HCI Bibliography Home | HCI Conferences | BCSHCI Archive | Detailed Records | RefWorks | EndNote | Hide Abstracts
BCSHCI Tables of Contents: 858687888991929394959697980001

Proceedings of the HCI'89 Conference on People and Computers V

Fullname:Proceedings of the Fifth Conference of the British Computer Society Human Computer Interaction Specialist Group
Note:People and Computers V
Editors:Alistair Sutcliffe; Linda Macauley
Location:University of Nottingham
Dates:1989-Sep-05 to 1989-May-08
Publisher:Cambridge University Press
Standard No:ISBN 0-521-38430-3; hcibib: BCSHCI89
Papers:35
Pages:502
  1. Conference Theme Invited Keynote Paper
  2. Other Invited Keynote Papers
  3. Invited Plenary Debate Papers
  4. Design Methods 1 -- Requirements and Task Analysis
  5. User Interface Management Systems
  6. HCI Tools and Applications
  7. Design Methods 2 -- IKBS and User Centred Design
  8. Hypertext and Hypermedia
  9. Evaluation 1 -- Concepts and Methods
  10. Evaluation 2 -- Tools and Practice
  11. Cognitive Ergonomics
Preface BIB 1-2
  Russel Winder
Editorial BIB 3-5
  Alistair Sutcliffe; Linda Macaulay

Conference Theme Invited Keynote Paper

Conceptions of the Discipline of HCI: Craft, Applied Science, and Engineering BIBAK 9-32
  John Long; John Dowell
The theme of HCI '89 is 'the theory and practice of HCI'. In providing a general introduction to the Conference, this paper develops the theme within a characterisation of alternative conceptions of the discipline of Human-Computer Interaction (HCI). First, consideration of disciplines in general suggests their complete definition can be summarised as: 'knowledge, practices and a general problem having a particular scope, where knowledge supports practices seeking solutions to the general problem'. Second, the scope of the general problem of HCI is defined by reference to humans, computers, and the work they perform. Third, by intersecting these two definitions, a framework is proposed within which different conceptions of the HCI discipline may be established, ordered, and related. The framework expresses the essential characteristics of the HCI discipline, and can be summarised as: 'the use of HCI knowledge to support practices seeking solutions to the general problem of HCI'. Fourth, three alternative conceptions of the discipline of HCI are identified. They are HCI as a craft discipline, as an applied scientific discipline, and as an engineering discipline. Each conception is considered in terms of its view of the general problem, the practices seeking solutions to the problem, and the knowledge supporting those practices; examples are provided. Finally, the alternative conceptions are reviewed, and the effectiveness of the discipline which each offers is comparatively assessed. The relationships between the conceptions in establishing a more effective discipline are indicated.
Keywords: HCI

Other Invited Keynote Papers

Feeding the Interface Eaters BIBAK 35-48
  John M. Carroll
An abiding question in the psychology of human-computer interaction (as elsewhere in science) is how knowledge can be applied. One view is that pure knowledge always finds application; another is that knowledge only exists through application. Both are problematic: the former is frequently false and generally over-optimistic; the latter is idiosyncratic and subjective. In order to efficiently contribute to the design of computer systems and applications, the psychology of HCI must energetically press pure knowledge toward invention and must cultivate disciplined interpretation of applications already in hand.
Keywords: User interface, Knowledge application
Judging Software Design BIBK 49-56
  Ernest Edmonds
Keywords: Software design, Design criteria
Designing Systems to Match Organisational Reality BIBAK 57-69
  Ken Eason
There is widespread recognition that a major requirement in the development of future Information Technology Systems is more explicit recognition that systems implemented in an organisational setting have to support complex forms of multi-user work. This paper reviews the evidence that current systems often fail because they do not match organisational reality and prove unacceptable or unworkable to their users. A central problem is the need to recognise the boundaries between work roles in an organisation and to support the responsibilities related to each work role by appropriate technical design in respect, for example, of functionality distribution and database access rules. The second part of the paper reviews the methods by which systems to serve organisational reality might be developed. It focuses upon the development of generic products by suppliers. It notes that whilst it may be possible to specify the functionality necessary to support an organisational task common across a market sector, it is not possible to specify the distribution of responsibilities between work roles. This may vary from company to company and, at a detailed level, may be subject to frequent change within an organisation. The paper proposes a five level model for the locus of design decision making which enables the changeable character of organisational reality to be matched. It starts with the control the individual user needs to configure his or her own interface, examines the needs of local management to configure a system to match the way they wish their staff to work, considers the needs of the application designer tasked with the requirement to select and develop a system specifically for an organisation and, finally, looks at the implications of these layers of decision making for the supplier of generic systems. It concludes that suppliers need to provide a tool box to match different organisational realities rather than try to design for a particular organisational reality.
Keywords: Information technology, organisations
UIMS: Promises, Failures and Trends BIBAK 71-84
  Joelle Coutaz
This paper is a reflection on the promises and failures of UIMS. It shows how the linguistic approach to the design of UIMS has failed in supporting the behaviour of the user as well as direct manipulation user interfaces. The multiagent model, which stresses parallel modular organizations, appears as a promising way for improving the flexibility of UIMS. However, this model needs to be tested against some upcoming technologies: multimedia communication, distribution over networks, and simultaneous access by multiple users. The paper points out one key factor for the next step towards effective user interface development environments: the transfer of knowledge between scientific disciplines. Obviously this transfer requires the integration of techniques from diverse fields in computer science. Equally important, it requires integration of techniques from cognitive psychology with those of computer science.
Keywords: UIMS

Invited Plenary Debate Papers

Integrating Cognitive and System Models in Human Computer Interaction BIBAK 87-103
  Phil Barnard; Michael Harrison
System and user modelling are means of improving the usability of interactive systems, enabling designers to discuss features of the system and implementers to produce full and coherent implementations. Both types of modelling have something to say about the usability of interface and content that can be capitalised in design. However the apparatus is fundamentally different in each case. Here we are concerned with the central problem of bridging the gaps between psychological representations of user behaviour and formal descriptions of how the computer system behaves. We argue that a third framework is required, the interaction framework, that will incorporate appropriate concepts and principles for representing properties of conjoint user-system behaviour. We propose an agenda for developing such a framework. The purpose of this paper is to stimulate discussion rather than present a concrete proposal.
Keywords: User modelling, System modelling, Interaction modelling
Bugs: The Issue Facing HCI BIBK 105-107
  Harold Thimbleby
Keywords: User interfaces, Bugs
Giving HCI Away BIBAK 109-117
  Dan Diaper
This discussion paper briefly outlines the current state of HCI with respect to its two major contributory disciplines of psychology and computer science. It is claimed that as an engineering discipline, HCI must make its products available to its industrial and commercial customers in a form that supports their requirements. HCI methods are identified as the primary product of HCI and it is suggested that wherever possible such methods are supported by software tools. The need for education and training in HCI is discussed.
Keywords: Psychology, Computer science, Engineering, Industry and commerce, Methods and tools, Education and training

Design Methods 1 -- Requirements and Task Analysis

From Users to Dialogues: Enabling Authors to Build an Adaptive, Intelligent System BIBAK 121-135
  Helen Tang; Nigel Major; Rod Rivers
The research described in this paper addresses the need for tools to support the development of adaptive dialogues in intelligent training. Despite the number of intelligent tutoring systems already developed, there are few tools available which allow the author to create the courseware for an adaptive environment. This paper proposes tools to model classes of users in an object-oriented fashion, allowing general information about larger groups of users to be inherited by individuals. These tools are currently in development and exist as prototypes. The fundamental components of adaptive systems can be defined for each class of users. Dialogue strategies can also be defined in terms of dialogue primitives. These strategies are then mapped on to the user classes to provide dialogue appropriate to the situation. The research has focused on providing an adaptive environment for intelligent training but the techniques developed can potentially be used in a wider range of intelligent interface applications.
Keywords: User modelling, Dialogue management, Intelligent tutoring, Object-oriented systems
A Family of Task Models for Interface Design BIBAK 137-148
  Ray Waddington; Peter Johnson
Given the increasing prominence of the role of task analysis and modelling in software design, this paper describes a technique wherein a family of task models is created whose members each complement, and are integrated with, the stages commonly found in decompositional design. An outline of each model in this family is given. The technique may be used by designers to explore the possibility of alternative user interface designs. In order to demonstrate this possibility the paper presents examples of how we have used the models to re-design the user interface of an existing application to produce a version with the same functionality as the first.
Keywords: Task analysis, Task modelling, Interface design, Decompositional design

User Interface Management Systems

Dialogue Specification in the GRADIENT Dialogue System BIBAK 151-168
  J. L. Alty; J. Mullin
The need for a more human-centred approach to the design of dialogues for dynamic systems is highlighted and the knowledge-based approach to the design of the dialogue system in the GRADIENT project is described. The system has been designed to take advantage of the benefits of a User Interface Management approach and some deviations from the Seeheim Model are discussed. A dialogue specification technique is described in which the specification is separated into an environmental specification and a control specification. Using SAVE as an example the technique is explained and the resulting tool strategy in GRADIENT is outlined. Three examples of the benefits arising from the use of this specification technique are described -- quicker implementation, the use of a Minimum Presentation Tool and Path Algebra analysis.
Keywords: UIMS, Dialogue specification, Human-centred design, Dynamic systems, Path algebras, Dialogue-presentation separation
A New User Interface Architecture BIBAK 169-189
  Yigal Hoffner; John Dobson; David Iggulden
This paper proposes a new user interface (UI) architecture which is intended to help designers with the internal structuring of the UI -- the mechanism which animates and facilitates the dialogue between the user and the application.
   The architecture is based on the assumption that the role of the UI is to bridge the gap between the language of the application and the language of the user. There are two major structuring concepts behind the proposed user interface architecture: the language stage concept and the language transformation step concept. The language stage concept advocates the introduction of several intermediate language stages between the user language and the application language. The language transformation step concept deals with the necessary transformations between the intermediate language stages and also the actions required within each language stage.
   An example of the use of the proposed architecture in the design of a UI of a Chess playing application is given. Finally, directions for future work concerning the proposed architecture are given.
Keywords: User interface architecture
Exploratory User Interface Design Using Scenarios and Prototypes BIBAK 191-201
  Mark van Harmelen
Exploratory user interface design consists of the experimental construction of, or depiction of, the use of a user interface to an interactive system, and the subsequent evaluation and incremental improvement of that interface in an iterative fashion, in a way that is similar to the incremental change that occurs during exploratory programming.
   Two uses for exploratory user interface design are to investigate ideas of the scope and function of a system being designed, and to try out alternative user interface designs. Exploratory user interface design is best performed using tools to construct scenarios of interactive system use, or to rapidly construct prototypes of interactive systems.
   Observations are made about this technique and the use of scenarios and prototypes in two projects; one being the development of a large Integrated Project Support Environment, and the other being a small editor for a MIDI sound source. In these projects the construction of scenarios and prototypes formed a means of exploratory user interface design that, it is postulated, forms a valuable part of user interface designers' design techniques.
   Finally some existing exploratory user interface design tools are examined in the light of a discussion of desirable features for these tools.
Keywords: Interface design, Scenarios, Prototypes

HCI Tools and Applications

A Software Development Environment for End-Users BIBAK 205-216
  R. J. Hendley; N. Jurascheck
In many areas the most serious obstacle preventing the wider use of computers is the need to learn and use traditional text-based programming languages and environments. End users are often unable or unwilling to invest the time and effort required to acquire programming skills themselves. Professional programming help can be expensive, difficult to obtain and all too often will produce results which do not match the original requirements.
   The provision of easily used and easily learned programming environments is one solution to this problem which can be shown to be very useful, for some tasks.
   The BASE system provides a windowed environment and a set of tools which are easy for non-programmers to use, but which are extensible as the need arises. The primary programming tools are a Graphical Programming Language (GPL) which provides a natural mechanism for program construction and a 'schema filler' which provides for instantiation of the schemata which are the building blocks of the GPL. Underlying the whole system is a database which can integrate all the aspects of programming and which allows arbitrary attributes to be associated with any object in the system.
   The other tools are presently oriented towards the general area of Computer Based Learning (CBL) material, but the techniques used are more generally applicable.
Keywords: Programming environments, Graphical programming
Evaluating a Colour Coding Programming Support Tool BIBAK 217-230
  Darren Van Laar
Program comprehension and ease of debugging may be improved by using colour coding and indentation in a programming language (Gilmore & Green [1988]). This paper evaluates the usefulness of a program developed to colour code Pascal control structures to see whether it held advantages over ordinary monochrome indented displays. Sixteen (8M, 8F) experienced Pascal programmers took part in the experimental assessment of the package. Subjects were asked comprehension quiz questions about unfamiliar Pascal programs which were presented in the four experimental conditions: colour coded structures with normal (4 space) indentation, colour coding with no indentation, monochrome presentation (yellow on black) with indentation and monochrome without indentation.
   It was found that subjects answered comprehension questions significantly faster in the colour coded conditions than in the monochrome conditions. Indented conditions were significantly faster than the unindented conditions. Subjects rated each of the conditions on 'ease of use' for the task. A non-parametric test found that colour coded displays were rated as significantly more usable than monochrome displays while indented displays were seen as easier to use than unindented displays.
Keywords: Colour displays, Colour coding, Psychology of programming, Prototyping
The Wizard's Apprentice: A Program to Help Analyse Natural Language Dialogues BIBAK 231-243
  Dan Diaper
The Wizard's Apprentice is a computer program designed to aid a person analysing natural language dialogues recorded between a user and an expert system. Such dialogues have previously been collected by simulating an advisory expert system using the 'Wizard of Oz' simulation technique. The background rationale to the Wizard's Apprentice is outlined and its utility is described with particular reference to its architecture and the possible architecture of an intelligent interface for expert systems. The implementation of a version of a Wizard's Apprentice program is described.
Keywords: Natural language dialogues, Analysis, Expert systems, Intelligent interfaces, Simulation
Menu-Based Extensions to GNU Emacs BIBAK 245-257
  Russell A. Ritchie; George R. S. Weir
This paper describes extensions to the GNU incarnation of the Emacs editing system which provide facilities for menu-based interaction. Following the Emacs philosophy, the described system is designed to allow for user-customisation and extension of menus, which are sensitive to the user's editing mode. The present system affords pop-up and pull-right menus on SUN workstations, and terminal menus on the wide range of terminals supported by Emacs. Additionally, a menu-based menu design tool has been implemented. This provides intelligent support to individual users who wish to construct or alter menus. The operation of this Emacs menu system is outlined and its range of facilities described in detail.
Keywords: Menu design, Menu interaction, GNU emacs

Design Methods 2 -- IKBS and User Centred Design

A Client-Centred Methodology for Building Expert Systems BIBAK 261-275
  Andrew Basden
Methodology for building Expert Systems is a hot topic. Since the old ad hoc days, there have been many attempts to impose structure on the process. Some have attempted to take advantage of methodologies and procedures in conventional software engineering, while others have argued that these are not suited to Expert Systems.
   Most such methodologies are centred on the technology, and on the activities that the technology dictates. This paper offers a different kind of methodology, which has been developed from the viewpoint of the client. It comprises seven stages, which are named, not according to knowledge engineering tasks like knowledge acquisition, but according to deliverables that would be meaningful to clients. Certain things, such as role analysis, assume a higher importance than they do in other methodologies, and an attempt has been made to integrate usability into the overall system development process.
   This methodology was developed from real-life experience, rather than from the work of others, but the relation to other methodologies is made clear.
Keywords: Expert systems, Knowledge based systems, Life cycle, Methodology, Knowledge acquisition
Developing a User Requirements Specification for IKBS Design BIBAK 277-289
  K. R. Howey; M. R. Wilson; S. Hannigan
This paper describes the development and application of a user-centred approach to the design of expert systems. It outlines work to date on a case study which aims to develop a fault diagnosis expert system within the Electricity Supply Industry. The work was undertaken as part of the Alvey Demonstrator Project, Mobile Information Systems.
Keywords: Expert systems, IKBS fault diagnosis, Human factors system design, User requirements specification

Hypertext and Hypermedia

Extending Hypertext for Learning: An Investigation of Access and Guidance Tools BIBAK 293-304
  Nick Hammond; Lesley Allinson
We argue that hypertext provides a basis for exploratory learning systems, but should be supplemented by more directed guidance and access mechanisms. A system with a variety of such mechanisms is outlined and is the subject of a study in which the provision of facilities is systematically varied. The study highlights some problems with bare hypertext, including inefficient information access and failure to provide an overview, and illustrates how these problems can be addressed.
Keywords: Hypertext, Learning support environment, Information retrieval
Towards a Rapid Prototyping Environment for Interface Design: Desirable Features Suggested by the Electronic Spreadsheet BIBAK 305-314
  Thomas T. Hewett
Despite the desirability of rapid prototyping to allow for empirical evaluation of interface design alternatives, a generic set of useful characteristics for an 'interface designer's interface' are not yet entirely clear. Based upon experiences using an electronic spreadsheet to develop user-oriented templates which model simple neural networks, this paper describes some advantages and limitations of an electronic spreadsheet as a rapid prototyping environment for development of those user interfaces. In addition to immediate feedback after changes, the advantages include a physical format which enhances the usability of the interface by aiding both designer and end-user in visualizing the model; a capacity for easy replication of neural models and their user interfaces so that different versions can be examined in parallel; and an environment for the modular development and testing of both interface and model prior to implementation in more powerful computational environments. The parallel between designing templates for interactive neural models and designing interactive interfaces in general suggests that several of these features are desirable characteristics which should be part of any rapid prototyping environment.
Keywords: Rapid prototyping, Design features, Design tools, Interface design, Electronic spreadsheet, Neural modeling
MEMOIRS: A Personal Multimedia Information System BIBAK 315-326
  M. W. Lansdale; D. R. Young; C. A. Bass
This paper describes the background and development of a computerised personal information and database system MEMOIRS, that is intended to investigate and exploit peoples' everyday memory in offices. The system is based on a hypertext-style database in which each information node has links to a time-structured network (a 'Timebase') and additional links to document labels known as attributes. The user interface to the system is complex in that it supports a wide range of strategies and methods for retrieval of information. The paper considers the issues that this raises in the design of the interface and the extent to which the system is able to support the users' processes of recall and recognition in retrieval of information.
   The rationale behind the system design and the psychological theory behind it are discussed. The iterative development of various aspects of the interface are then considered, together with the problems encountered in designing a multimedia environment for MEMOIRS.
Keywords: Information retrieval, Long-term memory, Multimedia cues

Evaluation 1 -- Concepts and Methods

A 'Late' Evaluation of a Messaging System Design and the 'Target' of 'Early' Evaluation Methods BIBAK 331-344
  John Dowell; John Long
This paper describes a Human Factors evaluation of a messaging system. The evaluation was performed in the course of research developing methods for the ('early') evaluation of system design specifications. The research aims to develop early evaluation methods with the capability for pre-empting ('late') evaluations of implemented designs. The research approach is to investigate empirically this capability of the early evaluation methods. Accordingly, concurrent early and late evaluations of designs are being conducted; their comparison is intended to substantiate the pre-emptive capability of the early evaluation methods. Characterisation of the researchers' late evaluations identifies the particular classes of developers' late evaluation practices they represent. Hence, the 'target' of the early evaluation methods -- that is, the late evaluation practices they can pre-empt -- is being declared.
   Evaluation of the 'Intermail' messaging system was one late evaluation conducted by the authors for the purposes of the research. In describing the evaluation, the paper also presents its characterisation enabling the target of the early evaluation methods to be declared. The paper discusses the general need for early evaluation methods to declare their targets as a means of qualifying their utility. The paper demonstrates how that qualification can be provided.
Keywords: Evaluation, Messaging systems, Specification, Prototype
Evaluation for Design BIBAK 345-358
  Peter Wright; Andrew F. Monk
In an iterative design methodology prototypes or mock-ups are built and evaluated by having typical users work through realistic tasks. Data are elicited from users in order to refine the prototype. This paper is concerned with the types of data which may be obtained and how they may be used. Two commonly used forms of data are compared, a behavioural record of system use, in the form of a time stamped system log, and verbal protocols. In each case the objective is to illustrate practical methods of data collection and the kind of usability problems likely to be revealed. It is argued that an account of system usage alone provides insufficient data for the identification of many important usability problems.
Keywords: Iterative design, System log, Protocol analysis, Critical incident, Breakdown
An Evaluation of the Usability of a Human Factors Based Requirements Capture Methodology BIBAK 359-371
  Chris Fowler; Linda Macaulay; Adrian Castell; Andrew Hutt
This paper reports the results of an evaluation of part of the USTM (User Skills Task Match) Methodology. USTM is a Human Factors based methodology which addresses the earliest stages of the product development cycle i.e., from the initial idea, through feasibility assessment to requirements specification. The paper discusses evaluation results from the first part of the methodology which is concerned with requirements capture.
   The users of the methodology are multi-disciplinary design teams typically consisting of senior designers, strategic marketeers and technical authors. The objective of the evaluation reported here was to establish whether the methodology met a number of usability criteria including ease of use, usefulness, ease of learning and enjoyability. The main methods of evaluation used were questionnaires and structured interviews.
   The results are generally very favourable in terms of the objectives set. In particular it was found that the attitude of the design teams towards Human Factors changed quite significantly after using the methodology. They found focusing on the user and the user's environment at this early stage most informative and felt that Human Factors had an important contribution to make. Although the responses to questions on 'ease of use' and 'ease of learning' questions were favourable to the methodology, there was a general feeling that too much paper was generated and that the methodology could be better supported by employing an automated tool kit. The design teams also appreciated the team building nature of the workshops associated with the methodology and reported improvements in team dynamics and in shared understanding of the product opportunity. The overall results of the evaluation show that the design teams found the methodology usable, enjoyable and valuable.
Keywords: Evaluation, Software development, Requirements capture, Human factors

Evaluation 2 -- Tools and Practice

System Monitoring: Garbage Generator or Basis for Comprehensive Evaluation System? BIBAK 375-394
  M. Maguire; M. Sweeney
This paper discusses the technique of monitoring user interaction data (e.g., keystrokes, mouse movements and screen images) and places it within a general framework of usability evaluation methods. System monitoring technology is generally concerned with the capture of large amounts of raw data which may at first appear to give a narrow view of user-computer interaction, lacking the richness of both qualitative and contextual information. The objective of this paper is to discuss how by interpreting interaction data in a relatively simple way, and by viewing it alongside other data streams such as an observation journal, video and audio recordings, enriched abstractions may be made. The paper also emphasises the need for software support to help with the reduction and interpretation of data within such a comprehensive capture system. A model for the full system is described and reference is made to the authors' work on the Alvey HIMS (Human Interface Monitoring System) project which succeeded in implementing certain key components of it.
Keywords: Usability evaluation, System monitoring, Evaluation methods, Evaluation metrics
Direct Manipulation Prototype User Interface Monitoring BIBAK 395-407
  Miles Macleod
A simple automated technique is described for monitoring interaction between users and computer programs with direct manipulation user interfaces, implemented using HyperCard. With its set of easily tailorable interface components, HyperCard can be used as a prototyping tool to construct direct manipulation user interfaces, for the purpose of comparing design alternatives. The work described here makes available an additional means for their evaluation.
   Interaction is recorded as actions on discrete interface objects (e.g., buttons, menu items, fields), rather than at the level of mouse coordinates and pixels. This grain of analysis provides a readily interpretable record, with the potential of being matched against predictions derived from formalisable interaction task models. A log of actions and times is created unobtrusively during interaction, and may be inspected or written to an external text file when desired.
   Two stages in the development of AutoMonitors (systems which monitor themselves) are outlined. Firstly, the construction of an AutoMonitor. Secondly, the implementation of a software device which can convert HyperCard programs into AutoMonitors, without additional programming effort. Conversion involves the automated modification of the code attached to each interface object, and the grafting on of a user interface for the monitor itself.
   The design, and automated installation, of a system for recording users' and experimenters' comments is also described.
Keywords: Prototyping, Direct manipulation, Monitoring
An Integrated Approach to Monitoring the Behaviour and Performance of Inference Systems BIBAK 409-425
  Mike Brayshaw; John Domingue; Tim Rajan
Recent research into the graphical visualization of program behaviour has contributed to program development by enabling the programmer to see immediately the direct implications of coding decisions. This presentation technique is based on the same principles as WYSIWYG word processors and electronic publishing packages which ease publishing by allowing the editor to see both the large and fine-grained implications of page and subsection layout. However, the behaviour of programs is only a part of the programmers job. Once the behaviour has been finalized the task of performance analysis starts, where the programmer fine tunes the program in order to determine the most efficient representation. To extend the benefits of program visualization to the task of performance enhancement, this paper presents an integrated approach to the visualization of both program behaviour and performance, building on our previous research into visualization of the behaviour of inference systems. Users can view the details of performance metrics at a chosen level of grain-size, and in addition be able to interpret the statistics in terms of the behaviour of the program. This enables users to rapidly find performance problems with long distance views, and then home in on the specific details using a close-up view. Performance may be monitored in several different ways, and can also be customized by the user. The result combines the graphical tracing of the behaviour of a program with statistical measurements of its performance, providing the developer with an integrated picture of program execution in one display.
Keywords: Program visualization, Expert systems, Performance monitoring
HIMS: A Tool for HCI Evaluations BIBAK 427-439
  C. J. Theaker; R. Phillips; T. M. E. Frost; W. R. Love
This paper describes the design of a HCI Monitoring system which is used as a tool by human factors analysts in the evaluation of the usability of computer based products. Originally based on the 'Playback' concept, the system developed supports data acquisition and the analysis of the interactions between a user and the computer, with facilities for replaying the computer responses. Its salient features include generality of application, physical portability, high fidelity replay of high resolution workstation screens, multi-channel recording, synchronised video and sound recording, and a human factors workbench to control the tool and support the analysis of the captured data. Experiences with the use of Monitoring system for HCI evaluations are described.
Keywords: Usability, Monitoring, Playback, HCI, Metrics, Tools

Cognitive Ergonomics

Cognitive Dimensions of Notations BIBAK 443-460
  T. R. G. Green
'Cognitive dimensions' are features of computer languages considered purely as information structures or notations. They therefore apply to many types of language -- interactive or programming, high or low level, procedural or declarative, special purpose or general purpose. They are 'cognitive' dimensions because they control how (or whether) the preferred cognitive strategy for design-like tasks can be adopted; it has repeatedly been shown that users prefer opportunistic planning rather than any fixed strategy such as top-down development. The dimension analysis makes it easier to compare dissimilar interfaces or languages, and also helps to identify the relationship between support tools and programming languages: the support tools make it possible to use opportunistic planning with notations that would otherwise inhibit it.
Keywords: Computer languages, Opportunistic planning, Cognitive dimensions
Relating Ideal and Non-Ideal Verbalised Knowledge to Performance BIBAK 461-473
  Philip Barnard; Judi Ellis; Allan MacLean
It is important to understand relationships between knowledge and performance. We need to establish what users really know about systems rather than simply modelling ideal knowledge. A picture probe task is used to elicit user's ideal and non-ideal knowledge of task-action mappings in two different interfaces supporting common functionality. The users of these interfaces articulated different amounts of both ideal knowledge and non-ideal knowledge. For a given interface, however, users who articulate more ideal knowledge of task action mappings generally perform well but their amount of non-ideal knowledge does not relate systematically to their performance. Non-ideal knowledge discriminated between interfaces but not between the relatively efficient and inefficient users. We discuss these results in relation to models which should ultimately help in system design, and in relation to the provision of diagnostic tests and adequate on-line support for users.
Keywords: Protocol analysis, Knowledge elicitation, User performance and knowledge, User models
Exploiting Natural Intelligence: Towards the Development of Effective Environments for Learning to Program BIBAK 475-486
  T. Boyle; B. Drazkowski
This study involved the development and initial assessment of a learning environment for programming. The approach was designed to avoid the didactic, rule based approach common in textbooks. The learning environment encourages students to engage in active problem solving and to form their own cognitive representations of programming constructs. An implementation of this approach for a sub-set of Lisp was developed and assessed through the reactions of a sample of over 50 first year students. The scores on objective tests were high and subjective reactions were positive. It is concluded that this approach is productive and should be developed further.
Keywords: Learning, Programming, Exemplars
Skill Levels and Strategic Differences in Plan Comprehension and Implementation in Programming BIBAK 487-502
  Simon P. Davies
A number of authors have proposed that the 'programming plan' be regarded as the major characteristic of programming expertise. Such plans are thought to represent the programmer's knowledge of generic and stereotypic fragments of programs that correspond to specific task goals or sub-goals. A range of empirical studies have been undertaken in order to provide support for the notion of the programming plan and to establish the relationship between such plans and expertise. Most of these studies, however, have been concerned with what might be characterised as a theory of plans rather than with a theory of 'planning' in programming. Such studies have tended to examine only the static elements of plans -- attempting to show merely, for example, that plans are related in some way to expertise or to the notation of a particular language, rather than providing a means of looking at the way in which plans might be combined or refined. In addition they have neglected to examine the question of whether plans are implemented differently in different circumstances or with respect to different skill levels. This paper considers the results of two experimental studies which suggest that the relationship between programming plans and expertise is by no means straightforward. This work highlights the need to examine strategic differences in plan generation and comprehension that exist at different skill levels. In conclusion some tentative requirements are proposed for a theory of 'planning' in programming.
Keywords: Programming plans, Skill differences, Planning in programming