the construction of a thematic hypertext guide to science

99
The Construction of a thematic hypertext guide to science fiction by Elaine Watson BSc (Hons) A Master's Dissertation, submitted in partial fulfilment of the requirements for the award of the Master of Science degree of the Loughborough University of Technology September 1994 Supervisor: Mr Alan Poulter, BA, MA, MSc, ALA Department of Information and Library Studies .... .",. , © Elaine Watson, 1994 I .... ", ·"t.. r ·· t U,,,-' . . .. ... ·}'our. i ;"V.3fS:ty" ,; of ,0 ".,." I:· ... ,... . , 11'0 ,I 03'8:1 ::£-; ..... ....

Upload: others

Post on 06-Apr-2022

1 views

Category:

Documents


0 download

TRANSCRIPT

The Construction of a thematic hypertext guide to

science fiction

by Elaine Watson BSc (Hons)

A Master's Dissertation, submitted in partial fulfilment of the requirements for the award of the Master of Science degree of the Loughborough University of Technology

September 1994

Supervisor: Mr Alan Poulter, BA, MA, MSc, ALA Department of Information and Library Studies

~:r.';r;. .... .",. ~-==~~~lI~~~,~. ,

© Elaine Watson, 1994

~ I .... ", ·"t.. r ·· • t U,,,-' . . .. i.~""':~!j" ... ·}'our. i ;"V.3fS:ty" ,; of T~~·l,~ ,0 ".,." I:· ... ,... .

, ··'-i-~:ieb$~~:~'=.:. ::~ 11'0 ,I 03'8:1 ::£-;

..... .... 4i~iQi''''~l~~J:!'l;..~

Abstract

Abstract

This dissertation describes the design and development of a HyperCard system to

guide users via themes, to science fiction works. The classification of fiction is

examined, highlighting the problems in this area. The underlying concepts of

hypertext and the essential features ofthe HyperCard software are discussed. Design

considerations pertinent to the system are examined and the subsequent

implementation described. The dissertation concludes by considering the results of a

formal evaluation of the system.

11

Acknowledgements

Acknowledgements

I would like to thank the following people:

• My supervisor, Alan Poulter for his guidance and suggestions

• Andy Sawyer for his time and assistance at The Science Fiction Foundation Collection

• My parents, Grace and Alan, without whom all this could not have been possible.

111

Contents

Abstract Acknowledgements Table of Contents

Chapter

1 Overview 1.1 Introduction

1.2 Scope of Dissertation

1.3 The Realm of Science Fiction

1.4 Why Hypertext?

1.5 Aims

1.6 Objective

1.7 Methodology

1. 7. 1 Literature Search

1.7.2 Selection of data to be used in the system

1. 7.3 Data Processing

1. 7.4 Design

1.7.5 Construction of Prototype

1.7.6 Construction of System

1. 7. 7 Formal Evaluation

2 Historical Background 2.1 Introduction

2.2 The Role ofIndexes

2.3 Creation ofIndexes

2.4 Fiction Indexes

2.5 The Problems of Fiction Classification

2.6 The Historic Legacy

2.7 The Perceived Merit of Fiction

2.8 Fiction Impossible to ClassifY?

2.9 The Principles of Fiction Classification

IV

Contents

11

111

IV

1

1

1

2

3 3 4

4

4

5 6 6

6 7

9

9

10

11

11

12

12

13

13

2.10 Content Classification of Fiction

2.11 The BOOKHOUSE Project

2.11.1 Background to the BOOKHOUSE

2.11.2 Classification based on users' need

2.11.3 Search Strategies

2.11.4 The BOOKHOUSE Metaphor

2.11.5 Use ofIcons in the BOOKHOUSE

2.11.6 Navigating in the BOOKHOUSE

2.11.7 Search Strategies within the BOOKHOUSE

2.11.8 BOOKHOUSE Evaluation

2.12 Time taken for Fiction Classification

3 Hypertext and HyperCard 3.1 The History ofHypertext

3.2 Memex

3.3 NLS

3.4 Xanadu

3.5 Advantages of Using Hypertext

3.6 Disadvantages of Using Hypertext

3.7 HyperCard

3.8 The Basic Elements ofHyperCard

3.9 The Building Blocks

3.9.1 Stacks

3.9.2 Backgrounds

3.9.3 Cards

3.9.4 Fields

3.9.5 Buttons

3.10 The Navigation Aids

3.10.1 The Home Stack

3.10.2 User Preference Card

3.10.3 Recent

3.10.4 The Help Stack

3.10.5 The Message Box

3.11 HyperCard Menus

3.11.1 The Edit Menu

3.11.2 The File Menu

3.11.3 The Go Menu

3.11.4 The Object Menu

3.11.5 The Tools Menu

3.12 HyperTalk

v

Contents

15

17

18

19

21

21

21

23

23

24

25

29

29

30

30

30

32

33

33

34

34

34

34

34

35

35

35

35

36

36

36

37

37

37

37

37

37

39

Contents

3.13 System Design Considerations 39 3.13.1 Audience 39 3.13.2 Presentation of Subject Matter 39 3. 13.3 Consistency 39 3.13.4 Card Layout 40 3.13.5 Text 40 3.13.6 Independence of Screens 40 3. 13.7 Visual Effects 40 3.13.8 Feedback 41 3.13.9 Use of Sound 41 3.13. \0 Stack Opening Sequence 41 3.13.1\ Introduction and Help 41

3.14 Evaluation 41 3.15 The Use ofHyperCard for The Science Fiction Pilot 42

3.15.1 Availability 42 3.15.2 Authoring 42 3.15.3 Ability for user to browse 42 3.15.4 Cross Referencing 42 3.15.5 Flexible Interface 43

4 Designing the System

4.1 Introduction 45 4.2 Classification of Science Fiction 45 4.3 Problems of Classification 47 4.4 OCLC FirstSearch: BookRevDigest 47 4.5 The Lysator Science Fiction Archive 48 4.6 Specialist Science Fiction Magazines 48

4.6.1 Foundation 48 4.6.2 Vector - The Critical Journal of the BSFA 48 4.6.3 Locus - The Newspaper of the Science Fiction Field 49

4.7 Review Source Chosen 49 4.8 Advantages of Using Reviews 49

4.8.1 Speed 49 4.8.2 Range of Viewpoints 49

4.9 Problems of Using Reviews 50 4.9. I Personal bias of reviewer 50 4.9.2 Deducing plot from review 50 4.9.3 Deducing themes from reviews 51

4.10 Schedule for Classification of Science Fiction 51 4.11 Structure and Scope of The Science Fiction Pilot 51

V1

Contents

4. 11.1 Scope 51

4.11.2 Size 51

4.12 HyperCard Design Considerations 52 4.12.1 Audience 52

4.12.2 Visual Style 52

4.12.3 Visual Effects 53

4.12.4 Use of Sound 53

4.12.5 Text 53

4.12.6 Buttons 53

4.12.7 Card Layout 54

4.12.8 Navigation 54

4.12.9 Organisation ofInforrnation in Stacks 54 4.12.10 Evaluation 55

5 Implementation of the Design 5.1 Introduction 57

5.2 Structure of the System 57

5.3 Opening Sequence 58 5.4 IntroductionlHelp 59

5.5 Themes 60 5.5. 1 The Basic Search 64 5.5.2 Advanced Search 64

5.6 Glossary 65 5.7 Index 65 5.8 Fiction 70 5.9 System Protection 71

6' Evalnation and Conclusion 6.1 Introduction 73 6.2 Subjective data 73 6.3 Selection of Subjects 74 6.4 Questionnaires 74 6.5 Designing the Questionnaire 74 6.6 Formal Evaluation Results 75

6.6.1 Evaluators Computer Literacy 75 6.6.2 Fiction Experience 75

6.6.3 Overall reaction to The Science Fiction Pilot 76

6.6.4 Learning the System 76 6.6.5 Presentation of the System 76 6.6.6 Best and Worst Features of the System 76

V1!

6.6.7 Additional Comments

6.7 Conclusion

6.7.1 Strengths

6.7.2 Weaknesses

6.7.3 Opportunities

6.7.4 Threats

Bibliography

Appendix 1 - Themes Appendix 2 - User Evaluation Questionnaire

viii

Contents

77

78

78

78

79 79

81

85

87

Chapter I - Overview

1 Overview

1.1 Introduction

While extensive work has been done in the classification and indexing of non-fiction texts, fiction works have been severely neglected. This is particularly surprising when you consider that the majority of books borrowed from libraries are fiction. [I] Fiction books are usually arrayed on shelves alphabetically by author, and catalogues whether hard copy or electronic, usually follow the same convention. Unless· a reader is familiar with an author, or has been recommended a book, this fonn of arrangement does not suggest what a text is about, and whether they will find it of interest.

1.2 Scope of dissertation

By constructing a hypertext system, this dissertation attempts to increase access to science fiction, using themes as a method of selection. The field of fiction is wide ranging and covers a number of different genres, from the classics to romances. The science fiction genre was chosen as the area of fiction to focus on. As the system did not follow the convention of an index, it was decided to name it lhe Science Fiction Pilot. This was to reflect the guidance the system could offer, as well as link in to the visual style which was incorporated into the design.

1.3 The Realm of Science Fiction

Suppose an individual is intrigued by the genre of science fiction and would like to dip into it. How are they to separate the good from the bad, or having developed a

taste for science fiction find other books they might enjoy? Readers have only the title and a dubious cover illustration to give any clues to the content. Possibly they could refer to one of the published guides to science fiction, such as the Good

Reading Guide to Science Fiction and Fantasy [2]. These introduce the reader to quality science fiction, recommending titles and authors. However, no matter how well written the guide, this is a rather an insipid method to discover fiction.

Fiction is a statement of imaginary events which relate to one another and have a

Chapter 1 - Overview

narrative quality. As is the nature of fiction genres, it is difficult to pin science fiction down to a neat definition, even the authors themselves give a range of descriptions. Issac Asimov said, "science fiction is that branch ofliterature which is concerned with the impact of scientific advance upon human beings" [3] and Frank Herbert thought it "represents the modern heresy and the cutting edge of speculative imagination" [4]. There is no good reason to expect that a workable definition of science fiction wiU every be establish. For the purposes of incllJsion and exclusion in The Science Fiction Pilot, works that are consciously written as a contribution to the field, wiU be considered as part of the science fiction genre.

Needless t9 say science fiction has developed its own shape and vocabulary, which encompasses subdivisions and themes which readers often returned to time and again. It is thus along thematic guide lines that The Science Fiction Pilot was is hoped to be developed. Giving the individual a choice of themes, and then book titles and authors that closely correspond to them.

While indexing by theme gives a clearer idea to what a book is subsequently going to be about, this method has its disadvantage in that how fragmented do themes become, and yet still retain their use as a descriptor. By its very nature, a novel has many topics within the plot, therefore how accurate is a core theme in defining the book and encouraging a reader who may be interested in what an author has say about it. Too broad, and the theme loses it quality as a descriptor; too narrow, and elements of the text may have been neglected. At best using a thematic system can only hope to give a flavour of the what the book wiU offer.

1.4 Why Hypertext?

Browsing is a visible activity III libraries and bookstores. Morse explains the behaviour of a library browser as;

UsualIy not seeking a specific book; he is looking through the shelves to see what catches his eye. Nevertheless he does not allocate his search efforts purely at random; he goes to that section of the library that he estimates has the highest probability of containing a book or books his immediate interests would find to be worth borrowing. [5]

Unlike other databases, where searches through the stored information rely on knowing what you require, hypertext can be developed to link "chunks" of information together; a1Iowing the user to jump from one section to another. This gives the systems author the flexibility to create links by whatever criteria they wish. The user is able to browse through the data available, until a suitable item is

2

Chapter 1 - Overview

discovered. This mirrors the process nonnally used in selecting works of fiction, and yet allows lines of guidance to be incorporated.

Because of this flexibility in linking infonnation, the user interface can be tailored to the designers needs. For a fiction index, this is especially useful as there is no guarantee the user is computer literate, and thus the system can be designed to be attractive, unintimidating and easy to use. Therefore, an individual wishing to explore the genre of science fiction need not be put off by the tool to enable them to do so.

1.5 Aims

1. Establish a system to enter material into the Science Fiction Pilot

- work out themes to be allocated to material - establish a method to allocate themes to science fiction works - find a suitable set of data to enter in the system - ascertain a reliable source of reviews, to establish themes for data set

2. Construct the Science Fiction Pilot using HyperCard

- design system which uses themes as a retrieval method. - construct a prototype - infonnally evaluate prototype - construct a working model - input data

3. Evaluate the Science Fiction Pilot

- devise a questionnaire - find subjects - schedule evaluations sessions

1.6 Objective

The objective of this study is to produce a hypertext system which would use themes as a selection method for pieces in the science fiction genre, and then evaluate its effectiveness for users.

:I

Chapter 1 - Overview

1. 7 Methodology

1.7.1 Literature Search

A literature search was undertaken, in order to find previous research and information

within the numerous areas which were under study.

To find previous examples ofHyperCard systems and their construction, and articles

on the design of user based applications, the following CD-ROMs were inspected:

- INSPEC (ProQuest) - Applied Science & Technology Index ( Wilson ) - CTI+ Current Technology Index (Bowker-Saur )

For data on browsing behaviour as an information retrieval mechanism, and its use

within hypertext and the problems associated with it:

- PsychLit ( SilverPlatter )

- Social Science Index ( Wilson ) - ASSIA+ Applied Social Science Index Abstract ( Bowker-Saur )

-ERIC

For aspects about fiction classification and indexing, as well as information retrieval

problems;

- LISA Library and Information Science Abstract (Bowker-Saur)

- LibraryLit ( Wilson)

Searches were also made within the Internet. The most valuable sources of

information about science fiction were found in the following:

- Science Fiction Resource Guide

(ftp:// gandalf.rutgers.edulpub/sfl/sf-resource guide.html)

- The Lysator Science Fiction Archive (http://www.lysator.liu.se/sf-archive/sf-main)

1. 7.2 Selection of data to be used in the system

Due to time constraints, it was decided only a limited selection of data could be

entered into the system. Award nominees seemed to offer a selection of quality works within the genre, as well as giving a selection of representative science fiction

4

Chapter 1 - Overview

themes. They also provided pieces in the often neglected area of short fiction formats,

as well as novels.

Two of the most authoritative and well known awards are the Hugos and the

Nebulas. Votes for the Hugos taken at a large science fiction convention; the

Worldcon, while the Science Fiction Writers of America vote for the Nebulas. Thus,

nominees from these awards give a selection of fiction which is considered worthy by

both professionals and fans of science fiction. Due to time constraints nominees

published during 1992 were selected to provide the source of texts to be used in the

system.

However, as fiction published in single year is some what limiting, a retrospective of

one of the selected nominated authors was also included. The author James Morrow

was chosen as his work encompassed a range of themes, and again because of time,

was of a manageable quantity.

1. 7.3 Data Processing

In order to allocate themes to each item of fiction, a knowledge of the plot was

required. As reading each item of fiction would be time consuming, reviews ~

appeared an ideal method of providing plot details, as well as in some cases indicating

themes within the piece. A number of different sources were explored to find the

best and most reliable stock of reviews:

- OCLC FirstSearch: BookRevDigest

- The Lysator science fiction book archive (the internet)

-Specialist Science Fiction Magazines:

- Foundation

- Vector

- Locus

The source chosen was Locus, a specialist science fiction magazine, as it provided

extensive up-to-date reviews, including ones for short fiction. Additional information

was provided by the The Lysator science fiction book archive.

From the reviews, themes were deduced, and a synopsis prepared for each piece. In

order to establish a controlled vocabulary, the list of themes to used was devised in

from a list in The Encyclopaedia of Science Fiction [6]. (see Appendix 1 - Themes)

This encyclopaedia is considered one of the most authoritative reference works in the

field, and therefore the list of themes devised considered credible and worthy.

5

Chapter 1 - Overview

A system was devised to help in the allocation of these themes. Subdivisions were created to aid establishing areas of the plot to give themes. These subdivisions were:

• Environment • Character

• Science

• Ideas • Moods • Action

Each subdivision contained themes appropriate to it. Subdivision were not mutually exclusive, and a work could be defined by themes from a single subdivision.

1. 7.4 Design

In order to design a system, the potential users and their needs must be defined. Planning started by listing the information which fulfilled these needs and devising ways in which it could be linked most productively. As disorientation is the most common problem of a hypertext system, care was taken to try and establish links which users would be able to follow easily. The inclusion of periphery facilities into

the system, such as help for users, an introduction and glossary, was also devised.

1.7.5 Construction of Prototype

A prototype system was constructed for a number of reasons. Firstly, in order to aid learning of the authoring tools available and the HyperCard scripting language HyperTalk, and secondly, to create a skeleton system in which to explore the processes and links, which were desired to be used in the working system. Within the prototype it was possible to experiment with a number of different design arrangements and visual effects.

1. 7.6 Construction of System

After reaching a satisfactory level of development with the core of the system in the prototype, construction started on the working system. Processes involved in the

system production included, scanning in images, entering data and linking facilities. During the construction the system was informally evaluated to ensure it was not deviating from its aims and objectives. Individuals were asked to try out the system, and point out any difficulties they encountered, as well as being observed in their efforts. Revisions and improvements offacilities were taken in accordance with these appraisals.

Chapter I - Oven-iew

l. 7.7 Formal Evaluation

Following infonnal appraisals of the system throughout its development, a more fonnal evaluation was required. Subjective evaluation methods were chosen, as users opinions and judgement of the system would indicate if it was perceived as easy to

operate, and had assisted users to discover texts in a pleasing and understandable manner. The chosen method of evaluation was a questionnaire, as subjects could undertake the assessment at their own pace, avoiding experimenter bias or influence.

The evaluation of the Science Fiction Pilot involved ten subjects, who were supplied with a questionnaires of eighteen questions, divided into two sections. The first section explored the subjects experience of computers and science fiction, while the second section provided the subjects with realistic tasks to perfonn with the system, and then asked for their views via a series of mainly closed questions. The evaluation took each subject approximately fifteen minutes to complete.

7

Chapter 1 - Overview

References

I. Baker,S.L & Shepherd,G. W. Fiction classification schemes: the principles

behind them and their success. RQ. 1987,27(2), p.245.

2. Zool,M.H. Good Reading Guide to Science Fiction and Fantasy. Gt Britain:

Bloomsbury, 1989.

3. Manhall,B.T. The Science Fiction Reference Book. USA: Starmort House,

1981, p.497.

4. Manhall,B.T. The Science Fiction Reference Book. USA: Starmort House,

1981, p.499.

5. Hildreth,C. The concept and mechanics of browsing in an online library

catalog. [n: Williams,M.E & Hogan, T.H, ed. National Online Meeting.

Proceedings 1982. New Jersey: Learned Information, 1982, p.182.

6. Clute,J & Nicholls,P, ed. The Encyclopedia of Science Fiction. London:Orbit,

1993.

Chapter 2 - Historical Background

2 Historical Background

2.1 Introduction

The classification of fiction has been mainly restricted to simply that by author. The following chapter explores the reasons for this somewhat limited classification, in relation to non-fiction subjects and the effect this has on fiction indexes. Alternative methods of fiction classification are described and especially in some detail the BOOKHOUSE project, one of the most thorough and innovative approaches to date.

2.2 The Role of Indexes

The main objective of any index is that it is able to retrieve the records or documents. [1] The Oxford English Dictionary defines it as "that which serves to direct or point to a particular fact or conclusion; a guiding principle". [2] Thus an index is a pointer or guide, and as such does not generally supply the required information itself. It usually contains a set of tags or descriptors, which earmark the source of information which is being searched for. Indexes systematically lead the enquirer to previously detected information whenever it is needed. More expansively it can be thought of as "an orderly guide to the intellectual content and physical location of knowledge records". [3] Therefore, an index is only of any value if it ensures the information, whether documents or records, can be traced satisfactorily, with no impediments and irrelevant material. [4] As Pope stated, "Index-learning turns no student pale, yet holds the eel of science by the tail". [5]

Subject indexes are a series of index records with each record incorporating a word or phrase describing the subject acting as the access point, and further details. These details are likely to include the title and/or citation of the appropriate works, and may be supported by more details in another list. [6] The basic intellectual problem in assembling such an index, is to accurately portray a document, which may be composed of many thousands of words, with a dozen or less index terms. These few index terms must be chosen in such a way that the essential meaning and objectives of

the document can be conceptualised from these terms. [7] In this way an index normally hinges on a series of headings or access or entry points arranged in some

9

Chapter 2 - Historical Background

recognisable order. Each access point is supplemented by some other information, which is designed to indicate where further information or document details may be located. [8]

2.3 Creation of Indexes

All indexes must cater for the variety of potential approaches to a subject. This fundamental need causes two categories of problem .

• the indexer must label or name the subject effectively. • the indexer must find some avenue by which to signal the

relationships between subjects so that search strategies may evolve. [9]

These require considerations such as the use of synonyms (terms with the same or similar meanings). The merging of synonyms carries implications for the effectiveness of the index in terms of precision and recall. With homographs, as the words have the same spelling, their diverse meanings are established with in the context of the text. In isolation of an index this is lost. Nouns have plural and singular form, the question has to be decided whether to include both. [I 0]

The problem of specif'ying relationships between subjects lies with the fact there are two kinds. Syntactic relationships, which arise from the syntax of the document which is undergoing analysis; and semantic relationships, which show aspects of genus-species relationships and reflect assumed and well-recognised hierarchies.

The indexing process itself involves at least three stages:

Familiarisation -+ Analysis -+ Conversion of concepts to index tenns. [I I]

For the first process of familiarisation, in order to achieve good consistent indexing, the indexer must become conversant with the subject content of the document, and have a thorough appreciation of the structure of the subject, and the nature of the contribution that the document makes to the advancement of knowledge. [12]

The human analysis of a document, and decisions concerning which subjects are sufficiently significant for indexing, is difficult to codifY. Some features of the process can be specified, but others rely to a large extent upon experience and intuition. Some topics in a document represent the main theme of the document. This must obviously be represented in indexing, but to what extent need minor or secondary themes be indexed? [13]

10

Chapter 2 - Historical Background

Control is necessary in respect of the terms used in an index because of the variety of natural language. The easiest way to exercise this type of control over indexing

terms, is to list or store the acceptable terms in a vocabulary. A thesaurus is such a list. It aims to achieve some, or better coincidence, between the vocabulary of an indexer and the vocabulary used by the searcher. A more formal definition of a thesaurus might be, " an organised list of terms from a specialised vocabulary, which has been arranged to facilitate the selection of index terms". [14]

Terms from an indexing language can be assigned in one of three ways:

• analytically - where the human indexer analyses subject content and selects and assigns index terms which accurately reflect subject content.

• clerically - according to some prescribed procedure. • automatically - where a computer selects the terms. [15]

2.4 Fiction Indexes

However with works of fiction the most simplest form of indexing is used. An author index, comprises of a series of alphabetically organised author names. In some cases each of these names will be accompanifd by document reference numbers or classification numbers and document citations. [16] Sources which index works of fiction topically, although admittedly better than nothing, do not provide the user with the type or degree of subject access that a catalogue could provide. The major shortcomings of such sources is that they tell what books exist, but not which ones are owned by the library, much less what is currently on the shelf or where they are located. [17] Simplicity of indexing is due in part to the method of classification of

fiction.

In most public libraries, fiction is commonly shelved alphabetically by authors surname, a method coined as "classification by creator" by Clare Beghtol. [18] This has seemed both "convenient and satistying", the major reason being that library users want access to "works by" as well as "work about". [19]

2.5 The Problems of Fiction Oassification

The majority of books which are borrowed from libraries are fiction. Therefore, the most effort should logically go into making these extremely accessible to the library client. However, less time has been spent on the classification of fiction than on any other category of books. [20] The access to works of imaginative literature, in academic and especially public libraries, is clearly inadequate in comparison to the detailed access provided for works of non-fiction. [21]

11

Chapter 2 - Historical Background

Both the Dewy Decimal Classification (DDC) and the Library of Congress Subject

Headings (LCSH), contain explicit instructions not to provide topical subject access to individual works of fiction, with a few exceptions. [22]

In DDC, individual works of literature may be classed according to subject only if they are "redeemed" by sufficient informational content. The manual states "works of imagination intended to delight are classed in 800, but works that are essentially informational are classed according to subject in other parts of the schedule". If fiction has to be classified, it is either grouped with the literature about it or, if necessary for the production of catalogues, special subdivisions are sometimes added for that purpose. In the British National Bibliography, for instance, extra symbols

are added to the Dewey literature class marks to identifY novels as opposed to works about them. [23] In LCSH, "topical headings are only assigned to biographical fiction, historical fiction, and animal stories". [24]

This lack of attention can be traced to three main reasons. Firstly, a legacy of historic disregard of fiction; secondly, a question of fictions' merits and worthiness to deserve classification, and thirdly and most practical, the problematic nature of it classifYing a work of fiction.

2.6 The Historic Legacy

In early days of public libraries, librarians tended to take an uncompromisingly negative attitude to the provision of fiction. Although not a contemporary off this era Mary Baker Eddy, encapsulates the thoughts of the time in saying "Novels, remarkable only for their exaggerated pictures, impossible ideals, and specimens of depravity, fill our young reader with wrong tastes and sentiments". [25] This attitude mellowed as times, and social values changed, and in the late 19th century, novels were commonly accepted in public libraries for the sake of their moral and educational value, especially for readers from the "lower" classes. [26] Today of course, disapproval of popular fiction is not much in evidence. Librarians no longer see their mission in life as the improvement of their patrons' minds and lives through the provision of "good" i.e., non-fiction, books. [27] However so traditional is the lack of enhanced access to fiction in our libraries that many librarians do not even perceive it as an issue. [28]

2.7 The Perceived Merit of Fiction

Shelia Intner, in her article "The Fiction of Access to Fiction", speculated that one of the reasons why there is such a cursory access to fiction in public and academic libraries, is that fiction is perceived as being non factual or unreal, and hence, untrue.

12

Chapter 2 - Historical Background

Being untrue, fiction is thought to be less important, or at least, less worthy of serious study than non-fiction, which has the perceived merit of descending the actual, or real world. [29]

The Oxford English Dictionary defines a catalogue as, "now usually distinguished

from a mere list or enumeration, by systematic or methodical arrangement, alphabetical or other order, and often by the addition of brief particulars, descriptive, or aiding identification, indicative oflocality, position, date, price or the like." [30]

2.8 Fiction Impossible to Classify?

The reluctance that surrounds the provision of access to fiction stems in part from a perception that the whole endeavour will be impossible to accomplish. The major obstacle, is deciding what a work of fiction is about. Indeed, although "aboutness" remains an elusive concept even as far as non-fiction works are concerned, fictional works are even more resistant to categorisation, because they can exist on several different levels at once: the literal, the symbolic, and the thematic. [31]

This mixture of "concrete and abstract levels of meaning" is typical of fiction's multidimensional character. Whereas in older genres, such as fables, themes are often directly expressed as morals, in more modem fictional genres such as the novel or the short story, themes are commonly expressed indirectly through connotative elements such as imagery or point of view. [32] Unfortunately, most information embedded in fictional works must be regarded as inaccessible, stumbled upon serendipiously when reading for pleasure. [33]

2.9 The Principles of Fiction Classification

~ring the late nineteenth and early twentieth century, when the majority of classification scheme which are still in use today were invented, a number of attempts were made at the classification of fiction. These attempts identified five key principles for the classification of fiction.

The first and most important principle is that fiction classification should make it easier for library users to find the types of fictional work they want. Baker recognised, as early as 1899, that although fiction was in great demand, the users' assessablity to it was extremely limited. At this time, even when classification schemes were being developed and tested, Baker noted that "all our artificial arrangements of books, skilfully constructed catalogues, open access systems, and other methods of alluring and improving the reader .... stop short at fiction".[34]

Chapter 2 - Historical Background

The early theorists reCognised that such schemes should be designed for use by people who were just browsing among the shelves for something to read. Indeed, this function of guiding browsers to the types of books wanted has been mentioned by all those who wish to classify fiction. [35]

A second principle considered, was that readers are complex beings who select materials in a variety of ways and who do not always base their selection on any single dimension of a work; therefore any method of subdivision that serves to guide patrons to the type of book desired should be used. The early schemes suggested at least four types of subdivision, by the following:

I. broad subject 2. genre area 3. format of work 4. literary quality [36]

The third principle often mentioned was that fiction classification should perform the function of exposing readers to authors they might otherwise overlook. Borden an early writer in this area, noted that most fiction readers choose well known, best selling works simply because they had been made aware of the existence of titles

through advertisements, reviews, best seller lists and the like. Thus Borden suggests that classifYing lesser known fictional works by type of genre will at least help narrow the readers choices to some extent. [37] Wilson & Robinson asserted that

"description of genre or kind is important because the idea of a range of conventional procedure that guides both the performance of producers (if only in the direction of deliberate flouting of convention) and expectation of users". [38]

However in criticism to this, Pejtersen noted that genre systems consist of co-ordinate classes which are not in the least mutually exclusive. Nevertheless, exclusivity is

frequently assumed when novels are organised by genre: each novel is only classified by one of its characteristics and cannot be either retrieved via other, equally dominant, features. [39]

Two other principles have been mentioned, but are not universally agreed upon. That fiction classification schemes should not physically separate works of each type. Advocates say all fiction works should be interfiled, alphabetically by author, in one single scheme. Spine labels would provide the appropriate class markings. The main support for this is for purely practical reasons, as this aids the search for specific titles. Another argument is that separation of the various categories would limit the reader's opportunity for cross-category browsing and could result in the choice of only "poorer quality" genre titles. This could be disputed in that often users are only

14

Chapter 2 - Historical Background

interested in finding one type of book. [40]

On a similar track it is thought that a classification scheme should not separate works by the same author, even though they may have works in different areas. This could be contested by the fact that each work has separate though sometimes overlapping clientele. [41]

2.10 Content Classification of Fiction

Over the past ninety years or so, a limited number of people have experimented with the content classification of fiction. Three typical approaches can be identified:

a. Universal non-fiction classification schemes were adapted for fiction. b. Special classification schemes were designed for fiction

c. Fiction was organised into genre-systems, based on the reading appeal of different features of novels[ 42]

a. Frank Haigh's 1930's version of the Dewey Decimal Classification, is one example of a universal system adapted to fiction. He kept all main classes intact, but the original subdivisions were replaced by an incoherent mixture of familiar genre headings and newly invented subject categories, interspersed with some of the original subclasses. In each main class, subclasses were arranged in numerical, non-hierarchical order. The exact contents of the classes were not easy to establish or predict; for instance, ghost stories were classed under 'Philosophy', whereas mystery stories and thrillers were found under 'Sociology'. Haigh realised that it would be difficult to mark clear boundaries between his classes. When problems of overlap occur, he suggested that the classifier should 'make decisions'. [43]

b. C.A.Burgess in 1936 used his personal ideas about fiction to design a hierarchical decimal classification. It is not altogether clear how Burgess defined his classes, but they seem to represent a combination of genres and literary styles:

Examples of a main class and its subdivisions

3 love stories and erotic romances 31 simple non-tendentious, love stories 32 'Cinderella' formula 33 straightforward domestic formula; including stories of marriages 34 'Triangle' and 'Polygonal' formulae 35 Erotic: domestic 36 Erotic: exotic, topographical and primitive ('cave-man' etc.)

15

Chapter 2 - Historical Background

37 Erotic: picaresque ('love 'em and leave 'em' etc.) 38 Erotic: abnormal 39 short stories oflove and erotic romance

Independent of this subject classification, Burgess also devised four evaluative categories:

A = Literary or 'Highbrow' B = 'Middlebrow', competent and having literary ambitions C = 'Middlebrow', without literary ambitions, but competent 'best-seller' D = Frankly 'Lowbrow', commercial and mechanised

Librarians were expected to use these categories when guiding users, either towards a category of higher quality, or towards new subjects in a preferred value-category.

Burgess stressed the point that,

" ... these symbols are intended solely for staff use - to make , accessible to every member of the staff the mature judgement of the librarian or his book selectors. They should never be used as arranging symbols or appear in public records." [44]

c. There are several examples of fiction being organised into genre-systems, often with further study to examine the effect on book circulation.

In an early example, Borden during 1906, noticing that only new fiction was circulating well in his library decided this was due to lack of familiarity with the nature of older titles. Noting that queries for help were often of the nature of what kind of book, he classified a large part of his fiction into separately shelved categories. Borden observed circulation patterns for two years after his reclassification. It was found that now 57% of fiction circulation came from the classified shelves, suggesting that patrons were indeed helped in finding the type of book they wanted, regardless offamiliarity. [45]

Briggs in 1973 reported a study she conducted in a junior high school in San Jose, California. Fiction was classified into eight categories, and then books were shelved

alphabetically by author within these subject areas. Two years later students who had used the scheme from the beginning were asked for their opinion. In the survey, 88% found the new scheme easier to use than the previous alphabetically arranged section.

Those who used both the school and public library, 78% found the schools classified system easier to use. [46]

16

Chapter 2 - Historical Background

Harrell in 1985 studied the fiction schemes' used in forty nine large library systems (those serving 100 000 or more) in the United States. The major finding was that

parts of fiction collections were arranged and organised using genre categories. The most popular categories used were science fiction (used by 98% of those who classified their fiction), westerns (96%) and detective-mystery-suspense (93%). In all, 26 different categories were listed, by forty-six libraries, however, fifteen of these were being used by only one or two libraries. [47]

In the 1980 survey of 500 readers in four British public libraries, a portion dealt specifically with fiction categorisation. Respondents were asked if they often looked for a particular kind of novel, 69% said they did (although not exclusively). Patrons were asked if they would rather have different categories physically separated on the shelves or have all fiction shelved in one alphabetical author sequence. 59"10 wanted classified fiction shelved separately, while 24% opted for one alphabetical arrangement. Those favouring the first said that physical separation of the various fiction classes,

• saved time • was better for those who chose books primarily from one category

Those preferring a single alphabetical sequence of books classified only by spine labels said such an arrangement was

• better for those searching for particular authors • provide "serendipitous" opportunities for cross-category browsing • was a better arrangement for titles that were difficult to classify

Unprompted, 26% of the library patrons said they generally discovered new authors by browsing for the type of book they liked. [48]

2.11 The BOOKHOUSE Project

One of the most exciting and innovative approaches to fiction cataloguing and indexing in recent times has been undertaken by Annelise Mark Pejtersen. The BOOKHOUSE, was a computer-based search system for fiction which aimed at

supporting library users in finding fiction. It is an example of a system design using a cognitive work analysis which, in turn, builds on earlier studies of the information retrieval task based on actual user-librarian negotiations. These studies lead to the identification of a set of user search strategies for retrieving literature as well as multidimensional framework for classifYing fiction. Both of these have been necessary prerequisites for the design of the BOOKHOUSE, a flexible on-line

17

Chapter 2 - Historical Background

database system for novice users. It employed icons in the display interface in order to enhance the utility of the system for casual users in public libraries. [49]

2.11.1 Background to the BOOKHOUSE

Reviewing early fiction schemes Pejtersen summarised they were difficult to use in practise, because:

I. designers of new systems, as well as adapters of existing ones, drew heavily upon their personal sets of values and perception of fiction when they constructed their systems;

2. Genre-classes were regarded as true and adequate reflections of users' needs; 3. these systems did not reflect the typically complex nature of fiction requests,

where the subject is only one among several possible and equally relevant criteria (such as period, theme, accessibility);

4. the fundamental incompatibility between the multidimensional character of users' needs, and one-dimensional shelving of books, was not understood and systems were therefore intended for shelving as well as for retrieval purposes.

[50]

She noted that the two main problems of fiction retrieval were: firstly identifYing the user's often subconscious and intuitively formulated need, and that secondly the problem of formulating a relevant search strategy among documents that are not indexed and represented according to needs, but according to formal bibliographical data. Difficulties may arise here, since the main part of the "request to book" connection is usually rather intuitive, and)t is most likely that many of the habits of the users are based on ill-defined attitudes and associations. These define the fundamental frame of reference for comprehension and interpretation of most users' requests and criteria of satisfaction. [51]

Pejtersen believed these problems could be solved, by the design of a data base with access to document contents that was compatible with users' request/need criteria, and by the design of an interface to the database that supported their choice of search strategy. The design of a database, with a representation of document contents that reflected end users' formulations of their information needs, required a new notion of

the information content of the problem domain and a new kind of indexing scheme for analysis of document contents. This was to involve a new approach to the design of classification system derived from empirical analyses of users' inquiries in libraries. [52]

lR

Chapter 2 - Historical Background

2.11.2 Classification based on users' need

The design of the indexing system was to integrate two principle assumptions:

I. The concept of what a document is "about" should be defined on the basis of an empirical analysis of the formulations and needs of groups of users within a specific information retrieval environment.

2. Users' needs are multidimensional and the definition of the information content of the problem domain should reflect the different aspects of a general reader's needs. An indexing system should be closely related to the users' formulations of their needs and to their criteria for choosing documents.

Due to psychological factors .there is often a "gap" between users request and their real needs. In order to define the boundaries and attributes of a number of aspects within several of the dimensions which characterise the documents, empirical studies of users' needs based on their requests regarding document contents were conducted. [53]

The study showed how users tended to characterise book contents from a number of different angles and led to the conclusion that novels need to be classified in a multidimensional way. Four main dimensions were identified in users' requests; subject-matter, frame, author's intention and accessibility. [54]

The classification scheme for Analysis and Mediation of Publications ( AMP) therefore had four independent, facet-like main classes called "dimensions". Each dimension characterised fiction according to its own set of criteria, and was Subdivided into a few broad categorises. (see Figure 2.1- The AMP Classification Scheme for Fiction [55] )

The dimensions were not mutually exclusive but supplemented each other, since all characteristics may be found in the same book but with different dominance when a novel is classified. This multidimensional approach meant that the classification could not be used for shelf order. [56]

It may be argued, that it is incorrect to call this system a "classification". With its non-exclusive, multiple-place structure, and the lack of a notation, it certainly differed a good deal from traditional library classifications. It would be just as incorrect to use the label "indexing system" in its traditional sense, since book features are not listed alphabetically but organised into meaningful groups. But an appropriate name has not readily been presented. [57]

19

Chapter 2 - Historical Background

For the BOOK HOUSE designers, a natural consequence of supporting access to fiction was that it should be possible for the user to identifY and select search terms from one or more of these facets of "need" as the basis for a subsequent search. In this way the database and accessing requirements began to resemble those of non-fictional retrieval systems. [58]

2.11.3 Search Strategies

A second important ingredient of a cognitive work analysis was in the identification of the strategies which were employed by users in navigating through the work space to support their decision making. [59]

In the design of an information system that supports the user's choice of search strategies, a database is required to present information that matches the user's need and the requirements to their performance of the task. Secondly, a computer-aided user-system dialogue is needed which shares the tasks between the user and the computer, depending on their different resource characteristics in relation to different strategies.[60] (see Figure 2.2 - Review of Strategies [61] )

2.11.4 The BOOK HOUSE Metaphor

In the BOOK HOUSE, an attempt was made to exploit the flexible display capabilities of computers. By creating a kind of storehouse (or book house) in virtual space, it was possible to relate information in and about the database as well as the various means for communicating with it. The many dimensions were allocated locations in appropriate rooms or sections of rooms with in the storehouse. This type of metaphor was felt to be easily understood and remembered by both children and adults. [62]

2.11.5 Use ofIconsln The BOOK HOUSE

The most important issue regarding the use of icons, is their comprehension by the

user. This is not only a function of their form and content but, more importantly, the context within which they are viewed - e.g. where in the retrieval task one is, as well as the experience level of the user. How a user interprets an icon in a given situation depends on their intentions/experience at the given moment. The challenge for the designer is to provide a match between the context (of the information retrieval task) and the icons so that the user perceives the latter in the intended fashion. [63]

21

Figure 2.2 - Review of Strategies [61]

1. Bibliographical Strategy User: selects and decides author!, explores retrieved items, compares bibliographical data with need and compares content with bibliographical data.

2. Analytical Strategy User: communicates information on need. Accepts and rejects proposed documents.

3. Empirical Strategy User:communicates characteristic demographic, personal characteristics. Accepts and rejects proposed documents.

4. Search hy Analogy User:communicates information on need through a model document. Accepts and rejects proposed documents.

5. The Browsing Strategy User:selects and skims subsets of database information. Communicates field of interest and explores potential ofthe database.

6. The Check Routine User:requires selective information on documents. Copares match with need. Accepts and rejects retrieved documents.

7. Shift of Strategies User:accepts or rejects retrieved documents, or postpone decision about match with need. communicates new information on need or selects and decides known item.

22

Chapter 2 - Historical Background

Computer:assists with identification, verification and location of retrieved items. Communicates information contents for verifaction of identified items.

Computer:explores users' need. Compares need with representation of information contents of documents in the database and selects documents. Displays document contents for user's consideration.

Computer:explores user characteristics by

questions and proposals. Classifies user according to user characteristics and associates to stereotypical sets of items.

Computer:explores attributes of contents of model document,compares to document representations in database and selects,suggests documents for user's consideration. OR:

classifies document and associates to stereotypical set of items in the database.

Computer:scans through information content of subsets of database and compare match with profile of user interest.

Computer:displays representations of retrieved documents in appropriate formats matching the requirements of strategy.

Computer:checks the user's choice of strategy,probes the user's need and compares choice of strategy with need, suggest switch of strategy.

Chapter 2 - Historical Background

Figure 2.1 - The AMP Classification Scheme for Fiction [55]

DIMENSION 1: Subject matter a. action and course of events

b. psychological description

c. social relations

The subject content of the novel: What the story is about.

DIMENSION 2: Frame a. time: past, present, future

b. place: geographical, social environment, profession

The setting in time and place chosen by the author as the scenario of his book

DIMENSION 3: Author's intention a. emotional experience

b. cognition and information

The author's attitude towards the subject. The set of ideas and emotions which the

author wants to communicate to his readers.

DIMENSION 4: Accessibility a. readability

b. physical characteristics

c. literary form

d. main characters

e. age of main characters

The level of communication in terms of those properties which facilitate or inhibit

communication, such as difficulty of contents and language, composition, typography

etc.

Bibliographical data: Title, author, illustrator, translator, editor, publisher, pages, year of 1 st edition, year

of classified edition, cover of book, serial, illustrated, screen version, title country and

year of original edition, etc.

20

Chapter 2 - Historical Background

A pilot experiment was conducted to determine whether it was feasible to draw small icons to represent the index terms chosen from the different dimensions of the classification. Analysis of the data showed users were able to associate most of the icons with one or more of the terms used as the design basis for creating the icons. Four of the 96 icons caused problems. [64]

Difficulties arose when icons were too detailed to communicate an overall message, which at the same time denoted a relatively abstract notion. Distributed responses seemed to be related to different factors. One factor was the nature of terms associated with the icon, especially when icons were based on terms having a strong semantic relation. Some icons were not strong enough, or attempted to represent an abstract notion such as "feelings" or "togetherness" by referring to concrete objects which also tended to result in less concentrated agreement. [65] The experience gained from these experiments was used in the further design of the icons as well as in the implementation of further terms with icons in the interface.

The design had been focused on the use of icons as:

• signs indicating action alternatives during a search, and/or • symbols providing a symbolic representation of the semantic content of the

database - i.e. the substance of book contents and features. [66]

2.11.6 Navigating in the BOOKHOUSE

When a user entered the BOOKHOUSE through a virtual space hall, they were presented with three connecting rooms, all of which contained books on shelves. The left room had books for children, the right room, books for adults while the centre room had books for both. [67]

After the user had chosen a database, a new room appeared showing people busy searching for books in different ways. The user can thus select one of four different search strategies based on the results of the empirical studies mentioned previously.

[68]

2.11.7 Search Strategies within the BOOKHOUSE

1. Work room for Analytical search

The user enters this room and takes the place of the figure sitting at the table in the picture. On and around the table are icons representing the different dimensions of the classification system. There are thirteen of them. The user can select one of these

23

Chapter 2 - Historical Background

at a time, and thus get access to an open book display, with the textual listing of the particular set of search tenns which belong to the selected dimension. The user can also combine search tenns from the same or different dimensions. Thereafter the user can see descriptions of the books, which the BOOKHOUSE has found. [69]

2. Search by Reference book

To find a similar book, another room is entered which shows a book containing a title or author index for the database. This can be opened to help the user identifY their reference book. After selection, the system will automatically attempt to find other books in the collection which include as many of the same indexed attributes of the reference book as possible. The description of the ten most similar books will be presented one at a time in an open book format in decreasing order of relevance. [70]

3. Browsing icons and books

In this strategy, an open book representation with randomly chosen book description is given. The user can thereafter continue to step through other descriptions. [71]

The browsing strategy also includes an iconic version where the user can browse though small pictures describing book content; six pages of small icons; eighteen icons on each page representing the book contents. Icons are an economic way of supporting intuitive searches since a single picture can communicate many different facets of meaning and thus can include tenns from more than one dimension. Thus the 108 icons cover more than 1000 index tenns.[72]

2.11.8 BOOKHOUSE Evaluation

The general conclusion was that the BOOKHOUSE was accepted by a large majority of users as a helpful and pleasurable aid for finding good fiction. [73]

1. Classification system

The experiment demonstrated that a highly structured and selective access to content keywords, helped the user to fonnulate their needs more precisely, and therefore lead

to a better search result. Non traditional access modes, such as readability and physical book characteristics, proved to be exceedingly useful. The different dimensions and their related keywords were judged to be in good agreement;

overlapping categories such as genre, subject matter and theme were however difficult to distinguish between. [74]

24

Chapter 2 - Historical Background

The iconic representation of the classification system had a strong effect in helping the user to distinguish among the different kinds of information. It was difficult to

convey complex and abstract information with icons.

The only listing of index terms which proved to be annoying, was that for the time

period dimension. The numerical look-up facility limited the user to specifYing the

first two digits in the desired year. Thus much page turning was necessary to reach say the 1980's. [75]

2. Search strategies

The analytical search was the most popular among users and librarians. Thus compared to the results of work analyses in traditional libraries, the BOOKHOUSE

has completely reversed strategy preferences. As a result browsing was least popular

and the analytical approach the most popular in the sense that, although it was not the easiest, it was the most helpful. Users indicated, that the representation of these strategies in a room was too

crowded, and that it was difficult to determine the meaning of the various activities

which were depicted. [76]

3. Other conclusions

The majority of users were found to prefer direct manipulation through the mouse, to

the use of a command language through the keyboard.

Pictorial displays were found more suitable than text for differentiating between

information.

The design of command icons was not straightforward, especially when seen from the point of view of what they represent semantically and what functions they activated.

[77]

2.12 Time taken for Fiction Classification

While the previous methods of classification have attempted to address some of the the problems which fiction presents, one major obstacle remains. In order to be

familiar enough with a text to classifY it, especially with such a through scheme

outlined in The BOOKHOUSE, the work has to be read. Even for the fastest reader, this is a time consuming arduous process. Therefore, no matter how effective the

classification method devised, operators are still faced with the Herculean task of

reading and analysing the vast number of fiction books available.

25

Chapter 2 - Historical Background

References

1. Rowley.J. Abstracting and Indexing. UK:Clive Bingley Limited, 1988, p.48.

2. The Oxford English Dictionary 2nd Edition Vol VII Oxford:Clarendon

Press,1989,p.852.

3. Cleveland.D.C & Cleveland.A.D. Introduction to Indexing and Abstracting. USA:Libraries Unlimited Inc, 1990, p.26.

4. Ibid,p.143. 5. Stevensons Book 0/ Quotations. USACassell & Company Ltd, 1964, p.279.

6. Rowley.J. Abstracting and Indexing. UK:Clive Bingley Limited, 1988, p.49.

7. Cleveland.D.C & Cleveland.A.D. Introduction to Indexing and Abstracting. USA Libraries Unlimited Inc, 1990, p.26.

8. Rowley.J. Abstracting and Indexing. UK:Clive Bingley Limited, 1988, p.48.

9. Ibid, p.49. 10. Ibid, p.49.

11. Ibid, p.49.

12. Ibid, p.51. 13. Ibid, p.57.

14. Ibid, p.57.

15. Ibid, p.57. 16. Ibid, p.48.

17. Hayes.S. Enhanced catalog access to fiction: A preliminary study. Library Resources and Technical Services, 1992,36(4), p.444.

18. Ibid,p.441. 19. Ibid,p.446. 20. Baker.S.L & Shepherd.G.W. Fiction classification schemes: The principles

behind them and their success. RQ, 1987,27(2), p.245.

21. Hayes.S. Enhanced catalog access to fiction: A preliminary study. Library Resources and Technical Services, 1992, 36(4), p.446.

22. Ibid,p.441.

23. Pejtenen.A.M & Austin.J. Fiction retrieval: Experimental design and

evaluation of a search system based on users' value criteria (part 1). Journal 0/ Documentation, 1983, 39(4), p.231.

24. Hayes.S. Enhanced catalog access to fiction: A preliminary study. Library Resources and Technical Services, 1992,36(4), p.441.

25. Stevensons Book o/Quotations. USA:Cassell & Company Ltd, 1964, p.662.

26. Pejtenen.A.M & Austin.J. Fiction retrieval: Experimental design and

evaluation of a search system based on users' value criteria (part 1). Journal 0/ Documentation, 1983,39(4), p.231.

27. Hayes.S. Enhanced catalog access to fiction: A preliminary study. Library

26

Chapter 2 - Historical Background

Resources and Technical Services, 1992,36(4), p.443. 28. Ibid,p.444.

29. Ibid,p.443 30. The Oxford English Dictionary 2nd Edition VolI/ Oxford:Clarendon

Press,1989,p.967.

31. Hayes.S. Enhanced catalog access to fiction: A preliminary study. Library Resources alld Technical Services, 1992,36(4), p.445.

32. Ibid.,p.445.

33. Ibid,p.454. 34. Baker.S.L & Shepherd.G.W. Fiction classification schemes: The principles

behind them and their success. RQ, 1987,27(2), p.246. 35. Ibid.,p.246.

36. Ibid.,p.246.

37. Ibid,p.246.

38. Hayes.S. Enhanced catalog access to fiction: A preliminary study. Library

Resources and Technical Services, 1992,36(4), p.451. 39. Pejtersen.A.M. The BOOKHOUSE: Modelling user needs and search

strategies as a basis for system design. Denmark:Riso National Laboratory, 1989, p.17.

40. Baker.S.L & Shepherd.G.W. Fiction classification schemes: The principles behind them and their success. RQ, 1987,27(2), p.247.

41. Ibid.,p.247.

42. Pejtersen.A.M & Austin.J. Fiction retrieval: Experimental design and evaluation of a search system based on users' value criteria (part 1). Journal of Documentation, 1983, 3 9(4), p. 23 1.

43. Ibid.,p.23 1. 44. Ibid,p.232.

45. Baker.S.L & Shepherd.G.W. Fiction classification schemes: The principles behind them and their success. RQ, 1987,27(2), p.247.

46. Ibid.,p.247.

47. Ibid.,p.249.

48. Ibid.,p.248.

49. Pejtersen.A.M. The BOOKHOUSE: an icon based database system for fiction retrieval in public libraries. Information and Innovation Proceedings of the 7th Nordic Conference for Information and Documentation, 1989, p.165.

50. Pejtersen.A.M & Austin.J. Fiction retrieval: Experimental design and evaluation of a search system based on users' value criteria (part I). Journal

of Documentation, 1983,39(4), p.233. 51. Pejtersen.A.M. The BOOKHOUSE: Modelling user needs and search

strategies as a basis for system design. Denmark:Riso National Laboratory,

1989, p.13.

27

Chapter 2 - Historical Background

52. Ibid,p.13. 53. Ibid,p.2l.

54. Ibid,p.26. 55. Ibid,p.29. 56. Ibid,p.30. 57. Ibid,p.30. 58. Pejtersen.A.M. The BOOKHOUSE: an icon based database system for

fiction retrieval in public libraries. Information and Innovation Proceedings of the 7th Nordic Conference for Information and Docume11lation, 1989, p.166.

59. Ibid,p.167. 60. Pejtersen.A.M. The BOOKHOUSE: Modelling user needs and search

strategies as a basis for system design. Denmark:Riso National Laboratory,

1989, p.59.

6l. Ibid,p.60.

62. Pejtersen.A.M. The BOOKHOUSE: an icon based database system for

fiction retrieval in public libraries. Information and Innovation Proceedings of the 7th Nordic Conference for Information and Documentation, 1989, p.169.

63. lbid,p.169. 64. Goodstein.L.P & Pejtersen.A.M. The BOOKHOUSE: System functionality

and evaluation. Denmark:Riso National Laboratory, 1989, p.43.

65. Ibid,p.44. 66. Pejtersen.A.M. The BOOKHOUSE: an icon based database system for

fiction retrieval in public libraries. Information and Innovation Proceedings of the 7th Nordic Conference for Information and Documentation, 1989, p.169.

67. Ibid,p.170. 68. Ibid,p.170. 69. Ibid,p.170. 70. Ibid,p.17l. 71. Ibid,p.17l. 72. Ibid.,p.l71. 73. Goodstein.L.P & Pejtersen.A.M. The BOOKHOUSE: Systemfil11ctionality

and evaluation. Denmark:Riso National Laboratory, 1989, p.133.

74. Ibid,p.133. 75. Ibid,p.134. 76. Ibid,p.135.

Ibid.,p.136.

28

Chapter 3 - Hypertext and HyperCard

3 Hypertext and HyperCard

3.1 The History of Hypertext

Hypertext has added a new dimension to infonnation management. Hypertext has

been described by Yankelovich and Meyrowitz as:

"non-sequential writing and reading ... allowing authors to link

infonnation, create paths through a corpus of related material,

annotate existing texts, and create notes and point readers to either

bibliographic data or the body of the referenced text ... {it} can allow

the creation of an automated encyclopaedia." [1]

And more technically as:

"at its most basic level, a database management system that lets you

connect screens of infonnation using associative links. At its most

sophisticated level, hypertext is a software environment for

collaborative work, communication, and knowledge acquisition.

Hypertext products mimic the brains ability to store and retrieve

infonnation by referential links for quick and intuitive access." [2]

3.2 Memex

A "machine" that operated in a similar way to the human mind was envisaged in 1945

by, Vannevar Bush.

"The human mind .... operates by association. With one item in its

grasp, it snaps instantly to the next that is suggested by the

association of thoughts, in accordance with an intricate web of trails

carried by the cells of the brain .... the speed of action, the intricacy of

trails, the detail of mental pictures, is awe-inspiring beyond all else

in nature. Man can not hope fully to duplicate this mental process

artificially, but he certainly ought to be able to learn from it." [3]

29

Chapter 3 - Hypertext and HyperCard

Bush illustrated this theory in his virtual system called a Memex. This device would

resemble a mechanised private file and library. The electronic system would support

browsing and note-making by embedding links to assist the reader. The linkages

would be predetermined, in anticipation of patterns of association and would result in a non-linear document acting as a supplement to the readers memory. [4]

3.3 NLS

Influenced by Bush's concept of associative links and browsing, Douglas Engelbart centred his research during the early 1960's at Stanford Research Institute on using

computers to augment human intelligence. He developed an On-line System, or

NLS( oN-Line System) which served as a storage receptacle for documentation, as a

communication network, since on-line conferencing was possible, and as a shared

work space where researchers could plan and design projects. An important feature

was viewing filters, which let a user move quickly through a database, scanning for

pertinent data. Such was its success the system was marketed commercially as

Augment. [5]

3.4 Xanadu

Also in the 1960's, Theodore Nelson launched his Xanadu Project at Brown

University. Named after the place of literary memory in Coleridge's Kubla Khan, the

system was designed to be an ever expanding publishing environment that would

allow users to create, interact and interconnect with linked electronic documents and

other forms ofhypermedia. As a repository of published documents that allows users

to to link to other documents, use or follow quotations to their origins, this removes

the need to seek permission to republish, and royalties are paid automatically. [6]

Nelson, who coined the term "hypertext", concluded that:

"today's conventional databases will not satisfY the information

needs of the non-computing public, nor can they provide methods

for publishing the ever more interconnected writings now being

placed on electronic networks. It is our hope to bring the power of

electronic access to the new and sweeping literary medium of

hypertext, in all the forms that the mind can devise." [7]

3.5 Advantages of Using Hypertext

Conklin defines a hypertext system as:

30

Chapter 3 - Hypertext and HyperCard

"a computer-based medium for thinking and communication .... To

qualifY as hypertext, a system should require no more than a couple

of keystrokes (or mouse movements) from the user to follow a single link. In other words, the interface must provide links which

act like "magic buttons" to transport the user quickly and easily to a new place in the hyperdocument." [8]

This early definition helps to indicate some of the advantages to be offered by a

hypertext system. Using a printed reference volume, a researcher is required to scan through the index or contents page to find the relevant topic. On finding the passage

of interest, the researcher may have to find the definition of key words in the glossary

or dictionary, refer to footnotes, consult the bibliography, or locate information in the appendices.

This richly interlinked hierarchical structure transfers easily to hypertext, reducing the considerable page turning activity of traditional printed material to a fraction of the

time. Thus a researcher is able to to follow references forward to their referent, or

backwards to their reference, often at only the click of a text embedded button.

Unlike a traditional printed text the user is not confined to the linear structure prepared by the author. A reader may move through the document selecting the most

pertinent pieces of information for their needs.

For information from traditional sources, hypertext can:

• make the existing structure of the information more apparent.

• use the links between disparate parts of the same document, to easier access to remote but related information structures. [9]

Thus a hypertext document caters for the main information seeking behaviours:

• Goal Oriented

• Background

• Browsing

information seeking for a particular purpose; implies familiarity and specificity.

information seeking for fuller understanding; implies Information non-familiarity but specificity.

information seeking with no particular purpose; implies

serendipity, maybe familiarity but no specificity. [l0]

31

Chapter 3 - Hypertext and HyperCard

Browsing, being far more rewarding and in hypertext than a printed document, due to ease with which the user can jump between topics, trace links and discover information previously not thought of as the search began.

3.6 Disadvantages of Using Hypertext

Swimming around in a sea of information has its problems. Users can get progressively lost in the hyperdocument; there may be difficulties of comprehension

using a facility which offers both ramified subject content and novel search strategies. Thus·there are two main problems in finding your way round a hypertext document. One is the general problem of disorientation and the other is the failure to locate specific items. [11]

Linear text has inbuilt clues which guide the user through text e.g. page numbers,

paragraphs, footnotes, punctuation, layout, etc. Non linear texts have not. Rather than adopting the prompts of the linear environment, hypertext systems uses appropriate clues such as highlighting, flashing signals, reverse text systems and

others embedded in the text. An aid to prevent getting lost in hyperdocuments is to allow users to know where they are; where they have been; where they can go from here. The hypertext should be structured in such a way that users quickly learn where to expect recurring features. [12] to

This highlights the prominence of which careful and thoughtful design has on the efficient working of a hypertext system. The user's ability to separate subject matter from multiple navigational trails, complex networks of concepts and relationships, should not be taken for granted. The facility must be in built to allow the user who goes off at a tangent to return to the point of departure and take up the main strand of the thoughts again. Where the system also allows serendipity or discovery learning, the presentation of the information must be such that the user can at any time recover from a browsing excursion to the point of departure. [13]

Care must be taken to avoid uncontrolled linkages. These can turn hypertext systems into a meaningless mass of obscure connections and references. On a similar theme, is the difficulty of breaking a thought or segment of information into a node. Themes in a document can be very tightly interwoven, so much so that breaking the information into discrete nodes would be detrimental. [14] It is vitally important that links are accurate or the function has no value at all. [15]

Therefore, not all literature is suited for a hypertext literary system. Similarly, even though information may have discrete components, the application developer may not be experienced to perceive these units. Information may be broken into nodes

Chapter 3 - Hypertext and HyperCard

prematurely. [16]

3.7 HyperCard

Hypertext gained a higher profile as commercial programs became available. Designed by Bill Atkinson, HyperCard was launched by Apple at the Boston Mackintosh Expo in August 1987. HyperCard was supplied with every new Apple Mackintosh, and applications created on HyperCard can be sold or distributed without further royalties or fees being negotiated. From 1992 HyperCard became a retail product, but the five years of magnanimousness by Apple has paid off, making HyperCard one of the most popular HyperText systems available.

When information is entered into a HyperCard, a user can jump to related data, via links. These links can be establish by whatever criteria the author sees as most relevant. Thus users are able to quickly browse through information, an essential quality of a hypertext system

HyperCard has limitation as a HyperText program. Words embedded in text cannot be made triggers for hypertext operations without some difficult programming. Only one card and document can be on the screen at one time. Text fields are limited to 30 000 characters, so long documents need to subdivided into chunks this size or smaller. [17]

3.8 The Basic Elements of HyperCard

HyperCard is formed from a number of basic elements. The basic building blocks of any HyperCard application are:

• Stacks • Backgrounds

• Cards • Fields • Buttons

As highlighted above disorientation is one of the major problems of a hypertext system, and thus HyperCard incorporates a number of navigation aids, which are available to help the user in difficulty.

Chapter 3 - Hypertext and HyperCard

• The Home Stack

• Recent • The Help Stack • The Message Box

3.9 The Building Blocks.

3.9.1 Stacks

Stacks are collections of cards, complete with its specific arrangement of backgrounds, fields and buttons, that are organised to store and retrieve information. Stacks are analogous to disk files created by other computer programs.

3.9.2 Backgrounds

Backgrounds are used to define buttons, fields, and graphics that will be common to a whole series of cards.

This facility is particularly useful, when creating a system, as it allows a collection of objects to be presented in exactly the same position over a number of card, saving a great deal of time during construction.

3.9.3 Cards

Cards represent unique collections of buttons, fields, and graphics arranged on an

area the size of the original Macintosh screen. In a similar manner to text, cards can be cut, copied, and pasted, however a card does not have to be selected, as commands affect the card currently visible on the screen.

The major appeal to the user of cards is their resemblance to the familiar printed page, being a set area of text or images which requires some action to transfer to the

next cards. The challenge to the designer of a HyperCard system is to express the information they wish to provide the user in the confides of a cards area.

3.9.4 Fields

Fields are containers for storing text and numbers. Text in fields can be selected, cut,

copied, and pasted. Fields come in several styles, which can be selected depending on what best suits the purpose of the application.

The flexibility of fields, greatly helps in the authoring process, enabling a designer to

34

Chapter 3 - Hypertext and HyperCard

freely choose the style, size and shape of a field, in order to best fulfil its purpose. Another advantage is that if changes need to be made during construction, moving and adapting fields can be easily achieved

3.9.5 Buttons

Buttons are most often used to initiate actions when the mouse is clicked while the mouse pointer is on a button. Buttons come in seven different styles; basically rectangular in shape, and can be labelled with text or icon to give an indication of its

action.

Buttons are one of HyperCards most powerful tool. They allow the designer of a system to create facilities which the user can easily access, with just the click of a mouse. However, careful thought has to be given by the designer to the appearance of the button in order to express to the user its function. Buttons that are easily identifiable and whose function is clear, allow the user to navigate and operate the stack with confidence. No matter how impressive the facility a button offers, it is ineffectual, unless the user is aware that it is available. For instance, at one exhibition of HyperCard, typical novices expressed bewilderment at the home button, and regularly asked what the "little house" was for. [18]

3.10 The Navigation Aids.

3.10.1 The Home Stack

The Home Stack is a special HyperCard document. It contains buttons for all the stacks that are included with HyperCard when supplied. The User Preferences Card is of particular interest.

3.10.2 User Preference Card

The user level is set by this card. Each level adds powers to the abilities of the previous level. The five levels and their respective powers are as follows:

35

Chapter 3 - Hypertext and HyperCard

'1im:I!J"7\~;; -'" I'" . I,. I" .~,'" .

information no text entry or editing .

• films text entry and editing on cards.

access to the Painting Tools.

access to Button and Field Tools.

access to

Table 3.1 - The Five User Levels of HyperCard

3.10.3 Recent

When selected, you get a display which shows a miniature versions of the last 42 cards you have visited. Clicking on any of these cards takes you directly to that card. Recent can be found in the Go menu, or by typing COMMAND-R.

For a well planned HyperCard system, this amenity should not be required. However, should an individual get lost or disoriented, this allows the return to a recognised point. From an authoring point of view, this facility is particularly useful where cards have been planned without navigation facilities, allowing the designer to move easily between cards.

3.10.4 The Help Stack

This is really a separate HyperCard stack. The stack contains introductory matter about HyperCard, and can be reached by a number of ways. From the Go menu, typing the Level COMMAND-?, and in some stacks help buttons, can be found which

take the user directly to the Help stack.

While this is useful when initially becoming familiar with HyperCard, a designer of an

application should provide a help system specific to their stack or stacks. Being sent to the HyperCard Help, the user may not know where they are or how to get back.

3.10.5 The Message Box

The Message Box is a small window into which you can type short commands. These commands can be call up simple or complicated actions. The box can be summoned via the Go menu or pressing COMMAND-M.

36

Chapter 3 - Hypertext and HyperCard

This is very useful when developing a stack, as it allows the quick and easy exploration ofHyperTalk commands, as well as navigating between cards and stacks.

3.11 HyperCard Menus

The User level, discussed previously, controls the appearance of some of the pull down menus from the menu bar across the top of a HyperCard card. Many of the items on the HyperCard menus have COMMAND-key equivalents.

3.11.1 The Edit Menu

This features typical text editing commands as well as commands specific to objects in HyperCard. When creating a stack these commands allow great flexibility to cut, copy and paste objects in or between cards, therefore giving a developer exceptional design power. (see figure 3.1 - Edit Menu)

3.11.2 The File Menu

This performs functions such as allowing the user to open, save, and print stacks. There are also features to print different aspects of a stack. HyperCard was not

designed to be a report generating facility and the printouts produced are very basic in nature, with little option for customisation. It is in this menu that the quit HyperCard command resides. (see figure 3.2 - File Menu)

3.11.3 The Go Menu

The items on this menu are used for navigation within a stack. A properly constructed HyperCard system should not require the user to resort to this menu. However, it is useful for the designer to be able go to different parts of a stack, easily and swiftly. (see figure 3.3 - Go Menu)

3.11.4 The Ohject Menu

This menu is used for manipulating HyperCard objects - buttons, fields, cards, backgrounds, and stacks. (see figure 3.4 - Object Menu)

3.11.5 The Tools Menu

HyperCard uses different tools to present you with different functions. The first row of icons concerns objects. For looking at stacks you always use the Browse tool. For creating and working with buttons, the Button tools is used, and similarly the

37

Objects Button Into ... Field lofo ••. Card I nfo .. . Bk:gnd I nfo .. . Stack: I nfo .. .

Bring Closer g€;+

Send Farther W-

New Button New Field New Back:ground

Figure 3.4 - Objects Menu

Chapter 3 - Hypertext and HyperCard .

Figure 3.5 . Tools Menu

Chapter 3 - Hypertext and HyperCard

Field tool for working with fields. Several tools are available for painting and graphic manipulation.

One of HyperCards' key features is the ability to include graphics within a stack. The tools menu provides the means to create and customise graphics, depending on the requirements of the designer. (see figure 3.5 - Tools Menu)

3.12 HyperTalk

HyperTalk is the programming language of HyperCard. It is relatively jargon free and the English-like syntax forms lists of instructions, which are carried out one at a

time. These "lists of instructions" are referred to as scripts. HyperCard objects such as buttons and fields can communicate with each other or with HyperCard itself by using scripts to create and send messages. Messages are ordinary English words that are intercepted and acted upon by message handlers.

One card can be linked to another even if that card is in a different field or stack. With help from HyperTalk, scripts can be created that allow browsers to go from one word to a linked card with a click of the mouse button.

3.13 System Design Considerations

3.13.1 Audience

In the planning stage of the a system, it is important to try and define the group, who will be using it. . This is in order to gauge the quantity and quality of information provided, and the method of searching it, at the correct level.

3.13.2 Presentation of Subject Matter

HyperCard presents information by means of an index card metaphor. Stacks are made up of a number of cards, each card consisting on one screen of information. The visual style and presentation of information, will effect users perception of the

system.

3.13.3 Consistency

Consistency within a stack is essential. When the look, usage and behaviour is the same throughout, it gives the user an integrated and stable impression of the stack, and makes it easier for the user to focus on the content of the stack. This lessens the number of errors and helps the user develop an accurate system model. Consistency

39

Chapter 3 - Hypertext and HyperCard

within and between procedures, reduces the volume of knowledge that has to be

assimilated.

3.13.4 Card Layout

A consistent card layout gives the user a feeling of stability and reduces the teaching

required. As a screen cluttered with information will only confuse the user, layout

should be as simple as possible.

It is best to try and maintain logical and functional relationships between the items on

any given card. As cards contain more than one piece of text, the user may find it

difficult to know what to read first. As the English language is read from left to right,

beginning at the upper left hand corner, it makes sense for the most important

elements to be structured following this convention.

3.13.5 Text

Text can be hard to read on screen. Legibility can be decreased by line length; lines

that are too long, reaching edge-to-edge across the screen, or too short, breaking

sentences every one or two words, are difficult to read. Long sections of text, should

be placed into scrolling fields occupying no more than half the width of each card.

Choice of font and font size has a big impact on the look of a stack. Screens should

be limited to the use of one or two fonts, with different sizes and styles providing any

necessary emphasis. Use of multiple fonts only serves to confuse users, making the

system difficult to use. The use of capital letters is best avoided except for a word or

two of emphasis. It is much easier to read a mix of upper and lower case letters.

Similarly plain text is more legible than italicised text.

3.13.6 Independence of Screens

As far as possible, screens should be made to be independent of each other. Users

see each card separately and so explanatory text on screen must be able to stand on

its own, with no reference to previous screens, that the user mayor may not have visited.

3.13.7 Visual Effects

Visual effects can be used to support the metaphor or navigation method of a stack.

HyperCard provides a wide variety of visual effects, but use of too many different

effects in one stack can confuse and disorientate the user. "Wipe left" can give the

40

Chapter 3 - Hypertext and HyperCard

impression of turning a page, as it reveals the next card as if it were underneath the

first one. In the same way "wipe right" conveys a sense of moving backwards

through a book.

3.13.8 Feedback

Users require feedback, to show that the system is carrying out the requested action. Facilities available in HyperCard include the use of the auto-hilite facility for buttons,

which inverts its image when the user clicks on it. The cursor can also be set to a spinning beach bailor a watch to indicate action is being taken.

3.13.9 Use of Sound

Sound can perform a variety of functions within a stack; provide transitions, illustrate content, convey progress, give feedback and provide a continuous environment for

the stack. However, sound can become distracting and irritating, and to an audience accustomed to sophisticated sound effects from television, an amateurish sound effect

can effect the whole tone of the stack.

3.13.10 Stack Opening Sequence

First impressions are important when using a new unfamiliar system. There are two common way in which stacks open.

• static opening:

• attract mode:

which starts with a title screen, followed by the main

menu, whose first item may be the introduction.

in which the stack opens to a rolling, inviting, animated, possibly audible display that attracts the user.

Experienced users need the option to by-pass introductory information required by

novice users.

3.13.11 Introduction and Help

The help function of a stack explains how to use it and how to navigate through it, at

specific request from the users. The help facility should be available from anywhere

in a stack.

3.14 Evaluation

It is important to involve users at all stages throughout the design and implementation

41

Chapter 3 - Hypertext and HyperCard

process. The main aims of evaluation are:

• to ensure that the required tasks could be accomplished • to ensure that no obvious errors had been made • to ascertain any necessary improvements • to ensure the system meets the requirements ofthe intended users

3.15 The Use of HyperCard for The Science Fiction Pilot

3.15.1 Availability

As mentioned previously, HyperCard was supplied free with every new Apple Mackintosh, making the software package one of the most widespread and easiest systems to obtain of the hypertext and hypermedia systems available. It is also the only hypertext system available in the Department of Information and Library Studies at Loughborough University.

3.15.2 Authoring

HyperCards' authoring tools allows the easy creation of objects such as cards, buttons and fields. The relative jargon free nature of HyperTalk and it close resemblance to basic English instructions, allows system construction to be achieved comparatively easier and faster than would be possible in another form of database.

3.15.3 Ability for user to browse

One only has to observe the behaviour of individuals roaming among shelves in a library or bookstore seeking a work of fiction, to see that the process is rarely one of systematic searching or immediate selection. Individuals are often only aware of the book they want when they have found it. Rather time is spent browsing the shelves, scanning the works available and selecting likely sounding titles.

Therefore an index for fiction must acknowledge this method of searching and somehow incorporate it into the core of its structure. HyperCard is well suited to this challenge as it is designed to help create applications which users can browse

through. This is emphasised by the user levels available on the User Preference card.

3.15.4 Cross Referencing

Unlike other databases, where searches of the amassed information rely on knowing what you require; hypertext can be developed to link "chunks" of information

42

Chapter 3 - Hypertext and HyperCard

together allowing the user to jump form one section to another. These links can be created by whatever criteria the author wishes, giving great flexibility. The users is able to browse the data available, until a suitable item is discovered. This mirrors the process normally used in selecting works of fiction, and yet allows lines of guidance to be incorporated.

3.15.5 Flexible Interface

Because of this flexibility in linking information the user interface, being highly visually oriented, is similarly adaptable to be tailored to the designers needs and that of the system. This is especially useful as for a fiction index there is no guarantee the user is computer literate and therefore the system can be designed to be attractive and easy to use, without being as intimidating as some other database interfaces may appear. Therefore an individual wishing to explore possible themes within science

fiction need not be put off by the system and can use the tool to to its fullest extent.

43

Chapter 3 - Hypertext and HyperCard

References

I. Baird, P. HyperCard opens an electronic window on Glasgow. The

Electronic Library, 1988, 6(S), p.344.

2. Fiderio, F. A grand vision. Byte, 1988,13(10), p.237. 3. Baird, P. HyperCard opens an electronic window on Glasgow. The

Electronic Library,I988, 6(5), p.344.

4. Ibid, p.344. S. Fiderio, F. A grand vision. Byte, 1988,13(10), p.238.

6. Baird, P. HyperCard opens an electronic window on Glasgow. The

Electronic Library, 1988, 6(S), p.34S. 7. Ibid., p.34S. 8. Ibid, p.34S. 9. Wilson, E. Link into hypertext. Vine, 1993,91, p.6. 10. Baird, P et al. What is this thing called hypertext? Aslib Information, 1990,

18(10), p.3IS. Il. Ibid,p.3IS.

12. Ibid, p.3IS. 13. Baird, P. HyperCard opens an electronic window on Glasgow. The

Electronic Library, 1988, 6(S), p.347.

14. Fiderio, F. A grand vision. Byte, 1988,13(10), p.244. IS. Langford, D & Brown, P. Creating hypertext documents: is it worth the

effort? Asli b Proceedings, 1993, 45(4), p. 92. 16. Fiderio, F. A grand vision. Byte, 1988,13(10), p.244.

17. Michel, S.L. HyperCard The Complete Reference. USA: Osbome McGraw-HiIl, 1989.p.ll.

18. McKnight,C et al. Problems in hyperland? A human factors perspective.

Hypermedia. 1989,1(2), p. 174.

Bibliography

Goodman, D. The Complete HyperCard 2.0 Handbook 3rd Edition. USA:Bantam Computer Books, 1990. Michel, S.L. HyperCard The Complete Reference. USA: Osbome McGraw-HiIl, 1989.

44

Chapter 4 - Designing the System

4 Designing the System

4.1 Introduction

The aspects involved m designing The Science Fiction Pilot comprise of, what information is to be included and how it will be selected; how this information is to be presented and how the user will be able to navigate around the data.

4.2 Classification of Science Fiction

The basis of The Science Fiction Pilot is to help the user select science fiction works by themes. This requires a scheme in order to allocate themes to works which will be added to the system. The most comprehensive fiction classification system derived most recently is the previously discussed AMP classification scheme by Pejtersen.

Not all the number of dimensions provided by Pejtersen in her AMP classification scheme, have sufficient relevance to science fiction works to be useful. For example, in the dimension of Frame the subdivision of time would certainly not give you any clues to the plot, the nature of science fiction being that it break the bounds of traditional time frames.

However it is useful to have some framework with which to try and draw out some of the main themes of a text. With modification the AMP does suggest some areas with which to start. While it would be extremely convenient to summarise a text with a single word it is impossible to allocate a theme which can encompass the whole meaning of the book. Therefore a number of non-exclusive themes would be selected. For simplicity, themes would not be considered as hierarchical in nature.

Important to flow of the plot from the dimension Subject Matter, is the subdivision: Action. This enables the reader to know where the source of the drama of the story is derived from.

Adapting the subdivision: Place, from the dimension Frame, would help to give some

45

Chapter 4 - Designing the System

idea of the Environment in which the plot of the book is taking place. In science fiction as the author has a completely free reign in their choice of setting, this is often chosen to have a strong influence on how the plot and characters interact.

Also from the AMP classification the Dimension: Author's intention and the subdivision emotional experience, would be helpful to highlight horror or amusing works. This could be expanded to the more general description of Mood.

In order to cope with the special characteristics of science fiction, other areas must be taken into consideration:

As the name applies, science fiction plots involve reference to some development in science, real or imagined. Scientific ideas therefore forms vital area with which to

categorise a book. Science would try to include familiar scientific and technological aspects but which may have been developed along imaginary lines, as well as areas which are visionary speculation. Hopefully, the title of scientific ideas would conveys this mixture of fact and fiction which neatly summarises the basis behind many science fiction novels.

As well as science and technology, science fiction also includes speculation about other facets of society and theoretical concepts such as politics and religion. Therefore, Ideas needs to be included as an area to help classifY a text.

Characters need to be included as due to the nature of science fiction they may not necessarily be human, and whose essence may have a strong bearing on the plot.

These adapted and new categories can be amalgamated with the themes devised in the The Encyclopaedia of Science Fiction, to provide a set vocabulary in order to allocate themes to science fiction works. (see Appendix 1 - Themes) It would not be necessary to find a theme in each category; in fact the most relevant themes maybe selected from a single dimension.

The Encyclopaedia of Science Fiction was chosen due to its considerable authority with in the genre. Since its first publication in 1979, rivals reference sources have not come close to replacing it, in scope or thoroughness. Therefore, the theme list compiled in this publication can be considered the most comprehensive and reliable, and a sound basis from which to incorporated into the classification scheme for this index.

46

Chapter 4 - Designing the System

4.3 Problems of Classification

As discussed in Chapter 2, an indexer has to be well versed in the material they are using. Once a book has been read is is easy enough at that point to allocate themes,

being conversant with the plot. However this is a time consuming process. For the

Science Fiction Index there is not sufficient time to read all the works to be included in the test data. Therefore, a way needed to be found to gain satisfactory knowledge

of the book, to see its main themes, in a quick and efficient manner.

Reviews seemed the ideal solution to this problem, as they included a synopsis of the

plot, and often allocated themes themselves, in order to interest the potential reader.

The quality of the review is dependent on it writer. It is important to find a source of reviews that had both authority and reliability. Sources investigated include.

4.4 OCLC FirstSearch: BookRevDigest.

The OCLC FirstSearch service is a collection of 18 electronic bibliographical

databases covering books, journal articles, theses, computer software and other types of material. Included in the 18 databases covered by this service is BookRevDigest which contains reviews of fiction and non-fiction books. Data is collected from

approximately 100 periodical published in the USA, Canada and Gt Britain; covering

over 7,000 adult and children's books each year. Reviews are from January 1983 to

the present and are updated monthly.

The database allows searching by subject, title and source; providing a list of suitable matches in a source, title and year format. Each entry can then be examined more

closely to provide such information as author, publisher and often an abstract of the review. Especially helpful for this dissertation is the inclusion of fiction themes

definition with in each record. The main one stating science fiction, a second the

main theme of the book. Unfortunately there is no list of the fiction themes used. Also it is impossible to attempt a Boolean search and state science fiction followed by

a secondary fiction theme.

Reviews are selected from publication such as; The Time Literary Supplement; The New York Times Book Review and the Library Journal. This has the advantage of

providing a number of reviews from different sources. However the publications used do not include any specialist science fiction magazines, so the inclusion of

science fiction books is somewhat erratic.

47

Chapter 4 - Designing the System

4.5 The Lysator science fiction book archive (http://www.lysator.liu.selsf-archivelsf-main )

The archive of Science Fiction & Fantasy that resides at Lysator is a collection of texts that have been saved from the Usemet during the last 7-8 years. It is mostly reviews and lists of different kinds. The index of the archive is in a hypertext format.

At present an experimental attempt is being made to make the archive searchable. Searches are allowed by a key word, which can include author or book title. The main contents include an authors list containing mostly bibliographies and views on their writing, and a books list, listing authors alphabetically and composed entirely of reviews of some kinds ..

As entries are on a voluntary basis, what is to be included and what is not is entirely depends on the whims of the potential reviewers. On the whole the quality of the reviews are very high .. Due to an agreed method of indicating well in advance if an important aspect of the narrative will be given away, these reviews often give full plot coverage. Therefore a judgement on allocation of themes can be made without neglecting an aspect, often missed out in other reviews for fear of spoiling a future readers enjoyment. In a couple of years time as internets use has increased, the probability is that most new books will end up being reviewed by the expanded readership.

4.6 Specialist Science Fiction Magazines

There are a number of specialist science fiction magazine which among other science

fiction oriented articles include reviews.

4.6.1 Foundation

Along with in-depth essays on science fiction, Foundation also includes very detailed reviews of recently published novels. These are often written by another science fiction author or an influential name with in the field, making them highly authoritative. However, the main focus of the publication is not to provide book reviews and only a select few are covered in each monthly issue; short fiction being rarely covered.

4.6.2 Vector - The Critical Journal of the BSFA

Similarly this journal includes essays and articles about science fiction issues, but contains far more book reviews. The reviews are concise and informative, however

48

Chapter 4 - Designing the System

the journal review short fiction only briefly and not to any great depth.

4.6.3 Locus - The Newspaper of the Science Fiction Field

This American publication contains articles about authors, conventions and anything else of interest with the "industry". Mainly however it contains reviews, covered in separate sections by half a dozen reviewers. This is advantageous in that the same book may be criticised more than once, giving another opinion on its themes and expanding knowledge of the plot. .

As well as novels the magazine also has a reviewer who deals only with short fiction. The section covers magazines such as Isaac Asimov's Science Fiction Magazine,

Fantasy & Science Fiction, Interzone and Analog. The reviewer shrewdly selects which magazines have a high calibre story or stories, so the selection of magazine

changes monthly.

4.7 Review Source Chosen

It was due to this comprehensive coverage of all science fiction formats that Locus magazine was chosen as the main source of reviews. The short fiction reviewer demonstrated his skill by selecting most of the Nebula and Hugo award nominees to cover in his column.

4.8 Advantages of using reviews

4.8.1 Speed

While the duration of time to read a single item of short fiction may not appear to be very great, if this process is considered accumulatively then it is still a time consuming process. Thus for all formats of fiction reviews allow a quick method of allowing the indexer to become acquainted with the plot, and hopefully its themes.

This makes it feasible to include large numbers of entries into the system, with out having to physically sit down and read each text before attempting to include it.

4.8.2 Range of Viewpoints

Every person has there own preferences to different aspects of stories; the character development, the exploration of new ideas, the action of the plot. Thus when reading a text it is often this aspect which is perceived as the one of importance. So, while one reviewer may praise a certain book, another writer will criticise it. Using reviews

49

Chapter 4 - Designing the System

therefore, overcomes the danger of personal bias in extracting themes. The more reviews that can be found therefore, the better. For novels, reading of three or four

reviews is still substantially quicker than reading the whole book.

4.9 Problems of using reviews

4.9.1 Personal bias ofreviewer

As previously stated every persons perception of a story is coloured by their own personal biases. The skill of a good reviewer is to try and overcome this and take a wider viewpoint; however this requires broad subject knowledge and experience. Therefore, a single review must be regarded cautiously, and if possible compared with a second to try and establish a consensus.

4.9.2 Deducing plot from review

Using reviews, you become reliant on the skills of the reviewer to confer the plot of text. To avoid giving too much away and spoiling the story writer are often deliberately obscure. Conversely they can provide too much detail, so the reader is unable to decide the wider themes that may be covered. To illustrate this compare the two sections of reviews from the same publication, about the same book; loan D Vinge. The Summer Queen.

The Complex events of The Summer Queen begin some eight years after Moon's accession. On Foursgate, BZ Gundhalinu struggles to understand the mutated smartmatter plasma which can restore the Hegemony and Kharamough to Imperial power. At Razuma on Ondinee, the criminal Brotherhood plot the theft of the stardrive, employing Reede Kullervo, who appears to be the avatar of

Vanamoinen, the Old Empire creator of the Sibyls and Sibyl net. [1]

While the second review starts:

The Summer Queen continues the story of the beautiful planet Tiamat, whose only valuable export is the water of life, a liquid distilled from the blood of the mers that grant eternal youth -immortality. Moon, the Summer Queen of Tiarnat, is relieved when the Black Gate closes and the Hegemony can no longer make it to Tiamat.[2]

so

Chapter 4 - Designing the System

4.9.3 Deducing themes from reviews

While some review writers will highlight themes in their articles, mostly they have to be deduced from the synopsis ofthe plot. This can be difficult, as the review does not give the complete flavour of the work.

4.10 Schedule for Classification of Science Fiction

• Receive book to be entered into The Science Fiction Pilot

• Find review or reviews published in Locus

• Find reviews in the Lysator Science Fiction Archive

• Allocate themes, guided by the six selected areas

4.11 Structure and Scope of The Science Fiction Pilot

4.11.1 Scope

In the space of time allowed for this dissertation it can not be hoped to construct a truly comprehensive collection of science fiction works. Limiting factors must be brought into bear to help whittle the data down to manageable proportion and yet allow an illustration of the objective undertaken.

Science fiction as a genre covers a variety of media; films, videos, artwork as well as the more commonly known printed works. Even in the printed media, the rapid emergence of graphic novels has created another category with in science fiction. However, this system will be limited to covering purely textual fiction; including both novels and short fiction.

Short fiction is considered to be the life blood of science fiction; where the exploration of ideas which is instrumental in this genre, is at its most fruitful. However, such fiction is mainly found is specialist science fiction magazines, mentioned previously in the Locus appraisal. This provides added complications, as an index is virtually useless if it informs the user of material which they are unable to obtain. Thus the index must not only identitY the story but also the publication in which it appeared.

4.11.2 Size

Constrained by time in which to complete the dissertation, it was decided to use a list of works nominated for awards. These would provide a list of quality fiction, with a representative range of science fiction themes, including the number of different

51

Chapter 4 - Designing the System

formats that science fiction is available in; novel, novella, novelette and short story. Science fiction has a number of awards such as Hugos, the John.W.Cambells and

Nebulas. The nominees and winners are selected by a different methods, which leads to criticism or loyalty to these awards.

For this system, both nominees for Hugo and Nebula awards were selected to provide the test data. Nebulas have a reputation for going to experimental work, being voted for by the Science Fiction Writers Association (SFW A) , while Hugos are voted for a one of the large science fiction conventions; the WorldCon and have been considered too popularist. In reality both nominee lists have considerable overlap, but have been chosen to help select books from all spheres.

A further limiting measure is to use nominees for books published in 1992; in order that these particular works can be studied in some depth.

To add another dimension to the data as a single years work was considered somewhat limited, one of the award nominated authors; James Morrow, was selected to have a retrospective of his work included. The author was chosen as his work encompassed a range of themes, and again because of time constraints, was of manageable quality.

4.12 HyperCard Design Considerations

In designing The Science Fiction Pilot using HyperCard, there were a number of considerations that were taken into account during the design process.

4.12.1 Audience

The potential users of this system were considered as individuals who had no knowledge of science fiction or rudimentary experience of the genre. Therefore, in order to cater for this range of ability in the audience, information and search capabilities would have to be stratified. Users should be able to chose a simple theme search process and basic information, with the option of a more elaborate searches and data being available.

4.12.2 Visual Style

One of Hypercards' great advantages is its ability to be able to include graphics;

whether created via the paint tools, or copied and pasted in. A system dealing with science fiction cries out to have graphics incorporated in its structure. However, a balance needs to be found between cliche, and beneficial images which can be used to

52

Chapter 4 - Designing the System

encourage and attract a user.

It was decided to use images from the Dan Dare comic strip, to convey a feeling of

fun, and make the system more enjoyable to use. Images would be confined to the background of facilities, so as not to interfere with the information being presented.

In the Introduction to the system however, to increase the facilities approachability,

the images would play a stronger role.

4.12.3 Visual Effects

For The Science Fiction Pilot, the number of visual effects were kept to a minimum.

One effect was consistently kept for a specific task, such as transferring between facilities, where the effect tried to instil a spaceage feel. For navigating between cards, the effect was kept more simple, imitating the turn of a page.

Some functions within the system took a few seconds to accomplish. As a user may grow concerned during this wait, a screen incorporating a message to be patient, was

displayed. The image reinforced the comic book visual style of the system.

4.12.4 Use of Sound

Even though HyperCard can offer a variety of sounds to integrate with a stack, these effects can become distracting and irritating, and it was decided to restrain its use to

the opening sequence, where it would enliven the simple animation.

4.12.5 Text

Making the textual information of the system clear, was imperative. The function of The Science Fiction Pilot revolved around the user making a number of selection

from text within fields, and therefore it needed to be easily comprehended. It was decided to make clicking on text instigate action, rather than buttons, in order to give

the users as wide a selection of choices as possible. Scrolling fields could be used, to

accommodate lists of themes or titles, allowing users a greater variety to select from.

4.12.6 Buttons

The use of buttons was limited to navigating between the facilities within the system. Care was taken to try and chose icons that suggested the quality of the facility it

represented. However, text describing the facility was also included within in the

button so its function was clear, and the user would be confident in using it. This combination of text and graphic was based on Muter & Myson [3] find that adding

53

Chapter 4 - Designing the System

icons to standard textual menu pages reduced user errors. It was suggested that the icons improved accuracy as semantic meaning may be extracted more effectively from graphics than words.

4.12.7 Card Layout

A consistent card layout gives the user a feeling of stability and reduces the teaching required. As a screen cluttered with information would only confuse the user, layout was kept as simple as possible.

As cards contained more than one piece of text, the user may find it difficult to know what to read first. As the English language is read from left to right, beginning at the upper left hand corner, it makes sense for the most important elements to be structured following this convention. A header, displayed the screens function to help orientate the user. Buttons were oriented at the bottom of each card. Similar to reaching the bottom of a page before turning to the next, the user will see these buttons control their movement to a different screen.

The largest element on a screen, will draw the eyes of the user and instantly gain attention. This effect was utilised by making the most important element on any screen the largest one.

4.12.8 Navigation

The system is constructed with the intention that the user makes a selection and is then taken to the appropriate card. Selection choices related to the fiction was by

text, indicating a author, title or theme. This was to imitate the way conventional indexes operated where the user looks up a term, and is given guidance where to find the relevant information. Spatial navigation was by buttons located at the bottom of the screen. Of all the button options available the rounded rectangle was selected as the stereotypical form of an electronic button. Even a user unfamiliar to computer systems would be able to guess its purpose. It was decided to use a combination of icons and text to illustrate the buttons function. As the purpose of some of these buttons did not have an image which would have a strong association, the icon selected was supported with a single word description of the button. Having an icon was thought to help regular users, as the image is more prominent on the button, and more unique than line of text to make at a glance recognition easier.

4.12.9 Organisation of Information in Stacks

The Science Fiction Pilot was composed of a number of stacks. Information was

54

Chapter 4 - Designing the System

separated into stacks, to allow users to browse through certain sections without stumbling upon an unexpected screen. For the designer, it allows different backgrounds to be created and changed without effecting another section.

4.12.10 Evaluation

To ensure that the design decisions made had created the expected effects, informal reviews were carried out with colleagues, who fulfilled expectation of the intended audience, throughout the building of The Science Fiction Pilot. Individuals were asked to use the system and their suggestions and difficulties noted, and acted upon. Fortunately, HyperCard allows changes to be easily made to the design of the system.

55

Chapter 4 - Designing the System

References

1. Morgan,P. Highest Order SF. SFRA Review, 1992, 198, p.69.

2. Hellekson,K. Another Queen. SFRA Review, 1992,196, p.79.

3. MacGregor,J.N. A comparison of the effects of icons and descriptors in videotex menu retrieval. International Journal of Man-Machine Studies,

1992,37, p.768.

56

Chapter 5 - Implementation ofthe Design

5 Implementation of the Design

5.1 Introduction

The construction of the system was driven by, what the user would require from it, what tasks it could accomplish for them and how they were likely to approach these tasks. As discussed the target audience is envisioned as novices to the science fiction genre, and those with a very rudimentary experience of it. Possibly the system would provide a novice with elementary experience. Therefore it was necessary to incorporate a way to cater to both groups, using a filter of some description in the implementation. As well as experience of fiction, the users experience of computing must also be taken in to consideration. A HyperCard system was chosen for its flexibility in allowing the implementation of an approachable interface.

5.2 Structure of the System

The system was divided into five sections:

• Opening sequence • IntroductionlHelp

• Themes • Fiction

• Glossary

Splitting the system into section allowed greater flexibility in designing the layouts of the screen for the different functions. A different background image could be given to different facilities, to establish a visual style for the stack, as well as help the user orient themselves within the system. The use of background fields and buttons saved time in constructing the facilities composed of identical cards, and helped maintain consistency in the system.

57

Chapter 5 - Implementation of the Design

IIntroductionlHelpl

I I Opening Sequence 1--- L-_M....;,a;.,in_M....;,e_nu;;.... ...

I 1..1 .....;U;.;;s.;;er...;L;;.;e..;,ve;;.;l.;..s ...... I l Theme Searches I --- L._.....;F;.,;i,;;.ct;;;;io;;,;n;;.... ......

L r-I -I-nd-ex--'I J Figure - 5.1 The links between facilities of the system

5.3 Opening Sequence

A short animated sequence starting from a static screen was chosen to open the

system. While something dynamic was needed to attract the user, practical

considerations, such as saving disc space for data, had to be taken into consideration.

A commercial package installed on hard disc would not have this problem. There is

also aesthetic considerations, that keeping the sequence short and simple ensures it

does not become irritating to continual users. It was decided to employ the systems

name as an integral part of the animation, to reinforce its function to the user. The

form of the animation tried to establish the visual image and theme of the system, as

well as entertaining the user so they felt confident about using the stacks. Also while

static scanned images look professional, it requires a great deal of care to manipulate

them in an animation, and not to end up with an amateur look. As the sequence was

so simple a short sound effect was used to liven it up.

The opening sequence finishes at the Main Menu. ( Figure - 5.6) This screen was

envisaged as the reference point which users could return to, and be confident they

could navigate from. It was attempted to keep this screen as simple as possible, with

the minimum of navigational button displayed and only three choices of destination:

• Introduction • User Level I: basic

• User Level 2: advanced

5R

Chapter 5 - Implementation of the Design

As a novice, the user had the option of choosing the Introduction, in order to help them understand the use and purpose of the system. An established user is able to ignore this facility and start their search immediately.

It was attempted at this point to establish layout consistency for the user. The Help button is positioned at the left hand side, while the Back button is on the right hand side. Even though at this point the user had no where to go back to, a way to hide the back button on the first visit could not be found.

As well as an opportunity to go to the Introduction, the user had the option of two user levels. These were displayed as two fields of descriptive text and two buttons. This was the method devised to try and cater for the anticipated abilities of the audience. The levels were numbered as well as being described as Basic and Advanced, to further differentiate them. A short sentence was also included to help the user decide which category they belong. There was some difficulty in choosing an icon which would be able to clearly represent user level.

As well as the buttons, the fields containing the descriptive text, were made clickable to ensure a user unfamiliar with the HyperCard convention, would enjoy an early success in their first choice, gaining results. Failure to gain a response at this point may cause a user to abandon the system.

The user level menus were differentiated by the level of theme search they offered. For example, User Level: I Basic provided the option of the basic theme search. (Figure - 5.7)

The Main Menu has a Help button, so the user can go straight to this if they prefer. There is· also a Quit button, as the Main Menu button is on all other cards in the system, the user is only two clicks away from leaving the system. To limit accidental quitting, an on screen message appeared giving the user the chance to cancel the operation. The decision to limit the quit button to the Main Menu and Index facilities was made, as at these destinations, it was thought the user may wish to leave the system; at the Index facility having finished a search.

5.4 IntroductionlHelp

As the Introduction and Help facility both explain aspects of the system it was decided to combine these function in one stack. This would not be visible to the user. This demonstrates the benefits of a hypertext system where links can be created between any areas of information.

59

Chapter 5 - Implementation of the Design

The Introduction started with a welcome card, to reassure the user they had been taken to their destination. This screen also included instructions on using the next button, for navigating through the system. The button giving access to the Main Menu was on the IntroductionlHelp cards. This screen was followed by one describing how the system used themes to find fiction, so the user understands how best to employ the facility. ( Figure - 5.2) The position of the navigation buttons; Next, Back and Previous were identical to where the user would find them in the system. Then a map of the process followed. ( Figure - 5.3) This showed the

various sections of the system and how they interact, illustrated with the buttons the users will meet. The curious user would find clicking on these buttons would lead them to explanations of their use within the system. This was not indicated as a possibility, as it was thought this would confuse a user unfamiliar with hypertext too early on in the introduction. It was hoped the map would give users a mental model of the system and prevent disorientation later.

Explanations of each aspect of the system had the button which was involved in the process, in the top left hand corner, in the hope this will reinforce its' use. (Figure-5.4 ) In a similar manner to the system map, if a button was contained in the explanation, clicking on it by a curious user would take them to its functional description. It was hoped to keep explanations limited to a single card. If information was spread over a number of cards, this was indicated at the bottom of the text. The Introduction finished telling the user to return to Main Menu by clicking on the appropriate button.

As stated, the Help facility used the same cards as the introduction. When the user clicked on the help button they were taken to a menu of topics. ( Figure - 5.5 ) These were displayed as text, in case the user found the use of icons confusing. Clicking on a topic sent the user to the appropriate section.

5.5 Themes

The system was based on a number of themes being allocated to each work. The allocation process is discussed in the previous chapter. The system then allowed a user to search for works by choosing a theme that appeals to them.

The two user levels each have a different theme search:

• The Basic Search

• The Advanced Searches

The same icon was used for both theme searches, being differentiated by the

60

Chapter 5 - Implementation of the Design

" "

TheScie,llce Fictioll Pilot will help ,.,""-'" select stories, by THEMES

: Once you select a theme the Pilot will : provide a list of titles which contain it.

ou can take a closer look at these titlesi to find out what other themes it contains and a synopsis of the story.

Please Click on The Next Button

Back N.xt

Figure 5.2 - Explanation of system in the Introduction section

, '. The Route the Science Fiction Pilot

Main ME-nu Books

IChoosel IChoosel Isends tol

~ ~:;-1*~~ Uso. Lovol I Choose I nomos lsends tol '--_-"'"

IndeH

Glossary

Main Menu Noxt

Figure 5.3 .. Map of system in the Introduction Section

61

Chapter 5 - Implementation of the Design

" .. '

The Main Menu has two useful buttons:

; The Help Button which is on all cards and gives you access to the help facility.

The Quit Button which allows you to quit from the Science Fiction Pilot

Figure 5.4 • A card from the Introduction Section

. ; H Menu

Please Dick a on Topic fm Help

I Navigation Buttons

Main Menu

Themes

Books

Index

Glossary

Figure 5.5 • The Help Menu

62

Back Next

Back

Chapter 5 - Implementation of the Design

Figure 5.6 - The Main Menu

Figure 5.7 - User Level: 1 Basic

63

Chapter 5 - Implementation of the Design

descriptive text below the image. It was decided a different icon would complicate the system and confuse the user, and so the risk of ambiguity was taken. There were

links to the theme section of the system, from all the other areas of the system so the user always had the opportunity to make this search. The basic theme search was taken as the default throughout the system. A button link enabled users access the

advanced searches.

5.5.1 The Basic Search

A novice user, with no knowledge of science fiction, will be reliant on themes to help them select a work of fiction. Therefore the theme section of the basic level: 1 needed

to be simple and approachable. The user was eased into making a selection by presenting them with six categories:

• Space • Characters

• Science • Action • Mood

• Ideas

It was hoped that by making a selection of one of these categories, it would guide the novice users to a theme they would also be interested in. ( Figure - 5.8) The category names were chosen to be simple but attractive, engaging the users curiosity. After this choice the user then found a list of themes under this category. (see Figure-5.9) They were simply required to click on the theme of their choice, which would then take them to a list of titles of work which contain this theme. Even though a glossary is available in this section the vocabulary used was filtered for terms only familiar to a seasoned science fiction reader.

5.5.2 Advanced Search

Selecting this level give the user two possible advanced searches ( Figure - 5.10 ):

• A single theme search • A multiple theme search

The single theme search did not differ greatly from the basic theme search, except the

user did not have to first select a category, as they were presented on the screen with the relevant themes beneath them in a scrolling field. ( Figure - 5. 11 ) The more experienced reader would not be overawed by the number of themes, and would

64

Chapter 5 - Implementation of the Design

probably enjoy browsing through what was available before making a selection. It

was hoped a user who started at the basic level, and had now moved up to the

advanced level, would be familiar with the categories. The themes included terms

that may only be encountered with reference to science fiction e.g. terraforming. At

this level it is hoped the user may have knowledge of these themes or wish to

investigate what they have to offer. Once a theme has been clicked on, the user was

taken to a list of texts.

The multiple theme search had a similar format to the single search format, with the

six categories and themes beneath them. However, the user could click on a number

of themes, and only after clicking on the "Find Themes" button were they given a'list

of those texts containing the group of selected themes. The choices made were

displayed in a field at the top of the screen. A cancel button was available for the user

to abandon the search. If no text matches the selection of themes the script displayed

a boxed message to this effect, and remains at the multiple search screen. As the

system stands the feature is most likely to admit that the selection of themes is not

available, but with a greater selection of texts this would increase the choice the user

has.

5.6 Glossary

As the system was using terms which the science fiction genre has adapted to specity

certain themes, it is important to have access to a glossary. Access to the glossary

was from points where there may be themes; the Theme Search, Books and at the

Index sections. The glossary started with a list of all the terms it contained ( Figure -

5.12). Clicking on a term sends you to that entry. Here the subject was foUowed by

the field containing its definition. There were also cross references to lead the user to

connected terms.

5.7 Index

Choosing a theme sent you to a list of texts at the Index section. ( Figure - 5.13 )

These are displayed by the works title and the authors name. Clicking on a title that

is attractive sends the user to that text in the books stack. If the user accidentally

clicked on a blank line, a message was displayed asking them to click on an item

A print button was available so a copy of the list of texts, which contained the theme

of the users choice, could be kept and used to find the books in a library or bookshop.

The option to cancel printing was given to ensure against accidental use.

65

Chapter 5 - Implementation of the Design

Main Me-nu Books

Figure 5.8 - The Categories of the BasicTheme Sean:h

Themes

Figure 5.9 - The List ofThemes fora Category

nn

Chapter 5 - Implementation of the Design

Figure 5.10 - The AdvancedTheme Sea«:hes

Themes

Figure 5.11· The Advanced SingleTheme Sea«:h

67

Anthropology

Genetic Engineering

Alternative Worlds Anthropology Communications

Chapter 5 - Implementation of the Design

Glo

, ' Colonisation of Other Worlds

Disaster ,: Dystopias

Ecology Fan.tasl:ic Voyages.

Figure 5.12 - The GlossalY

True Faces ( CADIGAN, Pat)

Books

Contact (OLTION, Jeny& GOODLOE, Lee) A Fire Upon the Deep (VINGE, Vernor)

Figure 5.13 - The Index

fiR

Chapter 5 - Implementation of the Design

Authors & Titles

Figure 5.14 - The CoverScreen for the Fiction Information

Authors & Titles

Books Back Ntxt

Figure 5.15 - Fiction Information

69

Chapter 5 - Implementation of the Design

5.8 Fiction

Even though the fiction covered was composed of short fiction as well as novels the section containing information on these were referred to as books. It was thought this term had stronger connotations, rather than the more formal term fiction which would give the user an instant conception about this section, as well as providing a stronger icon.

To ensure the user would be able to browse through this section when the mood took them, access to this section was available from a variety of sources.

Rather than plunging the user into the fiction information cards, the Books button took the user to a cover screen. (Figure - 5.14) This gave a number of ways to gain access to the data.

For simple browsing the best method was to use the horizontal alphabetical index. The user clicks on a letter and was then sent to the first author whose surname started with that letter. It would have been possible to send the user a list of authors under that letter, but this method was thought a better as it allowed the user to browse through the authors and titles stored. The purpose of the system being to help the user find a science fiction works, so while browsing, a title might catch their eye.

Also this horizontal index appears in the cards detailing the author and title information as a form of navigation, and so for consistency this method was kept. This index is best for a user for no knowledge, as it allows simple browsing, but is positioned at the bottom of the screen. To try and pick it out from the other items on the screen the descriptions and instructions are given in a larger field, in a larger font. The eye usually being caught by the largest item on the screen.

There was the option to view all the titles in the system listed alphabetically by author or title. At the moment while there are only a few items of test data in the system this method is quite useful, but as the data grows this may become too cumbersome to be useful.

The cover card also offered the possibility of finding a specific author or title. Even a novice may have read one or two science fiction books and would like the reassurance of finding them in the system. The more experienced reader may also want to see what titles are available from a favourite author, or what themes have been allocated to a favourite title. If the author has more than one title in the system, the button does not list them but takes the user to the first work in the books stack. As they are stacked alphabetically by author the user can then use the navigation

70

Chapter 5 - Implementation of the Design

buttons to work through the titles.

Using the alphabetical index, or clicking on an item in the index, sent the user to the section containing information about the science fiction works in the system. These cards were kept as simple as possible. (Figure - 5.15) The header was given as Authors & Titles, mirroring the conventional information provided by libraries or bookshops and the way most people characterise fiction. The reference points of the Main Menu and Help buttons have been consistently positioned in the bottom right

hand corner. Positioned in the left hand corner were the navigation buttons: Next, Previous and Back. Centre bottom of the screen contained button to go to other sections.

Displayed underneath the name of the author and the texts title, were the themes allocated to this particular work. This was another possible theme search. If an individual liked a certain science fiction text, they would be able to select the theme that appealed to them most. They would then be sent to a list of works containing that theme. This saved the user having to remember which themes have been allocated to a work, and then returning to the theme search section.

If the user finds the title interesting they may like to know more about the plot than the indication given by the list of themes. Also the format of the work maybe useful, as reader may prefer short fiction to novels. For finding the book, the availability of other information such as publisher and year of publication would be useful to know. However, to present this immediately to the user would clutter up the screen with information not immediately necessary to them. Therefore, the data was concealed and only revealed on clicking on the Bibliographical Details button. If the user is still enamoured by the work they can print these details using the Print button. HyperCard is not designed with report making facilities in mind and so the resulting printout is very basic.

If a user finds the style of a certain writer attractive they may wish to find out what other pieces they have written. Clicking on the authors name sent the user to the index and a list of their works.

5.9 System Protection

The HyperCard edit menu contains the protect stack command. This gives a number of options, from being unable to modifY, delete or abort the stack, to restricting the user access to browsing, typing, painting, authoring and scripting. However, using this facility was found to quite considerably slow down the links between stacks. Also the bar menu, through reduced in capabilities, was still given at the top of the

71

Chapter 5 - Implementation of the Design

screen, even for the browsing only protection. It was thought that the bar menu may

confuse users into thinking that this facility was necessary for the use of the system.

The Science Fiction Pilot was protected from alteration by locking all the fields to

prevent editing, and then scripting was used to hide the bar menu. This would focus

users attention on the commands available on the systems cards. An individual with knowledge of HyperCard, could alter the cards, therefore, it is vital to retain backup

copies of the system.

72

Chapter 6 - Evaluation and Conclusion

6 Evaluation and Conclusion

6.1 Introduction

A designer has an intuitive idea about their system, with the implicit assumption that "If

I can use the system, it's obviously easy to use". However, even if a designer has followed all the guidelines available to make a computer system, there is still a need to know how well it has succeeded. [1] The fundamental test of The Science Fiction Pilot,

was to evaluate its success with users. The evaluation would test that the system did not depart significantly from its user requirements of being easy to understand and use, and to make any changes to the design that were necessary. [2]

Evaluation criteria should include the functionality and usability of the system under study. Functionality relates primarily to the general objective of assessing the capabilities of the design, and refers to the tasks that the system enables the user to perform. In the case of The Science Fiction Pilot, its functionality encompasses the users ability to find science fiction texts by themes of their choice. Usability relates to the assessment of the impacts of specific design decisions and refers to the ease of use of the interface. [3] For The Science Fiction Pilot, usability is the ease the user can operate the system, selecting themes and viewing the works from this choice.

6.2 Subjective data

Being attractive to users, was one of the key design issues of The Science Fiction Pi/at, therefore, subjective data was considered appropriate as the systems functionality was centred on assisting users to discovered information. Thus, the main focus was that this task was accomplished in a pleasing and helpful manner to the users; opinions and feelings that subjective methods would be able to draw out.

Subjective evaluation methods are directly based on the user's judgement. The evaluators are asked questions, so that the impact of a design, and the extent to which the user requirements are being met, can be measured. [4] Subjective evaluation methods tend to yield subjective "soft" data about whether the system is comfortable, easy to use, manageable and comprehensible. [5]

7J

Chapter 6 - Evaluation and Conclusion

Subjective data is acquired using questionnaires, rating scales, interviews, and by human

observation. [6] The advantages of subjective methods are those of low cost; ease of

implementation; and an ability to pin-point unstructured problems. The drawbacks are a

tendency to produce exaggerations; the difficulty in avoiding leading questions and a

plethora of data. [7]

6.3 Selection of Subjects

Ten subjects participated in the evaluation. Virzi, [10] found that most usability

problems are could be detected with five subjects. As, well as usability problems the

evaluation was interested in participants feedback, through their opinions, attitudes and

understanding of the system, to ascertain whether it had achieved it function. Therefore,

ten subject provided a wider variety of viewpoints.

6.4 Questionnaires

Questionnaires can be used to elicit user's cognitions about the interface, or self reports

about their own behaviour when interacting with a user interface. Questionnaires usually

consist of a short series of questions on a given topic. [8] This survey method reduces

bias via the experimenter, by standardising questions and automating data collection. [9]

6.S Designing the Questionnaire

The questionnaire was made as brief as possible, consisted of eighteen questions. (see

Appendix 2 - user evaluation questionnaire) It was felt a longer questionnaire would

irritate evaluators, and the quality of response would suffer. The questionnaires was

divided into two sections:

• Information on subjects previous computing and fiction experience .

• Users evaluation of the system.

The first section was designed to establish what qualities and perceptions subjects

brought with them to the evaluation and whether this would effect their ability to use the

system.

The second section was composed of questions which elicited responses on how subjects

had found learning to use the system, and if the presentation of the system was clear and

efficient in providing users with information.

In order to encourage use of the system, subjects were given realistic tasks to undertake.

74

Chapter 6 - Evaluation and Conclusion

The tasks involved:

• finding what works were under a stated theme • finding the works by a specified author in the system • finding a definition of a theme

These were feasible tasks, which could be accomplished in a number of ways by users of all abilities. The tasks enabled the evaluators to see the interface as an overall system exposing any problems, difficulties and aspects of usability. [11] In order to elicit information on what subjects thought about the system questions, were of two types.

Closed questions with ordered choices, where answer choices were provided; each a graduation of a dimension of some thought or action. The respondent's task was to find the most appropriate place for his or her response. [12] And open ended questions, where the respondents had to 'create' their own answers and state them in their own words. These kind of questions allowed respondents to express themselves freely. [13] Open ended questions were restricted to the questionnaire section on the best and worst features, as well as providing the option to make additional comments. This was in order to prevent any influence closed questions may have introduced if used in this section.

6.6 Formal Evaluation Results

6.6.1 Evaluator's Computer Literacy

All the participants were found to have had previous computing experience ranging from one month to two years or more. However, when this experience was restricted to previous use of Macintosh computers, the length of experience fell, with 70% of subjects residing in the range of one month to less than a year. Despite this, all subjects had used HyperCard before.

6.6.2 Fiction Experience

Twenty percent of the subjects stated that they did not read fiction. But it was encouraging that these individuals still enjoyed using the system. Off the remaining participants, 80% were found to prefer period literature.

Feelings for science fiction were found to be quite diverse, with 50% of subjects claiming to dislike the genre, 20"10 selecting the like option, and 10% who claimed to strongly like science fiction. This was a good test of the system, to see how individuals, who would not persevere because of an interest in the subject matter, would find using it.

75

Chapter 6 - Evaluation and Conclusion

6.6.3 Overall reaction to The Science Fiction Pilot

Generally the overall reaction to the system appeared to be very favourable. Of all the responses returned in the second section, 46% were found to be within the easy option and 41 % in the very easy choice. Equally encouraging, was that participants continued to use The Science Fiction Pilot after they had finished their evaluation.

6.6.4 Learning the System

Responses indicated that subjects found it simple to learn how to use the system, with 50% registering very easy, and 40% selecting easy. Similar results appeared for opinions on exploration of the system. This indicated that design decisions had been successful in presenting the system in a effortless and understandable manner. Further support of this was that on screen instructions, were found easy by 50010 of participants and very easy by the same quantity.

Locating the help facility was found to be very easy by 80% of subjects. This was significant, in that even if the users does not need to use help, it is reassuring, especially for a first time users, to know that the facility is always available.

6.6.5 Presentation of the System

Evaluators were divided equally between deciding the function of the screen as easy or very easy. Seventy percent found following the sequence of the screens easy. This was very important to reduce disorientation. It means that the user can easily predict what the next screen will be, strengthening the usability of The Science Fiction Pilot. While 90% of participants found understanding the use of buttons easy and very easy. Comprehending the function of buttons was requisite for navigating between facilities, and was therefore vital to the successful operation of The Science Fiction Pilot by users. The quantity of data provided on each card was found to have been judged correctly, with 100% of subjects responding that there was enough information of each screen.

6.6.6 Best and Worst Features of the System

The evaluators considered the best features of the system to be:

• graphics (60% of subjects) • easy to move from section to section to browse ( 500/. of subjects) • different user levels (40% of subjects)

This signified that design decisions had been found effective by the evaluators. Visual

76

Chapter 6 - Evaluation and Conclusion

style was described as "excellent graphics and well laid out" Users appreciated" features for users of different levels", and that they were able to "move easily from screen to

screen".

The evaluators considered the worst feature of the system to be:

• length of sound effect in the opening sequence (20% of subjects) . • length oflntroduction (10% of subjects)

The sound effect in the opening sequence was considered too long as it continued after the main menu was displayed. Users, in anticipation that something further may happen, were forced to wait for the effect to finish. Steps were taken to try and edit the sound effect to occur only during the opening sequence and no further.

Evaluators found that the length of the Introduction meant that they had forgotten aspects by the time they had finished. This is natural, however, the planned effect of the Introduction was to give users a flavour of the facilities available, and the knowledge they could return for further instruction if necessary.

6.6.7 Additional Comments

Evaluators made some very interesting comments. A request was made for a sound effect when a button had been clicked, to reassure the user a selection had been made. Similarly, further indication for which text can be c1ickable, was suggested. Use was made of the auto-hi lite facility with buttons and text, to try and show when a selection had been made. By scripting it is possible to hilite a button or field when a mouse enters the target, but this was found to be somewhat slow in reacting, as the user sweeps the mouse across the screen. Sound was chosen not to be used, as it was considered that it may become irritating after continual use.

It was pointed out that, there was no indication that the alphabetical index only took the user to the first card of that letter, and that the next button was needed to be used to

access further cards. The script could have sent the user to a list of the subjects under that letter, however, it was hoped having to use the navigation buttons would encourage the user to browse through the cards. Possibly a future development would be to create a graphic which indicates to the user more author cards are behind the current one.

One evaluator found that as the position of the back button moves, it was possible to click back four times, and hit the quit button. The message to confirm that the user really wanted to quit would prevent leaving the system by accident, but appropriate action was taken to correct this oversight. However, the back button and link to main menu were

77

Chapter 6 - Evaluation and Conclusion

noted as vital if the user forgot where they where.

A suggestion that synopsis should be presented on the screen, rather than accessed by the bibliographic details button was made. As this information would indeed be useful for the user to view along side thematic details, appropriate alterations were made.

6.7 Conclusion

Hypertext provides an exciting addition to the traditional means of information access and manipulation. However, to produce an effective hypertext system requires considerable work and a remarkable diversity of skills. In addition to the skills required to process the information to be included in the system, the author requires abilities in user interface design, structuring, evaluation and maintenance.

6.7.1 Strengths

The Science Fiction Pilot can offer a range of potential benefits. It offers the user ease of access to information about science fiction works, which would not otherwise be available gathered together in one source. As access is via an approachable and attractive interface, the system should appeal to a variety of users.

Further advantages that such an electronic system can offer are, the facility to store large quantities of data in a compact and manageable manner, as well as offer the ability to update entries easily and simply. The Science Fiction Pilot would be suitable for use in a library or bookshop, to introduce users to the range of science fiction available in that establishment. Due to the ease HyperCard allows changes in card layout; the information presented could be tailored to what users of the organisations required. For example, a bookshop could include prices, a library shelf codes. However, a customer survey by these organisations would be necessary to establish accurately users requirements:

6.7.2 Weaknesses

The main problem encountered when processing data to be entered into The Science Fiction Pilot was that of fiction classification. In order to have knowledge of the plot, time consuming reading of the work was necessary, before classification could start. This problem was tackled for data entered into The Science Fiction Pilot, by the use of reVIews. However, if inclusion is dictated by reviews, then there is a danger of containing items because they are considered worthy by the publishing press to be reviewed, rather than giving the user the full choice of what is available. This is illustrated by shortage of satisfactory reviews for short fiction science fiction. Although

78

Chapter 6 - Evaluation and Conclusion

short stories are acknowledged as the life blood of the genre, there was only one specialist science fiction magazine which consistently reviewed this format.

6.7.3 Opportunities

The development of The Science Fiction Pilot has shown some of the achievements possible. Due to time restrictions, the scope of the system was limited to printed works, however, the science fiction genre encompasses other media. HyperCard allows sound, graphics and video to incorporated as data. When fully developed it may be possible for The Science Fiction Pilot to include sample section from, for example, graphic novels, television shows and films. However, before such media could be included, the problem of presentation must be addressed; how best to allow users to access images, and in the case of video initiate movement. Also copyright of these materials would have to be negotiated.

6.7.4 Threats

There are printed reference materials which cover the same area as The Science Fiction Pilot. What they lack in visual presentation over The Science Fiction Pilot, they make up for in breadth of information and view point. The Encyclopaedia of Science Fiction,

not only lists works within a theme, but includes critiques discussing the development of these themes, and the historical progression of authors work. While The Encyclopaedia of Science Fiction, tries to avoid personal bias, such opinions are more acceptable in

printed matter than electronic media where the tendency is to think of them as database stores of pure information, rather than a presentation of information as seen by an individual.

More work still needs to be done to turn The Science Fiction Pilot into an attractive product. If fully developed the system may provide a pleasant tool for access to science fiction.

79

Chapter 6 - Evaluation and Conclusion

References

1. Rubin. T. User Interface Design for Computer Systems. England: Ellis Horwood Limited,1988, p.137.

2. Ibid.,p.l37. 3. Ibid.,p.145.

4. Ibid.,p.159.

5. Reiterer,H & Oppermann,R. Evaluation of user interfaces: EV ADIS 11 - a comprehensive evaluation approach. Behaviour & Information Technology,

1993, 12(3), p.l38.

6. Rubin.T. User Interface Designfor Computer Systems. England: Ellis Horwood Limited,1988, p.150.

7. Reiterer,H & Oppermann,R. Evaluation of user interfaces: EV ADIS 11 - a comprehensive evaluation approach. Behaviour & Information Technology,

1993, 12(3), p.138. 8. Rubin.T. User Interface Designfor Computer Systems. England: Ellis Horwood

Limited, 1988, p.161.

9. ShackeI.B & Richardson.S.J. Human Factorsfor Informatics Usability. Gt Britain: Cambridge University Press, 1991 ,p.392.

10. Virzi,R.A. Defining the test phase usability: how many subjects is enough? Humanfactors, 1992,34(4), p.457.

11. Johnson,G.I, Clegg,C.W & Ravden,S.J. Towards a practical method of user

interface evaluation. Applied Ergonomics, 1989,20(4), p.257. 12. Yang,Y. Survey steered design: evaluating user recovery and command reuse by

questionnaire. Behaviour & Information Technology, 1989,8(6), p.442. 13. Ibid.,p.442.

80

Bibliography

Bibliography

Apple Computer, Inc. HyperCard Stack Design Guidelines. USA: Addison-Wesley Publishing Company, Inc, 1989.

Arents,H,C & Bogaerts,W,F,L. Concept-based retrieval ofhyperrnedia information: from term indexing to semantic hyperindexing. Information Processing

& Management, 1993,29(3),373-386.

Baird,P. HyperCard opens an electronic window on Glasgow. The Electronic

Library, 1988,6(5),344-353.

Baird,P, Davenport,L, & MacMorrow,N. What is this thing called hypertext? Aslib Information, 1990, 18(10),314-316.

Baker,S.L & Shepherd,G.W. Fiction classification schemes: the principles behind them and their success. RQ, 1987,27(2),245-251.

Boyle,C, Teh,S,H & Williams,C. An empirical evaluation of hypertext interfaces. Hypermedia, 1990,2(3),235-248.

Carmel,E, Crawford,S & Chen,H. Browsing in hypertext: a cognitive study. IEEE

Transactions on Systems, Man, and Cybernetics, 1992,22(5),865-883.

Carnahan,P.A. Beyond find: Booleari searching with HyperCard. Information·

Technology and Libraries, 1989,8(4),431-442.

Cleveland,D.D & Cleveland,A,D. Introduction to Indexing and Abstracting. USA: Libraries Unlimited Inc, 1990.

Oute,J & Nicholls,P, ed. The Encyclopedia of Science Fiction. London: Orbit, 1993

Fiderio,J. A grand vision. BYTE, 1988,13(10),237-244.

RI

Bibliography

Goodman,D. The Complete HyperCard 2. 0 Handbook. USA: Bantam Books, 1990

Goodstein,L.P & Pejtersen,A.M. The BOOKHOUSE: System functionality and evaluation. Denmark: Riso National Laboratory, 1989.

Gunn,J & Wolf,M.T. Science fiction: disturber of the literary peace. Library

Journal, 1988. 113(3), 129-133.

Hayes,S. Enhanced catalog access to fiction: a preliminary study. Library Resources

and Technical Services, 1992,36(4),441-459.

Hildreth,C. The concept and mechanics of browsing in an online library catalog. In:

Williams,M.E & Hogan, T.R, ed. National Online Meeting, Proceedings 1982. New

Jersey: Learned Information, 1982, pp.181-196.

Ingwersen,P & Wormell,I. Means to improved subject access and representation in

modern information retrieval. Libri, 1988, 38(2), 94-119.

Johnson,G.I, Clegg,C.W & Ravden,S.J. Towards a practical method of user

interface evaluation. Applied Ergonomics, 1989,20(4),255-260.

Kinnell,S.K. Comparing HyperCard and Guide. Database, 1988,11(3),49-54.

Langford,D & Brown,P. Creating hypertext documents: is it worth the effort? Aslib

Proceedings, 1993,45(4),91-95.

Leventhal,L.M. et al. Sleuthing in HyperHolmes: an evaluation of using hypertext

vs. a book to answer questions. Behaviour & Information Technology, 1993,12(3), 149-164.

MacGregor,J.N. A comparison of the effects of icons and descriptors in videotex

menu retrieval. International Journal of Man-Machine Studies, 1992,37, 767-777.

McKnight,C et al. Problems in hyperland? A human factors perspective. Hypermedia, 1989,1(2),167-178.

Michel,S.L. HyperCard: The Complete Reference. USA:Osborne McGraw, 1989.

Mohageg,M.F. The influence ofhypertext linking structures on the efficiency of

information retrieval. Human Factors, 1992,34(3),351-367.

R2

Bibliography

The Oxford English Dictionary. Oxford: Clarendon Press, 1989.

Pack, T. Shortcuts to finding short stories: searching fiction online. Database, 1992,

15(1), 59-61

Pejtersen,A.M. The "BOOKHOUSE": an icon based database system for fiction

retrieval in public libraries. Information and Innovation Proceedings of the 7th

Nordic Conference for Information and Documentation, 1989,165-178.

Pejtersen,A.M. The BOOKHOUSE: Modelling user needs and search strategies as

a basis for system design. Denmark: Riso National Laboratory, 1989.

Pejtersen,A.M & Austin,J. Fiction retrieval: experimental design and evaluation of a

search system based on users' value criteria (part 1). Journal of Documentation,

1983,39(4),230-246.

Reiterer,H & Oppermann,R. Evaluation of user interfaces: EV ADIS 11 - a

comprehensive evaluation approach. Behaviour & Information Technology, 1993,

12(3),137-148.

Rowley,J. A bstracting and Indexing. UK: Clive Bingley Limited, 1989.

Rubin,T. User Interface Designfor Computer Systems. England: Ellis Horwood,

1988.

Shackel,B & Richardson,S.J. Human Factors for Informatics Usability. Gt Britain:

University Press, 1991

Shell,B. Running HyperCard with HyperTalk. USA: Management Information

Source,Inc,1988.

Simpson,A. Navigation in hypertext: design issues. Online Information 1989,

Proceedings of the 13th International Online Information Meeting. 1989, 241-255.

Stevensons Book of Quotations. USA: Cassell & Company Ltd, 1964.

Sweetland,J.H. Beta tests and end-user surveys: are they valid? Database, 1988,

11(1),27-32.

Tomek,I & Maurer,H. Helping the user to select a link. Hypermedia, 1992, 4(2),

111-122.

83

Bibliography

Wilson,E. Link into hypertext. VINE, 1993, 91, 3-11.

Yang,Y. Survey steered design: evaluating user recovery and command reuse support

by questionnaire. Behaviour & Information Technology, 1989,8(6),437-459.

Zool,M.H. Good Reading Guide to Science Fiction and Fantasy. Gt Britain: Bloomsbury, 1989.

&4

Appendix 1 - Themes

Themes

Environment: Action: Mood:

Asteroids Disaster Absurbist

Atlantis Devolution Cyberpunk

Cities Evolution Dystopias

Colonisation of Worlds Fantastic Voyages Fantasy

End of the World Holocaust and After Gothic

Galactic Empires Invasion Horror

Game-Worlds Survivalist Fiction Humour

Generation Starships Sword and Sorcery Paranoia

Hollow Earth Time Travel Pastoral

Island Planetary Romance

Jupiter Characters : Satire

Living Worlds Aliens Science Fantasy

Lost Worlds Androids Space Opera

Mars Apes and Cavemen Steampunk

Mercury Children T echnothriller

Moon Clones Utopias

Outer Planets Computers

Parallel Worlds Cyborgs Ideas: Pocket Universe God and Demons Alternate Worlds

Ruritania Golem Anti-Intellectualism

Shared Worlds Heroes Conceptual Breakthrough S pace Habitats Hive-Minds Communications

Spaceships Messiahs Crime and Punishment

Stars Monsters Dimensions

Sun Mutants Discovery and Invention

Under the Sea Robots Edisonade Venus Superheroes Eschatology War Supernatural Creatures ESP

UFOs Far Future

Villains Fabulation

!l5

Feminism

Great and Small

Hitler Wins

Invisibility

Immortality

Life on Other Worlds

Linguistics

. Metaphysics

Near Future

Optimism and Pessimism

Origin of Man

Overpopulation

Money

Parasitism and Symbiosis

Politics

Prediction

PSI Powers

Pollution

Reincarnation

Religion

Sense of Wonder

Sex

Sleeper Awakes

Social Darwinism

Suspended Animation

Taboos

Time Paradoxes

Science:

Anthropology

Antigravity

Antimatter

Astronomy

Automation Biology

Black Holes

Cosmology

Cryonics

Cybernetics

Drugs

Ecology

Economics

Entropy

Force Field

Futurology

Genetic Engineering

Gravity

Hyperspace

Imaginary Science

Machines

Mathematics

Matter Transmission

Medicine

Nanotechnology

Neutron Stars

Nuclear Power

Perception

Physics

Power Sources

PseudoScience

Psychology

Rockets

Scientific Errors

Scientists

Sociology

Space Flight

Technology

Terraforming

Transportation

Virtual Reality

Weapons

86

Appendix 1 - Themes

Appendix 2 - User Evaluation Questionnaire

The Science Fiction Pilot A hypertext guide

User Evaluation Questionnaire

Thank you for agreeing to participate in this evaluation. The questionnaire is split into two sections; the first deduces your experience of computers and science fiction, whilst the second evaluates the system through your use and reactions. The evaluation should take you about fifteen minutes to complete. Please remember that it is the system being tested; not you.

Section 1 - Previous Computin& and Fiction Experience

1.1 How much previous experience of using computers do you have?

o less than a day o I month to less than I year

o I day to less than I week o I year to less than 2 years

o I week to less than I month o 2 years or more

1.2 How much previous experience of using a Macintosh computer do you have?

o less than a day o I month to less than I year

o I day to less than I week o I year to less than 2 years

o I week to less than I month o 2 years or more

1.3 Have you used HyperCard before?

o YES o NO

1.4 Do you read fiction?

o YES o NO ( if NO go on to Section 2)

R7

Appendix 2 - User Evaluation Questionnaire

1.5 What is your preferred type of fiction

D Period Literature

D Westerns

D Science Fiction

D Romances

D Murder mysteries

D Historical

1.6 Are your feelings about science fiction

Other. .•.••....••

Strong DislikeD Dislike D OK D Like D Strong Like D Don't Know D

Section 2 - Evaluation of the Section

. Please spend about five minutes familiarising yourself with The Science Fiction Pilot.

You will be shown how to enter the system, then continue at your own pace.

Please attempt the following questions. If you are unable to find an answer just state Don't Know:

Part 1 - Familiarisation

2.1 Select a story containing the theme of Aliens

What is the title: .................................................................................................................................... ............................................. ~ .................................................................................... . Who is the Publisher .................................................................................................. . (supply a printout if you wish)

2.2 List the story or stories written by Connie Willis

....................................................................................................................................

.................................................................................................................................... (supply a printout if you wish)

2.3 What is Nanotechnology

....................................................................................................................................

....................................................................................................................................

....................................................................................................................................

Appendix 2 - User Evaluation Questionnaire J

Part 2 - Comment

a) Learning the System

2.4 Did you find learning to use the system

Very Hard 0 Hard 0 OKO Easy 0 Very Easy 0 Don't Know 0

2.5 Was exploration of the system

Very Hard 0 Hard 0 OKO Easy 0 Very Easy 0 Don't Know 0

2.6 Were on-screen instructions

Very Hard 0 Hard 0 OKO Easy 0 Very Easy 0 Don't Know 0

2.7 Was answering the set questions

Very Hard 0 Hard 0 OKO Easy 0 Very Easy 0 Don't Know 0

2.8 Was locating help

Very Hard 0 Hard 0 OKO Easy 0 Very Easy 0 Don't Know 0

b) Presentation of System

2.9 Was deciding the function of a screen

Very Hard 0 Hard 0 OKO Easy 0 Very Easy 0 Don't Know 0

2.10 Did you find understanding the function of buttons

Very Hard 0 Hard 0 OKO Easy 0 Very Easy 0 Don't Know 0

2.11 Was the amount of information on each screen

Slightly Slightly

Too Much 0 too much 0 Enough 0 too Little 0 Too Little 0 Don't Know 0

R9

. .

, :

"

Appendix 2 - User Evaluation Questionnaire

2.12 Was following the sequence ofthe screens

Very Hard 0 Hard 0 OKO Easy 0 Very Easy 0 Don't Know 0

c) Users Comments

What do you think is (are) the best feature(s) of the system?

What do you think is (are) the worst feature(s) of the system?

Please write any additional comments you have in the space below:

.....................................................................................................................................

....................................................................................................................................

....................................................................................................................................

....................................................................................................................................

....................................................................................................................................

....................................................................................................................................

....................................................................................................................................

90

\ I