HCI Bibliography Home | HCI Conferences | CHI Archive | Detailed Records | RefWorks | EndNote | Hide Abstracts
CHI Tables of Contents: 8182838586878889909192X92Y92a

Proceedings of ACM CHI'83 Conference on Human Factors in Computing Systems

Fullname:Proceedings of CHI'83 Conference on Human Factors in Computing Systems
Editors:Ann Janda
Location:Boston, Massachusetts
Dates:1983-Dec-12 to 1983-Dec-15
Standard No:ACM ISBN 0-89791-121-0 ACM ISSN 0713-5424; ACM Order Number 608830; ACM DL: Table of Contents hcibib: CHI83
  1. Plenary Speakers
  2. Interface Design 1 -- Prototyping Techniques
  3. Interface Design 2 -- The Design Process
  4. Interface Design 3 -- Experimental Evaluation
  5. Interface Design 4 -- Analyses of User Inputs
  6. Interface Design 5
  7. Command Languages
  8. Graphics-Based Interaction
  9. Menu and Query Language Design
  10. Text Editors
  11. Intelligent Interfaces
  12. Cognitive Models 1
  13. Cognitive Models 2
  14. Programming 1
  15. Programming 2 -- Documentation
  16. Physical Interface Devices
  17. User Documentation

Plenary Speakers

Design Principles for Human-Computer Interfaces BIBA 1-10
  Donald A. Norman
If the field of Human Factors in Computer Systems is to be a success it must develop design principles that are useful, principles that apply across a wide range of technologies. In the first part of this paper I discuss some the properties that useful principles should have. While I am at it, I warn of the dangers of the tar pits and the sirens of technology. We cannot avoid these dangers entirely, for were we to do so, we would fail to cope with the real problems and hazards of the field.
   The second part of the paper is intended to illustrate the first part through the example of tradeoff analysis. Any single design technique is apt to have its virtues along one dimension compensated by deficiencies along another. Tradeoff analysis provides a quantitative method of assessing tradeoff relations for two attributes xi and xj by first determining the User Satisfaction function for each, U(x), then showing how U(xi) trades off against U(xj). In general, the User Satisfaction for a system is given by the weighted sum of the User Satisfaction values for the attributes. The analysis is used to examine two different tradeoffs of information versus time and editor workspace versus menu size. Tradeoffs involving command languages versus menu-based systems, choices of names, and handheld computers versus workstations are examined briefly.
Manual Dexterity: A User-Oriented Approach to Creating Computer Documentation BIBA 11-18
  Patricia Wright
This paper will not advocate a list of firm recommendations about document design because it is recognised that design decisions will vary with many factors. Instead, the present discussion will emphasize that when making these decisions it is necessary for designers to take account of how readers will use the information provided. In order to help them do this, a simple framework is proposed which outlines the rudiments of how people interact with technical documents.
   The advantages of this framework will be illustrated by using it to motivate design decisions at two decision levels. At a "macro" level the document designer must make broad decisions about the contents and format of the manual. At a "micro" level the designer must select particular combinations of linguistic, graphic and typographic options which will help readers locate, understand and implement the information given in the manual.

Interface Design 1 -- Prototyping Techniques

Soft Machines: A Philosophy of User-Computer Interface Design BIBA 19-23
  Lloyd H. Nakatani; John A. Rohrlich
Machines and computer systems differ in many characteristics that have important consequences for the users. Machines are special-purpose, have forms suggestive of their functions, are operated with controls in obvious one-to-one correspondence with their actions, and the consequences of the actions on visible objects are immediately and readily apparent. By contrast, computer systems are general-purpose, have inscrutable form, are operated symbolically via a keyboard with no obvious correspondence between keys and actions, and typically operate on invisible objects with consequences that are not immediately or readily apparent. The characteristics possessed by machines, but typically absent in computer systems, aid learning, use and transfer among machines. But "hard," physical machines have limitations: they are inflexible, and their complexity can overwhelm us. We have built in our laboratory "soft machine" interfaces for computer systems to capitalize on the good characteristics of machines and overcome their limitations. A soft machine is implemented using the synergistic combination of real-time computer graphics to display "soft controls," and a touch screen to make soft controls operable like conventional hard controls.
Building a User-Defined Interface BIBA 24-27
  Dennis Wixon; John Whiteside; Michael Good; Sandra Jones
A measurably easy-to-use interface has been built using a novel technique. Novices attempted an electronic mail task using a command-line interface containing no help, no menu, no documentation, and no instruction. A hidden operator intercepted commands when necessary, creating the illusion of a true interactive session. The software was repeatedly revised to recognize users' new commands; in essence, the users defined the interface. This procedure was used on 67 subjects. The first version of the software could recognize only 7% of all the subjects' spontaneously generated commands; the final version could recognize 76% of those commands. This experience contradicts the idea that people are not good at designing their own command languages. Through careful observation and analysis of user behavior, a mail interface unusable by novices evolved into one that let novices do useful work within minutes.
Executable Specifications for a Human-Computer Interface BIBA 28-34
  Robert J. K. Jacob
It is useful to be able to specify a proposed human-computer interface formally before building it, particularly if a mockup suitable for testing can be obtained directly from the specification. A specification technique for user interfaces, based on state transition diagrams, is introduced and then demonstrated for a secure message system application. An interpreter that executes the resulting specification is then described. Some problems that arise in specifying a user interface are addressed by particular features of the technique: To reduce the complexity of the developer's task, a user interface is divided into the semantic, syntactic, and lexical levels, and a separate executable specification is provided for each. A process of stepwise refinement of the syntactic specification, leading from an informal specification to an executable one is also presented. Since the state diagram notation is based on a nondeterministic model, constraints necessary to realize the system with a deterministic interpreter are given.
Formal Specifications for Modeling and Developing Human/Computer Interfaces BIB 35-39
  J. W. Roach; M. Nickson

Interface Design 2 -- The Design Process

Design Practice and Interface Usability: Evidence from Interviews with Designers BIB 40-44
  N. Hammond; A. Jorgensen; A. MacLean; P. Barnard; J. Long
Getting Into a System: External-Internal Task Mapping Analysis BIBA 45-49
  Thomas P. Moran
A task analysis technique, called ETIT analysis, is introduced. It is based on the idea that tasks in the external world must be reformulated into the internal concepts of a computer system before the system can be used. The analysis is in the form of a mapping between sets of external tasks and internal tasks. An example analysis of several text editing systems is presented, and various properties of the systems are derived from the analysis. Further, it is shown how this analysis can be used to assess the potential transfer of knowledge from one system to another, i.e., how much knowing one system helps with learning another. Several issues are briefly discussed.
Designing for Usability -- Key Principles and What Designers Think BIBA 50-53
  John D. Gould; Clayton Lewis
Any system designed for people to use should be (1) easy to learn; (b) useful, i.e., contain functions people really need in their work; (c) easy to use; and (4) pleasant to use. In this note we present theoretical considerations and empirical data relevant to attaining these goals. First, we mention four principles for system design which we believe are necessary to attain these goals. Then we present survey results that demonstrate that our principles are not really all that obvious, but just seem obvious once presented. The responses of designers suggest that they may sometimes think they are doing what we recommend when in fact they are not. This is consistent with the experience that systems designers do not often recommend or use them themselves. We contrast some of these responses with what we have in mind in order to provide a more useful description of our principles. Lastly, we consider why this might be so. These sections are summaries of those in a longer paper to appear elsewhere (Gould & Lewis, 1983). In that paper we elaborate on our four principles, showing how they form the basis for a general methodology of design, and we describe a successful example of using them in actual system design (IBM's Audio Distribution System).
Evaluation and Analysis of Users' Activity Organization BIBA 54-57
  Liam Bannon; Allen Cypher; Steven Greenspan; Melissa L. Monty
Our analyses of the activities performed by users of computer systems show complex patterns of interleaved activities. Current human - computer interfaces provide little support for the kinds of problems users encounter when attempting to accomplish several different tasks in a single session. In this paper we develop a framework for discussing the characteristics of activities, in terms of activity structures, and provide a number of conceptual guidelines for developing an interface which supports activity coordination. The concept of a workspace is introduced as a unifying construct for reducing the mental workload when switching tasks, and for supporting contextually-driven interpretations of the users' activity structures.

Interface Design 3 -- Experimental Evaluation

Computer Response Time and User Performance BIB 58-62
  T. W. Butler
Computer Communication System Design Affects Group Decision Making BIBA 63-67
  Sharon Murrel
The impact of computer-based communication on group performance depends upon the structure enforced by the communication system. While the ability to introduce structures which enhance human communication processes has been applauded, research to evaluate the impact of various design features is lacking. This research has explored the impact of two synchronous systems which vary in the role of immediacy of interaction and feedback on group decision making. One system is message-oriented, requiring a conferee to complete a message before interacting with others. The other displays what each group member is typing in a separate window on the screens of all participants. In this system, comments can be made as ideas are expressed. Groups were asked to solve a problem first individually and then cooperatively using one of the two systems. All groups produced decisions superior to the average initial individual solutions. Window system groups both improved more and produced significantly higher quality decisions. These groups focused on fewer topics at one time while spending less time discussing how to organize both system and task efforts. By influencing the group's ability to organize and focus its attention, the design of the communication system influenced decision quality.
A Methodology for Objectively Evaluating Error Messages BIBA 68-71
  Barbara S. Isa; James M. Boyle; Alan S. Neal; Roger M. Simons
Message quality is a critical factor in influencing user acceptance of a program product. Good error messages can reduce the time and cost to create and maintain software, as well as help users learn about the product. We have developed a methodology for conducting controlled usability evaluations of error messages. The Message Test Program is easily modified to adapt to different product situations, and messages can be evaluated even before working code exists. The Message Test Program can be used to test error messages for a batch product, as well as messages for an interactive product. It can also be used for stand-alone messages, for products that offer on-line help, or messages that provide additional information in a reference manual. Message testing enables us to objectively evaluate error messages and provide specific feedback about the difficulties users encounter and how error messages can be improved.
Human Factors Testing in the Design of Xerox's 8010 `Star' Office Workstation BIBA 72-77
  William L. Bewley; Teresa L. Roberts; David Schroit; William L. Verplank
Integral to the design process of the Xerox 8010 "Star" workstation was constant concern for the user interface. The design was driven by principles of human cognition. Prototyping of ideas, paper-and-pencil analyses, and human-factors experiments with potential users all aided in making design decisions. Three of the human-factors experiments are described in this paper: A selection schemes test determined the number of buttons on the mouse pointing device and the meanings of these buttons for doing text selection. An icon test showed us the significant parameters in the shapes of objects on the display screen. A graphics test evaluated the user interface for making line drawings, and resulted in a redesign of that interface.

Interface Design 4 -- Analyses of User Inputs

Playback: A Method for Evaluating the Usability of Software and Its Documentation BIBA 78-82
  Alan S. Neal; Roger M. Simons
A methodology is described for obtaining objective measures of product usability. The Playback program developed at the IBM Human Factors Center in San Jose collects performance data of the user interface without impact upon the user or the system being evaluated. While a user is working with the system, keyboard activity is timed and recorded by a second computer. This log of stored activity is later played back through the host system for analysis. An observer watching television monitors enters time-stamped codes and comments concerning the users employment of system publications. The advantages of this approach are: (1) data-collection programs are external to the product being evaluated, (2) no modifications of the playback program are required for testing different software applications, (3) the data-collection process does not intrude on the user's thoughts or activities, (4) problem determination is performed at an accelerated rate during playback analysis, and (5) all data collection is performed on line.
Questionnaires as a Software Evaluation Tool BIBA 83-87
  Robert W. Root; Steve Draper
This paper reports on a study investigating the strengths and weaknesses of questionnaires as software evaluation tools. Two major influences on the usefulness of questionnaire-based evaluation responses are examined: the administration of the questionnaire, and the background and experience of the respondent. Two questionnaires were administered to a large number of students in an introductory programming class. The questionnaires were also given to a group of more experienced users (including course proctors). Respondents were asked to evaluate the text editor used in the class along a number of dimensions; evaluation responses were solicited using a number of different question types. Another group of students received the questionnaire individually, with part of it presented on the computer; a third group also evaluated an enhanced version of the editor in followup sessions.
Changes that Users Demanded in the Human Interface to the Hermes Message System BIBA 88-92
  Charlotte D. Mooers
The Hermes Message System has evolved in response to the needs and criticisms of users. This paper gives examples of some less than successful features, many of which have been changed, so that future designers will know what didn't work, as well as what does. Principles derived from this experience are: (a) What you see should be what you can type. (b) Commands and syntax should be uniform. (c) Commands and objects should be organized into groups. (d) Hierarchy is great for organizing things you know about but much less useful for finding things you don't know. Even with careful design, it is impossible to predict what users will dislike so it is important to design programs so they can be easily changed.

Interface Design 5

Computing on a Shoestring: Initial Data Entry for Service Organizations BIB 93-97
  Martha R. Horton
The Consul/CUE Interface: An Integrated Interactive Environment BIBA 98-102
  T. Kaczmarek; W. Mark; N. Sondheimer
Consul and CUE are two systems that combine to support an interface to interactive computer services that is integrated across a variety of interface methods. Consul is an experimental natural language interface system designed to be customized to a set of specific interactive computer services: electronic mail, personal calendar, word processing, etc. CUE is a window- and object-based run-time support environment for interactive services with a command language, pointing device and menu interface. Using the Consul/CUE interface, the user sees a single system that is capable of handling a wide variety of input in a completely uniform service environment. The success of the combined system derives from a large knowledge base formalizing facts in the interactive service environment in an artificial intelligence network structure.
A Generalized Transition Network Representation for Interactive Systems BIBA 103-106
  David Kieras; Peter G. Polson
A general method for describing the behavior of an interactive system is presented which is based on transition networks generalized enough to describe even very complex systems easily, as shown by an example description of a word processor. The key feature is the ability to easily describe hierarchies of modes or states of the system. The representation system is especially valuable as a design tool when used in a simulation of a proposed user interface.
   In order to characterize the interaction between a user and a system, an explicit and formal representation of the behavior of the system itself is needed. To be of value in the design of user interfaces, the representation should be independent of the actual implementation of the system, but also reflect the structural properties of the system's behavior, such as its hierarchical form, the possible modes, and the consistent patterns of interaction. At the same time, the presentation must be easy to define and understand. This paper presents a representation notation with these properties.
Application of a Model of Human Decision Making BIBA 107-111
  Mark E. Revesman; Joel S. Greenstein
When a human and computer perform similar tasks in parallel, it is important that an effective line of communication exist between the two entities. Since overt communication may add to the human's workload, an implicit method of communication is suggested in which the computer has a model of human performance on which to base actions. A two-stage model of human performance is employed in an experimental situation in which both a human and a computer act as decision makers. Results indicate that the implementation of a model significantly improves the human's performance and the overall system performance, without degrading the computer's performance. Research into additional experimental and real-world situations is suggested.

Command Languages

Using Examples to Describe Categories BIBA 112-115
  Susan T. Dumais; Thomas K. Landauer
The successful use of menu-based information retrieval system depends critically on users understanding the category names and partitions used by system designers. Some of the problems in this endeavor are psychological and have to do with naming large and ill-defined categories so that users can understand their contents, and effectively partitioning large sets of objects. Systems of interest (like home information systems) often consist of new and frequently changing content in large and varied domains, and are particularly prone to these problems. We explored several ways in which one might name categories in one such domain (Yellow Page category headings) - category names, category names plus examples, and examples alone. We found that three examples alone were essentially as good a way to name these categories as either an expertly chosen name or a name plus examples. Examples provide a promising possibility both as a means of flexibly naming menu categories and as a methodological tool to study certain categorization problems.
A Featural Approach to Command Names BIBA 116-119
  Jarrett Rosenberg
A variety of aspects of command names have been studied, such as suggestiveness, memorability, and the use of icons. A single framework for these disparate studies is desirable, and it is proposed that the concept of featural analysis prevalent in linguistics and psycholinguistics be adopted as an approach to command name design. Examples of the breadth of application of this approach are given for the naming issues of suggestiveness, learning and memory, congruence and hierarchicalness, universal commands, the relationships of names to the command language syntax, and the use of non-words as names.
Command Use and Interface Design BIB 120-124
  Robert E. Kraut; Stephen J. Hanson; James M. Farber
Is There Really Trouble with UNIX? BIBA 125-129
  Lorenzo De Leon; William G. Harris; Martha Evens
Donald Norman has claimed that UNIX has cryptic and inconsistent command names. As Michael Lesk has remarked, the lack of objective data makes it difficult to evaluate the significance of Norman's criticisms. In an effort to explore this controversy we taught one group of novice users the UNIX command language and another group an English-based command language (NUIX). The number of errors and calls for on-line assistance were compared. The subjects in this study were 22 high school women with no formal exposure to computers. The results reveal that the UNIX group made fewer errors than the NUIX group in two training sessions a week apart. Although calls for on-line assistance for the two groups in the first session were comparable, the UNIX group made over twice as many calls for on-line assistance in the second session as the NUIX group. Our findings suggest that even though the UNIX command language may not be harder for novice users to learn, it is probably more difficult for them to use.

Graphics-Based Interaction

Enhancing the Usability of an Office Information System Through Direct Manipulation BIBA 130-134
  Alison Lee; F. H. Lochovsky
In Office Information Systems, the primary focus has been to integrate facilities for the communication and management of information. However, the human factors aspects of the design of office systems are equally important considerations if such office systems are to gain widespread acceptance and use. The application of design techniques from Human Factors can help enhance the usability of an office system. In this paper, we describe the user interface of an office system developed by adapting such design techniques.
An Assessment of Computer Generated Space Situation Map Projections BIBA 135-138
  Mark D. Phillips; James E. Allison; Valor S. Dodd
C3 environments have increasingly incorporated computer controlled maps as decision aids. The design of map displays in space oriented C3 system has taken on greater importance due to the complex spatial relationships among orbiting objects. The large number of objects orbiting the globe and their great speeds further complicates efforts to quickly and accurately portray their positions graphically.
   This paper describes an experimental plan aimed at evaluating a new 2D/3D "hybrid" space situation map display. The Hybrid display is created by opening a globe at the south pole and flattening it into a platter. A third dimension is obtained by tilting the platter. It was hypothesized that the Hybrid display would offer an advantage to C3 system operators and analysts dealing with three dimensional problems. The test plan focuses on perceptual parameters and user preference issues concerning conventional and Hybrid display techniques. The results of this study will be examined to direct future work on dynamic displays, and the impact of display design approach on cognitive performance.
An Effective Graphics User Interface for Rules and Inference Mechanisms BIBAK 139-143
  J. W. Lewis
As the technology of rule-based inference mechanisms matures, knowledge acquisition-the creation, structuring, and verification of rules-becomes increasingly important. The accuracy and completeness of the rules in the knowledge base determine expert system performance, and the cost of acquiring that knowledge base dominates all other hardware and software costs in practical systems.
   To reduce knowledge acquisition time and error rate, a new interactive graphics interface for rules is being designed and implemented in GE Corporate Research and Development. In the new system, each set of rules is represented as an AND/OR graph and parts of the rule base are displayed on a CRT screen as an AND/OR tree. A user -- even an unsophisticated user -- can navigate the AND/OR graph, identify nodes to be modified, analyze the behavior of the graph, verify its correctness graphically, and follow the execution of inference engines.
Keywords: Expert systems, User interfaces, AND/OR trees
Effect of Font and Medium on Recognition/Confusion BIB 144-149
  S. M. Gupta; L. H. Geyer; J. A. Maalouf

Menu and Query Language Design

The Effects of Positional Constancy on Searching Menus for Information BIB 150-153
  Richard C. Teitelbaum; Richard E. Granda
Usable Natural Language Interfaces Through Menu-Based Natural Language Understanding BIBA 154-160
  Harry R. Tennant; Kenneth M. Ross; Craig W. Thompson
Conventional natural language interfaces suffer from several ease-of-use problems. They require a user to type and to formulate questions in a way that the system can understand. They have high failure rates which often frustrate users, and users often do not use features of the systems because they are unaware of them or don't trust them. In addition, conventional natural language systems are expensive to build and require large amounts of storage to use. This paper describes a new approach to natural language interfaces called menu-based natural language understanding. This new approach solves the problems listed above. The paper compares the menu-based natural language approach to conventional natural language interfaces and to other forms of interface and discusses the advantages and limitations of this new approach.
Query Languages for the Casual User: Exploring the Middle Ground Between Formal and Natural Languages BIB 161-165
  William C. Ogden; Susan R. Brooks

Text Editors

A Comparative Study of Moded and Modeless Text Editing by Experienced Editor Users BIB 166-170
  Merle F. Poller; Susan K. Garter
Patterns of Experience in Text Editing BIB 171-175
  Mary Beth Rosson
How Interface Design Determines Who Has Difficulty Learning To Use a Text Editor BIBA 176-181
  Louis M. Gomez; Dennis E. Egan; Evangeline A. Wheeler; Dhiraj K. Sharma; Aleta M. Gruchacz
In previous studies two background characteristics of computer novices were consistently correlated with their success in learning to use a line-based computer text editor. Older people and those who scored low on a standard test of Spatial Memory had more difficulty than younger people and those with higher Spatial Memory test scores. In the present study, we observed computer novices as they learned to use a screen-based editor, which presumably reduced spatial memory load. Contrary to expectations, performance using a screen-based editor was again strongly correlated with Spatial Memory test scores. However, the correlation between performance and subjects' age was significantly reduced. Overall, subjects were able to perform the same text editing exercises almost twice as fast using the screen editor compared to subjects in previous experiments using the line editor. These results are discussed in terms of the different cognitive demands placed on users by line and screen text editors.
How You Tell Your Computer What You Mean: Ostension in Interactive Systems BIBA 182-185
  James A. Galambos; Eloise S. Wikler; John B. Black; Marc M. Sebrechts
An important part of communication is being able to point to an object without referring to its components or to the area surrounding it. How to do this is the problem of ostension. We observed many ostension errors in novices learning to use a full-screen text editor. Specifically, the novices erroneously tried to use keys that are appropriate for pointing when using a typewriter but incorrect in screen editors (e.g., space bar, backspace key, etc.), they frequently missed the location they intended by one character, they inadvertently pointed to the wrong occurrence of a string using a FIND command, they incorrectly specified boundaries by forgetting about "invisible" characters (e.g., formatting characters), and they mistakenly attempted to point to nontyping areas of the screen that were off-limits.

Intelligent Interfaces

A Qualitative Reasoning Approach to Mathematical and Heuristic Knowledge Integration BIBA 186-189
  Stephen E. Cross
Humans problems solvers use heuristic knowledge. Heuristics can be justified in a given problem solving context by reasoning about 'deeper' domain theories. A working computer program, an air traffic control expert system, uses a qualitative reasoning approach to justify heuristically generated plans. The justification is based on mathematical knowledge of aircraft performance which is computationally too complex for use in the normal planning process.
The Effects of Limited Grammar on Interactive Natural Language BIB 190-192
  James A. Hendler; Paul Roller Michaelis
An Empirical Methodology for Writing User-Friendly Natural Language Computer Applications BIBA 193-196
  J. F. Kelley
A six-step, iterative, human factors design methodology was used to develop CAL, a natural language computer application to help computer-naive business professionals manage their personal calendars. Language is processed by a simple, non-parsing algorithm having limited storage requirements and a quick response time. CAL allows unconstrained English inputs from users with no training (except for a 5 minute introduction to the keyboard and display) and no manual (except for a two-page overview of the system). In controlled tests of performance, CAL correctly responded to between 86% and 97% of the inputs it received, according to various criteria. This research demonstrates that the methodological tools of the engineering psychologist can help build user-friendly software that accommodates the unruly language of computer-naive, first-time users by eliciting the cooperation of such users as partners in an iterative, empirical development process.
Correcting Misconceptions: What to Say when the User is Mistaken BIBA 197-201
  Kathleen F. McCoy
Because people's knowledge is often partial and/or faulty, it is inevitable that misconceptions will be revealed in the course of a conversation. If recognized, the other person may say something to correct the misconception, and the conversation continues. Just as this is the case when people interact with each other, so must it be when users interact with a computer system. For example, in interacting with an expert system, a user may reveal misconceptions about objects modelled by the system. By failing to correct such misconceptions, the system may not only confirm the original misconception, but may cause the user to develop further misconceptions. It must therefore be up to the system to recognize and respond to misconceptions in an effective way. In this paper the space of possible object misconceptions is characterized according to the kind of incorrect information involved. It has been found that this characterization is often useful in determining how the user arrived at the misconception, and therefore the kind of information to include in the response. Using such a characterization, a system will be able to effectively correct object misconceptions in a domain independent way. Factors which affect the amount of information included in a correction (such as discourse and situational context) are also examined.

Cognitive Models 1

The User's Perception of the Interaction Language: A Two-Level Model BIBA 202-206
  S. J. Payne; T. R. G. Green
Users perceive consistency and inconsistency in syntax, and family resemblances among syntactic constructions. These factors are not captured in conventional BNF-like grammars. We argue that a generalised form of a two-level grammar is a better model of the user's perceptions, and show how the model relates to current psychological notions of organisation in recall and language learning. The model provides a unified interpretation of many previous results in HCI: we analyse here findings by Reisner (1981) and Barnard et al. (1981). Two preliminary experimental tests supporting the model are described.
Learning Text Editor Semantics by Analogy BIBA 207-211
  Sarah A. Douglas; Thomas P. Moran
This paper presents a cognitive model for one aspect of how novices learn text editors-the acquisition of procedural skill by problem solving in problem spaces and the use of analogy for building a representation of the semantics of text-editor commands (which we call operators). Protocol data of computer-native subjects learning the EMACS text editor suggests that they use their knowledge of typewriting to decide which commands to use in performing editing tasks. We propose a formal method of analysis that compares operators in two problem spaces and generates misconceptions. The comparison of these predicted misconceptions with verbal comments, error data, and task difficulty lends support to this analysis.
Mental Models and Problem Solving in Using a Calculator BIBA 212-216
  Frank G. Halasz; Thomas P. Moran
It has often been suggested that users understand and reason about complex system on the basis of a mental model of the system's internal mechanics. This paper describes an empirical study of how mental model knowledge is used in operating a stack calculator. One group of naive users were taught step-by-step procedures for solving typical problems on the calculator. A second group of naive users were taught the same procedures in conjunction with an explicit model of the calculator's stack mechanism. The users then solved problems on the calculator while thinking aloud. Analysis of the performance of these two groups indicates that the model had little effect in routine problem solving situations. But significantly improved performance for novel problems. Analyses of the think-aloud protocols indicate that the users employed five distinct modes of problem solving: skilled methods, problem reduction strategies, a conversion algorithm, model-based problem space search, and methods-based problem space search. Skilled methods, problem reduction strategies and the conversion algorithm were used for solving more routine problems and did not necessarily depend on mental model knowledge. Problem space search was used in the novel problems. For the model users, the states and operations of the stack mechanism served as the problem space to be searched for a problem solution. In contrast, the no-model users employed a less effective search strategy based on the recombination of pieces of known procedures. These results indicate that explicitly teaching naive users an appropriate mental model of a system can provide a psychologically effective and robust basis for operating the machine.
Planning Units in Text Editing Behavior BIBA 217-221
  Scott P. Robertson; John B. Black
The organization of text editing behavior can be characterized by graph structures containing goals, subgoals, goal outcomes, and actions. Here we propose a model to represent the goals and plans of text editor users based on goal-fate analysis (Schank & Abelson, 1977). The representation captures relationships between a user's multiple goals and shows how errors can result from badly formed plans. We discuss some data from a psychological experiment which supports the hypothesis that text editing behavior is chunked into distinct plan units. The cognitive components of pause times between keystrokes were revealed by statistically removing the physical time required between keystrokes. Finally, we suggest how a system which builds goal-fate graphs from keystroke input might be useful in providing specific help information that is keyed to a user's intentions.

Cognitive Models 2

Remindings and Their Effects in Learning a Text Editor BIBA 222-225
  Brian H. Ross; Thomas P. Moran
How can learning in text-editing be characterized? Much recent work has focused on the use of analogies from prior experience. In this paper, we investigate the retrievals of earlier experiences within the editor and how they might be used by analogy to accomplish the task and learn the editor. An experiment is presented that demonstrates the effects of these "remindings" on performance. In addition, some possible determinants of these remindings are investigated. This experiment points out the need to consider not only the general form of instruction, but also the specifics of the instructional sequence as well. Irrelevant aspects of the task may have strong effects on performance. We consider three teaching techniques, designed to take advantage of these effects in different ways.
Learning in an Instructionless Environment: Observation and Analysis BIBA 226-229
  Jeff Shrager; David Klahr
In an instructionless environment, there are neither teachers nor books. The only feedback comes from interaction with the target. All information appears from within the subject or from observation of the environment. In this setting, subjects rely upon experimentation to develop an understanding of the target. They form hypotheses by analogy or inference and test these hypotheses via experiments of calibration, replication, confirmation, exploration, and discrimination.
   This paper describes subjects' performance in a particular instructionless environment. The target object is a programmable toy robot tank. We use the hypotheses formed by subjects and form of the experiments performed, to assess subjects' knowledge of the system. This knowledge falls into distinguishable categories: syntactic knowledge of the programming language, semantic knowledge of the actions of the device, and model knowledge which addresses the structure of the device. Exemplary selections of our protocols are used to support the various aspects of the learning model.
Human-Computer Discourse in the Design of a PASCAL Tutor BIBA 230-234
  Beverly Woolf; David D. McDonald
An effective human-computer discourse system requires more than a clever grammar or a rich knowledge base. It needs knowledge about the user and his understanding of the domain in order to produce a relevant and coherent discourse. We describe MENO, A prototype tutor for elementary PASCAL, which uses a set of speech patterns modelled after complex human discourse and a richly annotated knowledge base to produce a flexible interactive system for the user.

Programming 1

What Do Novice Programmers Know about Recursion BIBA 235-239
  Hank Kahney
Recent research into differences between novice and expert computer programmers has provided evidence that experts know more than novices, and what they know is better organized. The conclusion is only as interesting as it is intuitive. This paper reports an experiment which was designed to determine precisely what novice programmers understand about the behaviour of recursive procedures, and exactly how their understanding differs from an expert's understanding of the process. The results show that different novices understand, or misunderstand, different things. Implications of the findings are discussed with respect to other research into novice and expert programming performance.
Beyond Numbers: Don't Ask "How Many" ... Ask "Why" BIBA 240-246
  Elliot Soloway; Kate Ehrlich; John B. Black
While programmers may differ in their assessment of the comprehensibility of a program, there are nonetheless some clear cut cases of programs that are truly difficult to understand. In this paper, we analyze three programs -- two of which are relatively incomprehensible -- using Halstead's Volume Metric, Propositional Analysis and Plan Analysis. We argue that only Plan Analysis provides a satisfactory explanation for why the programs in question differ with respect to understandability. Moreover, we suggest that a qualitative analysis, such as provided by Plan Analysis, is the desired type of evaluation: rather than simply providing a numerical ranking for programs, the qualitative analysis can pinpoint the troublesome area in the code and provide prescriptive information for correcting the difficulty.
Aesthetics and Programming BIBA 247-250
  Peter Molzberger
The paper at hand is based on interviews with a total of eight so-called "superprogrammers", software people, who show exceptional performance quantitatively as well as qualitatively. It becomes apparent that these people do not experience programming as a purely rational activity, but that for them it possesses strong intuitive components.
   Programs are visualized wholistically as three-dimensional structures. In this, aesthetics plays a special part: the structure must please optically, be elegant -- then it is functionally acceptable. Logical mistakes manifest themselves as interfering with this aesthetics.
   The author suggests that in the area of software as well there is something like the absolute beautiful: perfect solutions with a maximum of transparence beyond all rivaling design parameters. He has a feeling that the faculties described in this paper are widespread and may open up a totally new dimension in programming.

Programming 2 -- Documentation

On Enhancing the Interface to the Source Code of Computer Programs BIBA 251-255
  Ronald Baecker; Aaron Marcus
This paper addresses issues in the human factors of computer program documentation. We develop a framework for research on enhancing the interface to the source code of computer programs through designing and automating the production of effective typeset representations of the source text. Principles underlying the design research and examples of sample production are presented.
Documentation of Concurrent Programs BIB 256-261
  Deborah A. Boehm-Davis; Andrew M. Fregly

Physical Interface Devices

Use of Mouse Buttons BIBA 262-266
  Lynne A. Price; Carlos A. Cordova
Two experimental tasks were designed to test use of multiple-button mice. In the first, number of errors made and time to complete subtasks were measured as subjects attempted to depress one, two, or three buttons under three sets of conditions. In the second, subjects were asked to indicate true or false either by pressing one of two different buttons or by clicking a single button one or two times. People tended to be faster and more accurate using different buttons than different numbers of clicks.
Speech Recognition at Two Field Sites BIBA 267-273
  A. Rollins; B. Constantine; S. Baker
The performance of two speech recognition systems installed at two field sites was analyzed. The speech systems were part of larger computer systems that were performing real functions in industrial environments. The two sites appeared to be polarized in terms of expected suitability for speech recognition. The variables looked at included task complexity, memory load, requirements for verification and error correction, vocabulary and syntax, microphone, operator experience and complexity of host computer software. Accuracy and throughput were measured for the speech recognition system at each site. The same measurements were made for keyboard entry. Operator differences account for most of the variance in results. Accuracy with voice input was higher than with keyboard for most operators. The most accurate operators with keyboard also tended to be the most accurate with voice. Throughput data appears more sensitive to individual differences in dealing with voice input, although the throughput data was clouded by slow host system response times overall. The discussion suggests that one to one replacement of keyboard with voice overlooks some possible advantages of voice. It is also possible to find operators who work well with voice. For those who do not work well with voice, the problems appear to be related to general work habits and attitude, rather than to specific difficulties with speech.
Lighting Characteristics of Visual Display Terminals from an Ergonomic Point of View BIBA 274-276
  U. Brauninger; E. Grandjean
Measuring procedures were developed to assess those lighting characteristics of VDTs which are of importance for visual comfort and for legibility: Luminance oscillation, sharpness, contrasts, stability and dimensions of characters as well as reflections on the display surfaces. 30 different VDT models of various European and US manufacturers disclosed great differences, indicating a big potential for improving the ergonomic qualities of VDTs.

User Documentation

An Experimental Evaluation of On-Line HELP for Non-Programmers BIBA 277-281
  Celeste S. Magers
An interactive computer system was made easier to learn for non-programmers by modifying the on-line HELP and error messages of a system designed primarily for programmers. The modifications included supplementing the existing HELP command with a HELP key, making the content of HELP and error messages more concrete, responding to command synonyms, and more. The systems were evaluated in a between-groups experiment in which office workers with no programming experience were asked to perform a typical office task using one of the unfamiliar interactive computer systems. The results of the experiment supported the inclusion of the modifications. Non-programmers using the modified system completed the computer task in less time, with greater accuracy, and with better resulting attitudes than those who used the system designed primarily for programmers.
A Proposal for User Centered System Documentation BIBA 282-285
  C. O'Malley; P. Smolensky; L. Bannon; E. Conway; J. Graham; J. Sokolov; M. L. Monty
This paper outlines a set of proposals for the development of system documentation based on an analysis of user needs. It is suggested that existing documentation is not sensitive enough to the variety of levels of user expertise, nor to the variety of contexts in which on-line help is required. We outline three specific proposals for fulfilling these needs: a quick reference facility, a command-line database, and a facility for full explanation and instruction, and suggest a number of ways in which users might access these facilities. Finally, we suggest a way of combining these facilities into an integrated structured manual, offering more effective user support than is currently provided.
Autobiography of a First-Time Discretionary Microcomputer User BIB 286-290
  Marilyn Mantei; Nancy Haskell