a graph-plotting tool. Conclusions included (1) that weekly usage promoted better learning than daily usage, (2) that a break of five and six weeks led to an initial decrement in performance followed by rapid improvement, and (3) the change between dialogue styles must be predictable and consistent with the system's previous behaviour.
Use of Path Algebras in an Interactive Adaptive Dialogue System BIBA 351-354
  J. L. Alty
The CONNECT system-an adaptable dialogue delivery vehicle -is briefly described. The system casts interactive dialogues in the form of a transition network. Adaptability is achieved through a production system which monitors user transitions in the network and thereby enables or disables transitions in the network. A multilayered network representation technique is described and an example of use providing access to the CP/M operating system is outlined. Some preliminary user experience of automatic adaptation is reported. The importance of the path algebra approach is emphasised.
Personalising the Software Interface BIBA 355-361
  Roberto Sasso
Humans communicate with (application) programs through a software (or logical) interface. We believe there should be a common framework shared by the interfaces of these interactive systems. A model of interaction is presented together with a formal specification using a notation based on set-theory. The implementation of this model allows users to interact with their system in a uniform and consistent way not only to make use of the facilities provided but also to "personalise it", that is, to redefine the external appearance of the system in order to suit their personal needs and desires.
The Design and Implementation of Flexible Interfaces: A Case Study of an Operational System for Assisting the Diagnosis of Cerebral Disease BIBA 363-367
  P. R. Innocent; D. Plummer; D. Teather; B. A. Morton; K. M. Wills; G. H. du Boulay
The design of flexible user interfaces is discussed. An interface system for constructing menu type dialogues is described. The interface system has been utilised in the building of an extensive applications system to assist in the radiological diagnosis of cerebral disease.

Tools to Aid Interface Design and Programming

An Attempt to Ease and Simplify the Development of Interactive Software: The Menu Support Subsystem of INTERPRO BIBA 371-374
  Dimiter Novatchev; Yuly Gabrovsky
A general method for dialog organisation, supported by the Menu Support Subsystem -- MMS of an interactive environment INTERPRO, is described. The most distinguishing MMS features are pointed out. The Menu Description and Interpretation Language -- MDIL and the Menu Interpreter -- MINT are briefly examined, their use being demonstrated by a practical example. The authors argue that MMS can be regarded as a software engineering tool, aiding the design and partly the development of prototypes for interactive programs. They outline some basic directions for future work.
The SYNICS2 User Interface Manager BIBA 375-378
  Ernest Edmonds; Stephen Guest
SYNICS2 is a user interface mangement system. It is the latest development of the family of SYNICS systems and incorporates graphics and logging facilities. The paper provides a brief introduction to the concepts of the new system.
A Human-Computer Dialogue Management System BIBA 379-383
  H. Rex Hartson; Deborah H. Johnson; Roger W. Ehrich
The Dialogue Management System (DMS) of Virginia Tech is a comprehensive system for designing, implementing, testing, and maintaining interactive software systems with human-computer interfaces. A concept called dialogue independence, in which the dialogue component of a system is separate from the computational component, forms the fundamental philosophy of DMS. Dialogue independence is manifest at the highest level of DMS in three ways: by providing a comprehensive design methodology for the production of interactive systems, by providing separate sets of automated tools with which the dialogue and computational components are implemented, and by providing separate execution environments for each of the components.
An Improved User Interface for PROLOG BIBA 385-389
  Marc Eisenstadt; Tony Hasemer; Frank Kriwaczek
We have developed a series of prototype software environments for both beginning and advanced Prolog users in an attempt to provide a consistent and powerful user interface which is easy to understand, easy to use, and incorporates the best features of several different dialects of Prolog. Our user interface is characterised by Prolog-specific editing and debugging tools, and the ability of the user to interact directly with the internal representation of the database in a manner which is intuitive for beginners yet provides sophisticated debugging capabilities for experts. Prototypes currently exist for most portions of the system, and run under Prolog-10, POPLOG, and micro-PROLOG on a variety of machines.
A Simple User Interface for Interactive Program Verification BIBA 391-395
  R. J. R. Back; P. Hietala
A prototype system for interactive program verification is described, with special emphasis on its user interface. The system supports incremental and iterative verification of programs, employing a spread sheet like paradigm of direct manipulation.
The Programmer's Torch BIBA 397-402
  T. R. G. Green; A. J. Cornah
The Programmer's Torch is a projected software tool designed to illuminate the workings of programs by answering maintenance-like questions about data flow and the 'roles' of identifiers. We describe the novel methods of analysis used and the current state of the project.
A User Friendly Editor for the PLATO System for FAULT Troubleshooting Simulations BIBA 403-407
  Henry L. Taylor; William C. Entwistle; Charles F., Jr. Ziegler
A system editor was developed for the PLATO (Programmed Logic for Automated Teaching Operations) computer system to permit troubleshooting simulations to be created and edited. A user guide describes the system editor capabilities and discusses the editing options. To evaluate the user friendliness of the system editor, eleven subjects entered a troubleshooting simulation on PLATO. All subjects completed the task with some assistance from the experimenter. The subjects' responses to questions concerning the user friendliness of the editor were very favorable. They found the editor easy to use to enter the simulation and to correct mistakes. The subjects reported that the task of entering the simulation on PLATO was not difficult.
Design Considerations of an Intelligent Teaching System for Programming Languages BIBA 409-413
  Mark Elsom-Cook
The purpose of this research is to investigate the nature of the teaching interaction. A system intended to guide a discovery learning interaction about the programming language LISP is described. Representation of domain knowledge, skills of interaction and knowledge about the pupil are discussed.
Designing and Evaluating Computer Aids in Structured Programming BIBA 415-420
  J. M. Hoc; J. Guyard; M. Quere; J. P. Jacquot
Presentation of the first stages of research work aimed at developing and evaluating a design language for structured programs and software aids, intended for experienced programmers.
   In order to clarify some of the problems posed by programmer training, one of the targets of this research is to determine the variety of programming strategies employed by professional programmers.
Transferring Users' Responsibilities to a System: The Information Management Computing Environment BIBA 421-426
  Robert Neches; Bob Balzer; Neil Goldman; David Wile
User difficulties in developing and using useful system model appear to depend both on the number and complexity of rules they must learn, and the memory load of building and applying the required set of rules. The information Management computing environment, by presenting several non-traditional capabilities based on a fusion of AI and database technology, may help ameliorate some of these difficulties. The key concepts of the system consist of a uniform underlying database representing all information in the environment, support for retrieval of information objects by description, and the explicit specification of rules telling the system how to take over responsibility for consistency maintenance and routine user actions.
Computer-Supported Program Documentation Systems BIBA 427-434
  Gerhard Fischer; Matthias Schneider
One of the most neglected research areas in computer system development is how to produce effective materials and reference information. Beyond the design principles developed for printed materials, there are documentation opportunities unique to interactive systems that we do not yet understand how to exploit effectively.
   Program documentation systems with a high-bandwidth user interface in connection with program analysis systems are necessary tools for program designers and users because they allow the monitoring of both the program and its underlying design principles and ideas.

Language Design and Comprehension

The Analysis and Understanding of an Operative Language BIBA 437-441
  Pierre Falzon
When the operators of a system have to communicate verbally, they tend to build operative languages, molded by the characteristics of the task and its objective. A better knowledge of these languages could provide guidelines for the design of computer command languages. A method of analysis of such an operative language is presented, based on schema theory. Within each category of command messages the different expressions are considered as a collection of different instances of a single underlying schema. Their analysis provides a description of the schema, and allows the elaboration of a dictionary of words. Given the knowledge of the schemata, the understanding process can rely on a very limited syntax and on a very small dictionary in which all words are monosemous. This hypothesis has been tested by implementing these principles in computer programs. Results of an evaluation of the programs are presented and discussed.
Transformations of Software Design and Code May Lead to Reduced Errors BIBA 443-448
  Edward M. Connelly
This research investigated the capability of programmers and non-programmers to specify problem solutions by developing example-solutions and also for the programmers by writing computer programs; each method of specification was accomplished at various levels of problem-complexity. The results, which showed the superiority of using example-solutions with inductive feedback over writing code, suggests that the transformation process provided by the induction might be applied analogously to software development. Considering designs and code in multiple transformed forms may reduce software errors to a level found for example-solutions.
Flip and Luciflip: A "Tree" Programming Language and a "Tree" Editor for Introducing Novices to Structured Programming BIBA 449-454
  Alain Giboin; Alain Michard
Two empirical studies are presented which show that the simultaneous use of a small programming language, called Flip, and of its syntax editor, called Luciflip, by inciting the handling of known metaphor-the tree diagram, constitutes a meaningful and refined environment to introduce novices to the basic notions of structured programming (mainly, stepwise refinement [or] top-down development, and modular development): the syntax editor allowing to ignore programming syntactic details and to focus on semantic features of programming which are the most important.
A Cognitive Account of 'Natural' Looping Constructs BIBA 455-459
  Marc Eisenstadt; Joost Breuker; Rick Evertsz
We describe an investigation of the everyday skills which underlie a person's ability to write iterative programs. We look at the strategies used by naive and experienced programmers to perform repeated processing in both 'natural' and 'computer-like' settings, and develop a production system simulation of our subjects' performance. Our data suggest an overwhelming preference on the part of naive and experienced programmers alike to think in terms of function application to aggregate data objects rather than control flow and temporal sequence. Our model explains why this is so, and demonstrates that the various components of everyday knowledge which enable everyone to perform our tasks can be combined in systematic ways to account for some of the difficulties encountered by students of programming.
The Comprehensibility of Programming Notations BIBA 461-464
  D. J. Gilmore; T. R. G. Green
Empirical comparisons of the comprehension of four miniature programs were used to contrast the notational structures of four programming languages, concluding that an important determiner of comprehensibility is the match between notational structure and the task being performed by the programmer. None of the four languages was consistently the most comprehensible: the procedural notations (e.g. Pascal-like) were best suited to answering sequential questions and the declarative notations (e.g. production systems) to answering circumstantial questions. Software complexity metrics, like many models of program comprehension, do not allow for such structural and contextual effects and their validity is brought into question.
The Nature of Expertise in UNIX BIBA 465-471
  Stephen W. Draper
This paper discusses the nature of expertise in Unix, arguing that in certain senses of the word there are no experts. The consequences for interface design of revising the common-sense notion of expertise, particularly with respect to designing help facilities, are then discussed.
Novice-Expert Differences in Software Design BIBA 473-478
  B. Adelson; D. Littman; K. Ehrlich; J. Black; E. Soloway
In this paper we describe the results of analyzing protocols of expert and novice software designers as they performed a novel, non-trivial design task from a domain with which they were familiar. The protocols allowed us to develop a model which can account for several interesting and recurrent expert behaviors such as constraint gathering, balanced development, and the building and running of mental simulations of partially completed designs. We have also found what look like systematic differences between our novices and our experts.

Knowledge Based Techniques

Application Domain Modelling by Knowledge Engineering Techniques BIBA 481-488
  Wolfgang Dzida; Wilhelm Valder
From an ergonomist's point of view it is important to understand the tasks and objectives a user is concerned with. Before a systems designer starts developing an application system his understanding of the application case can be facilitated by means of a knowledge base. To provide this knowledge an ergonomist may help, since work-analysis is his professional research field. This paper is intended to illustrate how work-analysis and user participation may contribute to the development of applications systems so that some ergonomic design principles may be considered during the software development process. Our approach is not focussed on modelling application domains in manufacturing but in business context.
Knowledge Engineering for Expert Systems BIBA 489-493
  Mildred L. G. Shaw
The Japanese fifth generation computer development program is targeted on knowledge processing rather than information processing. Expert system developments have demonstrated that knowledge can be represented and processed to replicate human skills, but have also focused attention on the difficulties of knowledge engineering, of eliciting and representing human knowledge and its processing. This paper presents a system for interactive elicitation of expert knowledge that is based on a systemic psychological model of human knowledge acquisition and processing. The need for normative tests of such methodologies is emphasized and an experimental study studies on reproducing a standard business record-keeping methodology is described.
Maps for Computer Wayfinding BIBA 495-498
  Ila J. Elson
The purpose and design of a knowledge based system is described and compared to previous styles and methods for helping computer users. This expert system answers a user's questions about the various software entities stored in his computing environment.
Tasks, Skills and Knowledge: Task Analysis for Knowledge Based Descriptions BIBA 499-503
  Peter Johnson; Dan Diaper; John Long
A method for deriving descriptions of knowledge from tasks is described. Knowledge descriptions constitute the basis of a syllabus specifying the training requirements of Information Technology (IT). Task analysis for Knowledge Descriptions (TAKD) is a method which is first used to generate descriptions of tasks, and then to reexpress the descriptions in terms of knowledge. The resulting knowledge descriptions consist of action/object pairs that when combined represent the knowledge content of tasks. The potential application of TAKD to other design problems is discussed and in particular to the design of the Human-Computer Interface and Intelligent Knowledge Based Systems.

Modelling Users and User Interactions

Four Stages of User Activities BIBA 507-511
  Donald A. Norman
When a person interacts with a computer it is possible to identify four distinct stages in that interaction: Intention, Selection, Execution, and Evaluation. Each stage has different goals, different methods, and different needs. It is well-known that the task and class of user affects the requirements for an interface. In this paper I emphasize that these requirements vary according to the stage of the interaction, even for an individual user working on a single task. The analysis of these four stages shows that different support is required at different times within an interactive session.
Planning Nets: A Framework for Analyzing User-Computer Interactions BIBA 513-518
  Mary Riley; Claire O'Malley
During the course of interacting with a computer, a user has goals that correspond to tasks to be performed and must plan how to achieve those goals with the available commands. We present a framework for analyzing user goals, the mapping between those goals and available commands, and the factors influencing the success and efficiency of the resulting plans. We discuss the implications of our analysis for the development of principles for improving user-computer interactions.
Predicting Expert Slips BIB 519-525
  T. R. G. Green; S. J. Payne; D. J. Gilmore; M. Mepham
Task-Action Grammars BIB 527-532
  Stephen J. Payne
Representing the User's Model of an Interactive System BIBA 533-537
  Marion Wells
With the very rapid increase in the development of interactive systems the traditional data flow diagrams produced using structured methods are becoming less than adequate as a basis for obtaining user acceptance of a proposed software system. Enhancement of such diagrams with screen or report layouts and input forms, can only hope to present static models of the proposed system. The user's conceptual model of the dynamic nature of the system should form an essential ingredient of the design process, and to this end a notation has been devised which can be directly integrated with traditional data flow diagrams. The paper describes the notation and its application in a dialogue prototyper.
On the Modelling of Human-Computer Interaction as the Interface Between the User's Work Activity and the Information System BIBA 539-546
  Juhani Iivari; Erkki Koskela
This paper analyses human-computer interaction and its design as an integrated part of an information system design methodology, paying principal attention to its role as the interface between the user's work activity and the information/data system. It is also pointed out that due to the great number of factors influencing human-computer interaction design, this should be layered in a way which is consistent with the whole information system design process.
An Operationalized Model for Success in the User Role BIBA 547-554
  Jouni Simila; Risto Nuutinen
The paper presents a conceptual model for success in the user role and operationalizes it in a longitudinal case study of adp systems implementation in a large Finnish chemical enterprise. The model is based on a general conceptual SROE -- subject, role, object, and action environment -- framework. An analysis is made of the role and systems hierarchy associated with the use of data systems in order to identify the relevant role, object and action environment combinations. Success in the user role is defined to consist of three conceptual components: (1) the degree to which the data system satisfies the information needs associated with the user role, (2) the individual's demands and desires towards acting as a user and his experiences of being a user, and (3) the organization's demands and desires towards the users and the corresponding experiences from the organizational point of view. A preliminary causal model for success in the user role is defined in terms of independent and dependent variables. The operationalization is presented in the context of the case study. The validation and generalizability of the model is briefly discussed.
A Model of the Engineering Design Process Derived from Hearsay-II BIBA 555-559
  Andy Whitefield
A model of the engineering design process is proposed. The framework for the model is derived from the Hearsay-II speech recognition program, and therefore consists of a set of knowledge sources communicating via a central blackboard. Changes to the framework for the purpose of this model are outlined. The content of the model comes from an analysis of the verbal protocols of four engineering designers. A total of twenty knowledge sources are described, with examples from the protocols. The uses of the model in comparing computer aided and unaided design, and in the design of CAD system interfaces, are discussed.

Design -- Approaches and Methods

Defining Information Technology Systems for Electricity Supply Distribution BIBA 563-569
  J. C. Gower; K. D. Eason
This paper demonstrates a methodology for the introduction of Information Technology to the Electricity supply industry.
   A socio-technical systems analysis is made of the engineering function of the District office of an Area Board. This is used to identify areas where the introduction and application of information technology could be beneficial and generalisable to other District offices and other Area boards.
   The development of a pilot system, using the stated needs and requirements of the potential users as a basis for the specification is also described.
System ABC: A Case Study in the Design and Evaluation of a Human-Computer Dialog BIBA 571-575
  Catherine R. Marshall
Human factor specialists concerned with the human-computer interface in two worlds. In the world of theory we are concerned with the properties of the ideal interface -- one that is easy to learn and use, and results in performance that is efficient and error-free. In the world of practice we must design working systems in the presence of many constraints. This paper presents a case study in the design and evaluation of a human-computer dialog in a constrained environment. It also discusses the relationship between theory and practice in interface design, with particular emphasis on the role of standards and guidelines.
Building a Usable Office Support System from Diverse Components BIBA 577-581
  Nancy C. Goodwin
Organizations interested in providing computer-based office support systems now have several choices. They can select an integrated system that uses new equipment or that uses equipment already in-house, or they can select components from diverse sources and build their own system. The latter approach enables an organization to introduce office systems without major disruption to existing services, but requires a strong awareness of user-oriented issues. Among these issues are user-system interface design, text editing, data transfer requirements and user support requirements. The success of the system will depend on the organization's commitment to resolving these issues.
Task and User Adequate Design of Man-Computer Interfaces in Production BIBA 583-588
  H.-J. Bullinger; K.-P. Fahnrich; C. Raether
The interface of the man-computer interaction for CNC-control has been designed according to software-ergonomical principles. The design was based upon a layered man-computer interface model and an overall interface design methodology. With the incorporation of an object oriented design and programming methodology, a large degree of freedom is won in the design of several interface parameters (e.g. colour coding). Thus, at a relatively early stage an evaluation of the man-computer interface can take place due to the rapid prototyping approach chosen.
Guidelines for User Participation in the System Development Process BIBA 589-595
  Bernard C. Glasson
Taking the need for effective user participation in computer-based information system development as agreed, this paper argues that there are several user roles associated with the system development process, each being required to participate to varying degrees at different stages of system evolution, each having several parts to play in the system development process, and each requiring detailed advice on how they might best participate. The paper then suggests and illustrates how a detailed task-based set of guidelines for user participation relevant to each role could be derived from an organisation's technical approach.
Experiences on User Participation in the Development of a Conceptual Schema by Using a Concept Structure Interface BIBA 597-601
  Hannu Kangassalo; Pirjo Aalto
Conceptual modeling of the universe of discourse (UoD) has been gaining increasing attention as an efficient method in the development of information systems. An important part of a modeling process is the collection of information from which the conceptual model will be built up. This information consists of concept definitions and structural descriptions concerning the UoD. In this paper we will describe methods and tools used for the development of a conceptual schema in a life insurance company. We will also describe experiences on user participation in conceptual modeling.
   Two types of graphical notations were used: concept structure diagrams and conceptual schema diagrams. Finally we will describe a graphical workstation environment the development of which has been started on the basis of collected experiences.
Formalizing Task Descriptions for Command Specification and Documentation BIBA 603-609
  Paul Smolensky; Melissa L. Monty; Eileen Conway
We consider the problem of formally describing computer tasks not in terms of procedures that will accomplish them but rather in terms of the input given and the output desired. A feasibility study in the domain of printing suggests that task attributes provide a powerful language for such descriptions. We describe the constraints such attributes must satisfy, and the procedure we used to design the printing attributes and test their usability. Applications to attribute-oriented interfaces and documentation are discussed. It is argued that task description is important for moving the center of human-machine interface design away from the machine and towards the user.
Developing Interactive Information Systems with the User Software Engineering Methodology BIBA 611-617
  Anthony I. Wasserman
User Software Engineering is a methodology, supported by automated tools, for the systematic development of interactive information systems. The USE methodology gives particular attention to effective user involvement in the early stages of the software development process, concentrating on external design and the use of rapidly created and modified prototypes of the user interface. The key ideas and steps of the User Software Engineering (USE) methodology are described. The Unified Support Environment provides an integrated collection of tools to support the USE methodology.
An Approach to Information Needs Analysis BIBA 619-628
  Robert D. Galliers
"Information systems have value only if they contribute to improve the situation for people in the organization. They have no value of their own. It is therefore not enough that we study the contents of the information systems so that we can form an opinion about their values. We must instead study the activities that people perform in the organization and that somehow should be improved."
   Lundeberg et. al. [17, p.125]
   Systems development methodologies tend to be approached either from the human or the design aspects of the information system being developed. Regretably, more attention has apparently been focussed on design in the past. Regretably, because much evidence exists to suggest that the human considerations are of paramount importance.
   What is perhaps of greater concern is that many of those methodologies which purport to recognise the importance of the human considerations are misguided in their application of technique with the methodology. The focus of many of the techniques tends to be on the technical design of the existing system rather than on the information requirements of the human activity system which dictates the need for information. One could go further; in certain instances, no technique for determining information needs is provided at all.
   This paper concentrates its attention on a means by which information needs can be identified and compared with existing information provision, an approach which arises from the soft systems methodology developed by Checkland [6,8]. It relates to the very first steps to be taken prior to deciding on whether or not to proceed with an information system development. It describes in some detail a methodology which has been successfully applied in a range of actual studies with a view to identifying information needs as a precursor to design, and it relates some of the views expressed by those involved in the process. Further, it illustrates the way in which the approach can be used in the strategic sense of setting out a coherent plan for information systems development, organisation-wide, for a period of some years ahead.
Dialog Shell Design BIBA 629-634
  Brian R. Gaines; Mildred L. G. Shaw
Many rules have been proposed for dialog engineering effective human-computer interfaces. The underlying technology has been changing rapidly with the introduction of windows, icons, and natural language. It is not clear how coherent, complete and consistent are the various systems of rules, and how applicable they are to the new technologies. A systematic model of human protocols is needed where the principles and technology dependencies are clearly expressed. There is a need also for such protocols to be made available as application-independent processes, dialog shells, that implement effective human protocols. This paper gives a systematic exposition of the principles of dialog engineering, shows how these structure effective human protocols in different dialog technologies, and how this leads to the design of dialog shells.

Design -- Guidelines

The User Interface to Computer-Based Information Systems: A Survey of Current Software Design Practice BIBA 637-641
  Sidney L. Smith; Jane N. Mosier
From a survey of 201 people concerned with information system design, estimates for 83 systems indicate that on average 30-35 percent of operational software is devoted to the user-system interface (USI). In the design of USI software, survey responses indicate that improvements are needed in requirements definition, design documentation, and design guidelines.
Designing Interfaces for Different Types of Users - Experimental Criteria BIBA 643-647
  L. A. Macaulay; M. A. Norman
The nature of the user interface should be adaptive to accommodate variations in user skill levels and cognitive style. The paper includes a brief review of current research on dialogue design, and describes aspects of an experiment carried out into the use of functional simplicity in dialogue design. The experimental session is described briefly and results based on keystroke timings are reported upon. Further findings from the experiments are reported upon elsewhere, including [16].
Human Factors Guidelines for the Design of Computer-Based Systems BIBA 649-653
  Arthur Gardner; Tom Mayfield; Martin Maguire
The HUSAT Research Centre has been awarded a four year contract by the U.K. Ministry of Defence (Procurement Executive) to produce a new handbook of human factors design guidelines and associated training for use by the Royal Navy and Industry. The paper describes the project's aims and plans, reports progress and seeks to stimulate discussion by presenting an outline of the handbook's proposed structure and content.
The "Impact Analysis Table" Applied to Human Factors Design BIBA 655-659
  Tom Gilb
The "Impact Analysis Table" is a tool for multiple viewpoint analysis of any system design idea. It is used to estimate the impact of suggested design techniques on the target levels of multiple design objectives.
   On a single table, multiple techniques, or groups of techniques can be evaluated against multiple objectives.
   Using a series of tables, all suggested techniques can be evaluated against all design objectives.
Generative User-Engineering Principles for User Interface Design BIBA 661-666
  Harold Thimbleby
Generative user-engineering principles are assertions about interactive system behaviour and have equivalent colloquial forms. Current work shows that they are a promising contribution to the design of acceptable user interfaces, because they effectively bridge the conceptual gap between designer and user. In colloquial form a generative user-engineering principle can be used to help clarify requirements in participative design, or to explicate documentation. In rigorous form, generative user-engineering principles provide a constructive higher order consistency on user interfaces.
Empirical Guidelines and a Model for Writing Computer Documentation BIBA 667-671
  Darlene Clement
A model of the computer manual comprehension task is proposed in which four processes operate simultaneously: task-mapping of the structure of regular procedures onto the structure of computer commands, constructing a mental model of the computer system, inducing the command language grammar, and learning the structure of computer procedures. Findings from a study of five novices' comprehension problems with UNIX documentation are analyzed in terms of these four processes. The model and the findings from the study yield general heuristics and specific recommendations for document developers.

Evaluation -- Approaches and Methods

A Human Factors Evaluation of the IBM Sheffield Primary Care System BIBA 675-681
  Mike Fitter; Garry Brownbridge; Bob Garber; Guy Herzmark
The IBM Sheffield Primary Care System was an experimental system developed by the IBM UK Scientific Centre in co-operation with Sheffield University Medical School. It was installed in two General Practices for a trial two year period and evaluated by the Social and Applied Psychology Unit.
   In this paper we report on our evaluation objectives and techniques which focus on the 'human factors' aspects of designing and implementing a complex and comprehensive information system in an unfamiliar and relatively unstructured environment. We argue that an effective way of learning from the project is to focus on difficulties that arose and were experienced by users. We conclude that although certain specific problems resulting from the design of the user interface did occur, these were relatively minor, and in this respect the system was well liked. The most significant issues arose from the organisational impact and the mismatch between the system and the user organisations ie the 'organisational interface'.
Evaluation of Interactive Audiovisual Applications: Some Results and Perspectives BIBA 683-688
  Francis Kretz
News interactive media such as interactive audiovideography and videodisc lead to applications which combine audiovisual presentation and interactivity. Sound and audiovisual sequences much increase the creative capabilities of the media as compared to videotex. They tend to produce a "spectacle" attitude on the user who will also be asked through interactivity for an active motivation. More generally, results from the evaluation of four experimental applications will be presented, using a structured analysis schema to organize them.
It Is What It's Used For -- Job Perception and System Evaluation BIBA 689-692
  Oleg de Bachtin
A study involving very experienced people representing two different professions is presented. The data from this study indicates that such people regard the computer systems as mere tools for their jobs. If the tools are in agreement with the mental model a person has of her or his job, job satisfaction will increase. If not, the tools are rejected.
   Factors, commonly regarded as "good", like "ease of use" or "ease of learning", could have a negative influence if they are perceived as a threat to ones professional role. It is argued that developers of computer systems must consider the users mental job models when building their systems. They should also use the knowledge from the field of cognitive psychology in doing this.
Workstations Using Direct Manipulation as Interaction Mode - Aspects of Design, Application and Evaluation BIBA 693-698
  K.-P. Fahnrich; J. Ziegler
Some more recent developments of workstations have been using a new mode of man-computer interaction. As a term for this mode "Direct Manipulation" has been introduced in the literature. A characterization of direct manipulation is given on the basis of a suggested general man-computer interface model. The concept of "generic interaction modes" is introduced and discussed, where programming languages, direct manipulation and natural language are suggested to form generic interaction modes. Some work on the evaluation of an interface based on direct manipulation (Xerox STAR) is reported.
Randomly Sampled Self-Report Method for Collecting Field Data on Human-Computer Interactions BIBA 699-702
  Paul D. Tynan
A variant of traditional work-sampling methods, the randomly sampled self-report, has some advantages over other methods of collecting field data. These advantages are especially important when studying information-processing environments. Moreover, the method can improve collection of subjective data. The history, application, and validity of the method is discussed.
User Acceptance of Information Technology Through Prototyping BIBA 703-708
  Frances Clark; Peter Drake; Marco Kapp; Peter Wong
Prototyping is a method of complex systems design which is receiving more attention, as the tools and techniques which are now available have made the process more cost-effective.
   This paper provides some theoretical perspectives on prototyping and suggests why a need for this approach exists. A working definition is proposed and an indication given of the types of situations where prototyping is most appropriate.
   Next, two examples of successful prototyping are described, one in a military environment, the other in a commercial setting. Finally, some guidelines are offered, based on practical experience.
Problems in Evaluation of Human-Computer Interfaces: A Case Study BIBA 709-713
  Liam Bannon; Claire O'Malley
One of the most difficult aspects of interface design is evaluating new or changed features of an interface. In this paper we discuss methods of evaluation, their strengths and weaknesses, in the context of a program we developed to assist users in getting quick access to information contained in the UNIX manual. We outline the problems encountered both in the design and the evaluation of this user interface.
Human Aspects of Office Systems - User Acceptance Research Results BIBA 715-718
  Reinhard Helmreich
The vision that the "office of the future" conjures up is indeed an impressive sight: computer power at every office desk, within easy reach of every white-collar worker, electronic systems permitting worldwide communication by text, images and voice. But what real benefits can be expected from this new technology and how will they be exploited? Will the user be aware of the advantages technology has to offer? And how should advanced office systems be designed and organized if they are to gain acceptance?
A Computer-Based Tool for Evaluating Alphanumeric Displays BIBA 719-723
  Thomas S. Tullis
A computer program has been developed to measure six characteristics of alphanumeric displays: (1) the overall density of characters on the display; (2) the local density of other characters near each character; (3) the number of distinct groups of characters; (4) the average visual angle subtended by those groups; (5) the number of distinct labels or data items; (6) the average uncertainty of the positions of the items on the display. A study of 520 CRT displays that varied on these measures was conducted. Multiple regressions indicated that search times to locate items on the displays could be fit using these display measures (R = .71), as could subjective ratings of ease of use (R = .90).
Evaluating the Interface of a Document Processor: A Comparison of Expert Judgement and User Observation BIBA 725-729
  N. Hammond; G. Hinton; P. Barnard; A. MacLean; J. Long; A. Whitefield
Efforts to improve the usability of systems have resulted in the development of several techniques for interface evaluation. This paper explores evaluation through (1) assessment by Human Factors researchers and (2) analysis of user performance. Three pairs of researchers prepared reports on the interface of a document processor. Separately, five novice users were observed learning the system. The two evaluations generated overlapping but separable classes of information. User testing provided low-level information on procedural and conceptual difficulties, while experts provided a more integrated overview and hypotheses concerning the sources of problems.
The Cognitive Regulation of Human Action as a Guideline for Evaluating the Man-Computer Dialogue BIBA 731-735
  Michael Paetau
Human Factors Research is changing. In the fist phase systems designers defined and evaluated userfriendliness themselves, based on their own experiences and common sense. In the second phase computer scientists discussed the term "userfriendliness", as a recognized problem, and finally in the third phase human factor research is turning now to a more empirically oriented science. Due to the increasing problems with poorly designed interfaces, especially in application fields where users are non-computer-specialists, different institutions of computer science have started more empirical investigations about human-computer interaction. The Man-Machine Communication Research Group of GMD (Bonn) is engaged in controlled experiments. The GMD is presently building up a "Human-Computer Interaction Laboratory" in which selected elements of userfriendliness can be created in an rapid prototyping process and evaluated with social scientific methods. The general guideline of this research is the cognitive regulation of human action. On this theoretical basis we can derive the main criteria of software ergonomics and develop the stage for evaluating them.
Adapting a Psychophysical Method to Measure Performance and Preference Tradeoffs in Human-Computer Interaction BIBA 737-741
  Jonathan Grudin; Allan MacLean
An experimental methodology for contrasting certain design alternatives and quickly determining user preferences and performance tradeoffs is presented. It is shown how this experimental paradigm, used for psychophysical measurement, may be applied to the field of human-computer interaction. Where it can be applied, it promises a relatively quick determination of user preference and performance characteristics and tradeoffs on these measures with variation in parameters governing the user situation. Because the methodology is within-subject, it may also facilitate the study of individual differences.
The Iterative Development of Usable Computer Interfaces BIBA 743-748
  Kevin F. Bury
Of all the tools and techniques now available to aid in the design of usable computer interfaces, human-factors testing is by far the most comprehensive and reliable. For testing to be effective, however, it must be done early in the development process. This paper describes how user-interface prototypes can be used as vehicles for performing human-factors tests before actual product code exists. A specific type of test, namely iterative testing is described. An example will be given of how iterative testing with a prototype was used to evolve the usability of an interactive programming-language product. Lastly, this paper will discuss how other interface design tools can be used in conjunction with prototypes and iterative tests.

Learning and Training

A User-Interface for Teaching Piano Keyboard Techniques BIBA 751-758
  Martin Lamb; Veronica Buckley
Those aspects of piano technique which can best be taught through computer-aided instruction are identified. Two classes of programs to provide this instructional aid are described. The first provides real-time visual feedback which analyses rhythm and relative loudness. The second provides augmented visual and auditory feedback after the performance in which the computer functions as a 'magnifying glass' for the ear, by revealing details crucial to the music, but which are difficult to perceive. This then allows the pianist to analyse his own playing with respect to articulation ornamentation and rhythm, for example. Examples are presented in which the above aids are used to improve the playing of students and professional pianists, as well as to reveal a number of paradoxes contained in musical performances of the highest calibre.
Preparing Hospital Staff for the Changeover to Computerised Records BIBA 759-763
  M. J. Barber; J. W. Kempson
This paper describes the preparation, installation and implementation of a computerised medical records system in the West Midlands Health Region, with special reference to its introduction into the Nuneaton Hospitals of North Warwickshire Health Authority. This computer system is being run by hospital staff with no day-to-day assistance from specialist computer personnel, with the central computer routine operations being carried out by the hospital Medical Records Department.
Providing Online Assistance to Inexperienced Computer Users BIBA 765-769
  Robert C. Williges; Jay Elkerton; James A. Pittman; Andrew M. Cohill
Inexperienced users of interactive computer systems often need online assistance to complete their task successfully. The results of several research studies are reviewed as a means of specifying the human-factors design considerations for online assistance. Alternatives such as automatic error detection, browsing and comparison of help facilities, and online expert aiding need to be considered. It was concluded that these features need to be incorporated into an adaptive interface tailored to the specific needs of the inexperienced user of the software interface.
Learning Pascal after BASIC BIBA 771-775
  John D'Arcy
This study compares the learning of Pascal by University students who either had previous training in BASIC with those who had no programming experience at all. Pascal program comprehension and modification tasks were used to see if those who had initially learned BASIC suffered the 'mental mutilation' suggested by computing and education experts. Data gathered during coursework show little evidence, given a well structured course, of the anticipated negative transfer of BASIC onto subsequent learning of Pascal.
How Novices Learn to Program BIBA 777-783
  Ann Jones
This paper describes a study of the behaviour of novice programmers. A conceptual model is advocated as a framework for novices learning to program, du Boulay, O'Shea and Monk (1981), and this research is concerned with the detailed learning processes of novices learning with the provision of one such conceptual model. It presents some new results supported by protocol data of students learning SOLO, a data-base manipulation language designed especially for novices, (Eisenstadt, 1978). In particular this paper reports on learning about control statements and the hierarchical structuring of programming. The ultimate goal of this research is to improve instructional materials for teaching at a distance and the paper concludes with some of the implications of this research for teaching novices to program.
Simulators which Invite Users into Learning Conversations BIBA 785-793
  Sheila Harri-Augstein; Laurie F. Thomas
The development of a range of new learning aids for complex computer-driven learning systems is described. An Air Intercept Control Skills Trainer is used to demonstrate the theory and technology of "Learning Conversations". This applies to the process of learning as well as to its content. The person-centred conversation paradigm of research accepts learners as active collaborators who are modelling, developing, reflecting upon and reviewing their learning skills. This requires freedom to structure activities, choose personal styles of execution and for evaluation of performance. The evidence reported demonstrates the emergence of a capacity for self-organised learning.
The Use of a Colour Graphics Display and Touch Screen to Help Naive Users Understand and Control a Multi-Function Computer System BIBA 795-799
  D. E. Penna
The work described in this paper is aimed at the production of computer based systems which are understandable and controllable by computer-naive people. All options open to the user are displayed diagrammatically on a colour graphics display and a pointing device such as a touch screen or graphics tablet used for selection. A tree structure of selection frames is used with short cuts to frequently used items at the lower levels. The control technique is designed to give a good idea of context and eliminate some common types of user error.
   This paper describes the design considerations and application of this control technique.

Aids for the Disabled

Videotex for the Blind: Design and Evaluation of Braille and Synthetic Speech Terminals BIBA 803-808
  R. W. King; N. Cope; O. R. Omotayo
Videotex and similar computer based information services cannot be used directly by blind people without special non-visual terminals. We describe the design of such terminals with dynamic Braille and synthetic speech output. The interface has two main processes -- reformatting of the Videotex source pages into linearised text -- and translation of this text into contracted Braille and synthetic speech. The controls for the displays are also important for overcoming some of the deficiencies of the two processes. Subjective tests are described which show that, on average, Braille is approximately twice as slow to use as normal visual access, whereas speech is approximately four times slower than visual access to Videotex.
Can Deaf Children Interact with Computers? Evidence from Speech Acquisition Training BIBA 809-813
  E. Gulian; F. Fallside; P. Hinds
The paper discusses the ways in which deaf children learn to use computer-based systems in order to gain control over their articulators and produce intelligible speech. Two systems are presented which analyse and display features of speech. The stages of the children's interaction with the systems and their use of the available feedback are described.
Prediction and Adaption in a Communication Aid for the Disabled BIBA 815-819
  J. A. Pickering; J. L. Arnott; J. G. Wolff; A. L. Swiffin
Many physically disabled persons suffer communication impairment as part of their handicap. It is possible to improve their effective communication rate in any language task, such as typing, by exploiting the redundancy of natural language. Many systems which do this are often equipped with fixed knowledge bases determined by 'a priori' statistical data about the language. This paper describes a system which exploits language redundancy in an adaptive, predictive aid for the disabled that has been specifically designed to be modular and thus have general purpose applications.

Organisation and Social Issues

Conceptual Data Modeling as an Obstacle for Organizational Decentralization BIBA 823-827
  Clas-Olof Kall
During the last decade a very large number of more or less different modeling approaches to conceptual data modeling have been presented. These research efforts seem to be based on a technological research tradition. Theories concerning use of conceptual data modeling approaches and of information systems by human actors in an organizational context are usually not considered. However, the conceptual data modeling strategy is not "neutral" in regard to organizational and behavioural aspects. In this paper one of the most characteristic features of conceptual data modeling, the integration of local views is analyzed by using a theory on information systems based on social sciences denoted the language action theory.
The Legitimacy of Information Systems Development -- A Need for Change Analysis BIBA 829-833
  Goran Goldkuhl; Annie Rostlinger
The decision on whether to develop computerized information systems or not must be made in a rational and transparent way. This kind of decision process is called change analysis. If a computerization decision is taken without a proper change analysis, then such a change action is not organizationally legitimate. A methodology for change analysis is presented consisting of problem analysis, goal analysis, change requirement analysis, change action determination and activity analysis. Some applications of and experiences from method use are noticed.
Microcomputer Installations in "Computer-Untouched" Environments BIB 835-838
  Laszlo Tolnai; Mihaly Csako
Training for Subjection or Participation BIBA 839-846
  Niels Bjorn-Andersen
The work role is undergoing substantial changes when new microelectronic systems are introduced in the office and the factory. Based on a recent study of changes in work role in eight Danish companies, the training offered for this future work role is discussed.
   Most training programs made available in these and other companies for users are oriented towards making the user appreciate and operate the systems. In other words, subjection to the system.
   However, alternative training programs could be designed in order to allow the users to participate in the design, i.e. to evaluate equipment and to maintain, modify and even design new systems. This is illustrated in a project within local government, which turned out to be very successful for the organization. The training program and some of the methods used are described and discussed.
Work Organisation Implications of Word Processing BIBA 847-853
  S. M. Pomfrett; C. W. Olphert; K. D. Eason
This paper describes a study which aimed a) to classify the forms of work organisation used for word processing and b) to evaluate responses of job holders and organisations towards the different forms. The paper concludes that the picture of the use of word processing is a complex one with more and more pluralistic forms of work organisation being employed. There does not appear to be a strong relationship between the form of work organisation and the job satisfaction of operators or the satisfaction of authors with the service they receive. Organisations which adopted a mixture of forms of working tended to be more satisfied with their word processing than those which used only 'small group' working arrangements; these in turn were more satisfied than organisations which used only word processing 'pools'.
Implementing Computer-Based Information Systems in Organisations: Issues and Strategies BIBA 855-860
  R. Hirschheim; F. Land; S. Smithson
The history of computer-based information system implementation in organisations points to the fact that implementation is not a simple nor straightforward process. It is very much based on human beliefs, emotions, perceptions, and the like. It has to be treated with care and effective strategies. This paper explores the nature of implementation -- how people react to change and why -- as well as how the reactions are often manifested in terms of computer-implementation. A consequentialist perspective is advocated as one way to help understand the basis for such reactions.

Behavioural Issues in the System Development Cycle / Introduction and Overview

Behavioral Issues in the System Development Cycle BIB 863-864
  John L. Bennett

Behavioural Issues in the System Development Cycle / System Examples and Deductions Therefrom

The Concept of Architecture Applied to User Interfaces in Interactive Computer Systems BIBA 865-870
  John L. Bennett
Work is emerging that will influence the evolution of the interfaces presented to users of computer systems. A central question is: "What abstractions from current specific systems are needed to support transfer of productive user habits as people adapt to new hardware and software technology?" An orderly evolution requires that users recognize similarity of control functions (e.g., select an object) even though the details of object presentation and of the way the user invokes the function are clearly different in different products. Managing such an evolution requires that we understand what must be held in common across products. Details of work in industry are proprietary; this brief paper outlines some of the problems that are being solved to make concept of a "user interface architecture" become a reality.
The Visi On Experience -- From Concept to Marketplace BIBA 871-875
  George H. Woodmansee
The Visi On system [1,2] is a personal computer software operating environment for business oriented application programs. It was developed to increase the effectiveness of personal computers in the office by providing an easy to learn, use, and remember problem solving tool for office professionals. This paper describes its development from the perspective of interface engineering in a small market driven company where competitive time pressures substantially shape the development process.
QMF Usability: How It Really Happened BIBAK 877-882
  James Boyle; William Ogden; Steven Uhlir; Patricia Wilson
QMF is a query and report generator product that can easily be used by non-programmers. An easy to use human interface was a primary design goal, and this paper describes how this goal was met by a combination of techniques. First, and probably foremost, the design team consisted of a group of software engineers, quality assurance specialists, and human factors engineers whose design effort was focused on the required information and actions that users would need in order to complete their tasks. Next, there was a commitment to obtain empirical input to the early design process by testing representative subjects using early prototypes. Controlled tests were very useful, especially when they came early in the development cycle and were broadly focused. These proved to be more useful than pre-planned, formal, paper evaluations. In general, iterative testing was always useful and should be a part of every interface design effort.
Keywords: QBE, Query by example, SQL, Structured query language
Human Factors Roles in Military Systems BIBA 883-888
  John B. Shafer
A series of vignettes are used to illustrate various Human Factors roles in the life-cycle of military systems. As programs progress through requirements, design, development, evaluation, selling, installation and training; the Human Factors roles flex and change to meet project demands. The successful human interface design, produced on schedule within budget constraints is the result of a great deal of creativity and resourcefulness.
Man-Machine Interface Design Process BIBA 889-894
  Gregory V. Kloster; Kristine Tischer
Many systems are built by selecting computers and displays, developing software, and only then retrofitting the human, his procedures, idiosyncracies and experiences into the design. Such retrofits are done almost as an afterthought resulting in systems in which the Man-Machine Interface (MMI) is re-engineered within the constraints of a fixed hardware system. To mitigate the risks associated with these approaches, a structured process is being used which focuses on an interdisciplinary and user-involved approach to better engineer the MMI. This approach presumes that man is a starting point in the design process.
Operations Concept Formulation for Next Generation Air Traffic Control Systems BIBA 895-900
  Mark D. Phillips; Kristine Tischer
A comprehensive requirements derivation and validation methodology was used to define the user-system interface for a proposed upgrade to the United States Air Traffic Control (ATC) System. This paper focuses on one step in the overall methodology -- the formulation of the Operations Concept for this system. The analyses were performed to provide functional descriptions of what the system would look like to the end user -- the air traffic controller. Cognitive and perceptual aspects of displays, viewability criteria, implicit information coding and presentation and interaction techniques are documented. As such, the Operations Concept formally records the allocation of functions between controllers and machines.
Note: in association with Valerio R. Hunt and Andres Zellweger
ZOG and the USS CARL VINSON: Lessons in System Development BIBA 901-906
  Robert M. Akscyn; Donald L. McCracken
This paper contains recommendations for developing computer systems for other organizations using emerging technologies. These recommendations are based on our experience developing a computer-assisted management system for the USS CARL VINSON, a nuclear-powered aircraft carrier, during the past four years. We recommend that such projects be conducted in a highly cooperative manner between the users' organization and the developers' organization over a planned period not longer than 18 months, obtaining feedback via a series of instrumented prototypes that are exercised by both users and developers for actual tasks.

Behavioural Issues in the System Development Cycle / Evaluation Issues

Instrumenting a Human-Computer Interface for Development and Evaluation BIBA 907-912
  Elise Yoder; Donald McCracken; Robert Akscyn
The ZOG human-computer interface has been instrumented to collect data about the system's performance and the users' behavior. We explain which data are collected and how they are recorded. We then suggest that analyzing the instrumentation data is akin to archaeology, because one must infer behavior patterns from low-level data "artifacts". Finally, we provide some guidelines for instrumentation design.
Augmenting Generic Research with Prototype Evaluation Experience in Applying Generic Research to Specific Products BIBA 913-917
  Thomas M. Gruenenfelder; William B., II Whitten
Generic research in the area of human-computer interaction is aimed towards discovering general principles of user interface design that can be applied to a wide variety of specific interfaces. In this paper, we point out some important limitations of generic research that make some results difficult to apply to specific designs. The most important limitation is the insensitivity of generic research to the context of a full design. We suggest guidelines for the designer to use when evaluating the applicability of generic research. Finally, we discuss new approaches to generic research that may help overcome these limitations.

Usage Issues in Electronic Mail, Conferencing and Journal Systems / Introduction

Usage Issues in Electronic Mail, Conference and Journal Systems: Introduction to the Theme of 4 Sessions BIB 921
  B. Shackel

Usage Issues in Electronic Mail, Conferencing and Journal Systems / Overview of Systems

Survey of Computer-Based Message Systems BIBA 923-928
  Jacob Palme
This paper provides a survey of computer-based mail and conference systems. The paper discusses systems for both individually addressed mail and group addressing through conferences and distribution lists. Various methods of structuring the text data base in existing systems is discussed and the networks of interconnected systems (ARPANET, CSNET, BITNET, USENET, JNT-MAIL, EURNET, MAILNET etc.) are described. The emerging standards for the interconnection of message systems are described.
The Computer Conferencing System KOMEX BIBA 929-930
  Uta Pankoke-Babatz
It was in 1979 that an initial configuration of the computer conferencing system KOMEX developed in the GMD was presented to a broader public at the Hannover Trade Fair. Since the beginning of 1982 KOMEX has been installed on a GMD computer network and has been used by GMD members as an in-house communication tool.
COM/PortaCOM Conference System: Design Goals and Principles BIBA 931-932
  Jacob Palme
The COM/PortaCOM computer conference system has been designed to be both easy to use for the novice user and powerful for the experienced user at the same time. This goal is achieved by having a powerful general-purpose basic data base structure, and by providing a user interface where the novice can understand the data base structure based on a simple model in the beginning, and then enhance this model with more experience with the system. The command structure combines menus and commands to suit both novice users and experienced users with the same user interface.
Software Infrastructure for the BLEND "Electronic Journal" Experiment BIBA 933-936
  W. P. Dodd; T. I. Maude; D. J. Pullinger; B. Shackel
The rational is discussed for the initial selection of software to support the BLEND electronic journal project; the preferred solution was the NOTEPAD computer teleconferencing system. The mapping of a journal format onto the NOTEPAD information structure is considered, as are a number of experimental extensions to NOTEPAD to provide additional facilities for the journal editors and referees.

Usage Issues in Electronic Mail, Conferencing and Journal Systems / User Experience and Usage Results

Experience with the COM Computer Conference System BIBA 937-939
  Jacob Palme
The main emphasis of the results of this study of the experience with the COM computer conference system is that computer conferencing is not mainly a replacement for neither face-to-face meetings, letters or telephone calls. Rather, it is a new communication medium, providing new modes of human interaction which to a large extent were not possible without this new medium.
Results from Evaluation Studies with the Computer Conferencing System KOMEX BIBA 941-944
  Uta Pankoke-Babatz
- technical and organisational facts that influence acceptance of CBMS
  • - appropriate long term training procedure with KOMEX
  • - results from the evaluation study
  • - main user groups and main tasks performed with KOMEX
  • User Surveys in the BLEND-LINC 'Electronic Journal' Project BIBA 945-950
      D. J. Pullinger; B. Shackel; W. P. Dodd; T. I. Maude
    This paper describes two telephone surveys of users in the 4 year experimental programme on electronic communication organised jointly by two Universities as the Birmingham and Loughborough Electronic Network Development (BLEND). Several communities of users are being studied; herein is described the first community of initially about 50 scientists (the Loughborough Information Network Community -- LINC). Considerable problems have been experienced with the hardware available to LINC members, with communications equipment, with modifying and developing software to obtain an acceptable operating system, and with various unexpected bureaucratic and organisational difficulties. Nevertheless, more than 50 papers are in the system and successful teleconferences have been held.

    Usage Issues in Electronic Mail, Conferencing and Journal Systems / Problems and Issues of Present and Future Systems

    Structures for Group Working in Mailbox Systems BIBA 951-958
      Paul Wilson
    Basic electronic mailbox facilities allow individuals to send and receive messages. However, for more formal work (eg running meetings or projects) structured facilities are required. This paper provides a generic categorisation of mailbox structures and identifies those currently being used in 5 categories of mailbox group working applications. The structures used in a group working experiment on the BLEND system are critically described, and recommendations for further work in this area are made.
    An Experiment in Group Working on Mailbox Systems BIBA 959-963
      T. I. Maude; N. O. Heaton; G. N. Gilbert; P. A. Wilson; C. J. Marshall
    A computer conference on the subject of electronic mailbox systems was held over the BLEND system in 1983. After nine months the ideas generated had been developed, written up and submitted to an international meeting as a paper, entirely by means of the computer conferencing system. The work that was done to prepare that paper, the roles and structures adopted and the patterns of communication between the mailbox members are described. Comparisons between mailbox and face to face group working are made, and some of the strengths and weaknesses of mailbox group working are identified. The paper, which has itself been written using the BLEND mailbox system, concludes that mailbox systems can be used successfully by groups of otherwise unrelated people to produce work jointly and within a previously defined timescale.
    Methodological Problems of Human Factors Research in Long-Termed CBMS Field-Trials BIBA 965-969
      Michael Pieper
    With special purpose to reveal changes of habitual communication to be caused by introducing Computer-Based-Message-Systems (CBMS), sociologists of GMD's 'Impact Research Group' evaluated a field-trial with the Computer Conferencing System KOMEX. KOMEX was developed by GMD's former 'Institute for Planning and Decision Support Systems'. The trial involved as pilot-users five subgroups working at different locations on different aspects of a common scientific project, founded by the 'German National Science Foundation (DFG)'. Evaluating changes of habitual communication induced by CBMS-technology required a quasi-experimental research-design. According to technical and non technical preconditions for communication within the field setting, different methodological approaches were used to control different aspects of user-behaviour before and after the introduction of KOMEX. All of these methods centered upon an approach to content-analyse all 356 messages distributed via KOMEX during the field-trial. The paper will discuss the practical problems arising from applying these methodologies with regard to the validity, reliability and representativity of generalized results. Finally, it shall be questionized whether the requirements of the quasi-experimental research design could be met, and in how far the methodological approaches have been appropriate to answer the analytical question outlined above.
    Design Criteria for the Electronic Journal BIBA 971-973
      John W. Senders
    An electronic journal (EJ) must meet the needs and satisfy the expectations of the readers if it is to succeed. Some reasonable criteria for an EJ are derived from consideration of the qualities of a paper journal (PJ). None of the requirements can be shown to be impossible to achieve or to be very much beyond today's state of the art.