Resource Library: Multimedia

home *** CD-ROM | disk | FTP | other *** search

/ Resource Library: Multimedia / Resource Library: Multimedia.iso / hypertxt / msdos / hypdiss / chapt3 < prev next >

Wrap

Text File | 1992-04-11 | 22.0 KB | 394 lines

CHAPTER III METHODOLOGY AND RESEARCH DESIGN Research Method This study addresses the information retrieval potentials of the hypertext software technology. It investigates the effective transferability of traditional information access methods to this new electronic system. The research questions in this study focus upon present and potential information access facilities of a particular hypertext information system. The selected subject is a fully operational hypertext authoring system used for current production of commercial information products. There was therefore no opportunity for investigator control over system implementation features. This investigation of the information access facility planning within a hypertext system is also an initial, exploratory investigation. The investigator used insights gained in the study to make generalizations applicable to future hypertext information access design. The investigator selected the case study research method as the appropriate one for use in this study. This chapter provides background for that decision and provides an overview and summary of the research design for the study. Case Study Rationale Yin (1984) <refs -yin> defines a case study as an empirical inquiry that: * investigates a contemporary phenomenon within its real-life context; when * the boundaries between phenomenon and context are not clearly evident; and in which * multiple sources of evidence are used (23). These characteristics clearly match the conditions of this study. In addition, the case study method is generally regarded as particularly appropriate for exploratory studies (Busha and Harter 1980 <refs -busha>, Whyte 1984 <refs -whyte>, Yin 1984 <refs -yin>). Busha and Harter (1980) <refs -busha> consider the overall purpose of a case study to be the gathering of comprehensive information about the research object. They observe that the case study has potential for concentrated focus on a single phenomenon, and the use of a wide variety of data- gathering methods. They state that the usual data- gathering methods center upon direct observation. The authors also note that investigator analysis can identify valid relationships among the many variables that may be present. They point out that the case study method may also result in insights about the research question or hypothesis, and may stimulate further studies using other, more controlled, research methods. In his landmark work on case study methodology, Yin (1984) <refs -yin> compares five major social science research strategies, including experiment, survey, archival analysis, history, and case study. He considers the different methods in relation to: the form of the research question; the requirement for control over events; and need for investigation focus upon contemporary events. He characterizes the case study as particularly appropriate or advantageous for studies involving "how and why questions," being asked about contemporary events within some real-life context, over which the investigator has little or no control. Yin also identifies and answers several traditional prejudices against use of the case study: 1. Accusation - a general lack of rigor. Response - He proposes this be addressed by careful research planning and control. 2. Accusation - provide little basis for scientific generalization. Response - He points out that case studies, like experiments, are generalizable only to theoretical propositions, not to populations or universes. (This restriction follows the reasoning that neither case study nor experimental data represents a valid "sample.") 3. Accusation - case studies take too long and result in massive, unreadable documents. Response - He argues that this may be addressed by careful planning and quality in reporting and presentation of results. He points out that the "overlong time" portrayal incorrectly confuses the case study strategy with specific methods of data collection. He also notes that exhaustive methods may at times be necessary and appropriate. Hillway (1956) <refs -hillway> notes the similarity of the case study method to the survey method. He observes that both methods ascertain facts and analyze them in order to draw inferences and establish general principles. He comments upon the wide variety of data sources that may be used in the case study, including written records, as well as observation and interviews. Hillway also identifies potential defects of the case study approach. One is the difficulty of selecting representative cases for study. Another is the strong element of subjectivity that seems unavoidable. However, he concludes, "Nevertheless, the case study, especially when used in conjunction with a quantitative survey, often draws attention to information that cannot be obtained successfully in any other way, and thus can be justified scientifically" (Hillway 1956, 244-245). The exploratory nature of the case study can additionally serve as the basis for effective investigation or evaluation of working systems. Byrd (1975) <refs -byrd> defines systems as a possibly complex series or network of interconnected steps or stages, working together to achieve some desired end. He observes that it is sometimes necessary to evaluate alternate approaches or solutions for the different system units. It is often difficult or impossible to use true experimental methods to isolate and measure system variables in an operational system. The exploratory nature of the case study approach can result in identification of critical elements. Results of a case study can thus be valuable in definition of a model for advanced systems evaluation or operations research investigative approaches. In his introduction to Yin's book, Campbell makes a case for acceptance of the case study method as a valid approach for social science research. He notes that in the formalized case study approach, rival hypotheses must be specified and specifically controlled for. He likens the case study approach to the "experimental isolation" or "laboratory control" models of physical science research. This latter research model is unquestionably recognized as a valid alternative to the "randomized assignment to treatments" model of quantitative research (Yin 1984, 8) <refs -yin>. Tuma additionally points out that the case study is a highly cost-effective approach. He states that large-scale studies, attempting to measure a variety of factors, and simultaneously meeting the rigorous assumptions of other research approaches, are often simply not feasible, if applied on a large scale. He concludes, "The main point, then, is to choose a research strategy that is scientifically sound, yet feasible and cost-effective" (Tuma 1982, 80) <refs -tuma>. Agreeing with the other writers, he states that the case study is well-suited to generating new and working hypotheses. Upon review, the case study approach is regarded as a valid, useful, manageable and economical research approach for application to suitable research efforts. Based upon the recommendations given by various authorities, the case study method is well-suited to the requirements of this particular study. Selection of Subject The investigator surveyed a number of commercially available hypertext authoring systems packages, prior to final subject selection. He was most interested in hypertext systems compatible with the MS-DOS operating system, the most widely distributed personal computer platform, now approaching mass market penetration. He was also personally restricted to use of an MS-DOS system for the study, and therefore limited subject candidates to products written for that operating system. He personally used and evaluated the following hypertext products as potential study subjects: Guide (TM), by OWL, Inc.; Hyperties (TM), by Cognetics, Inc.; Folio Views, by Folio Corporation; HyperWord (TM), by Zaron Software, Inc.; HyperWriter (TM), by NTERGAID, Inc.; Knowledge Pro (TM), by Knowledge Garden, Inc.; MaxThink (TM), by MaxThink, Inc.; and xText (TM), by Flambeaux Software, Inc. The investigator selected the MaxThink authoring system package as the case study subject for the following reasons: 1. This system had the most sophisticated hypertext network building and information organizational modules. The system is a combination of various editor, outlining and network building programs. There is also a collection of separate utility programs, most serving information control purposes. Many of the other systems offered more sophisticated hypertext implementations and more attractive user interfaces. However, in the investigator's judgment, none achieved MaxThink's levels of hypertext linking- and network-control sophistication. 2. The system is currently used for production of a sophisticated hypertext product, the DaTa CD-ROM MaxThink, Inc. and Pondlife Systems, Inc. produce this hypertext system on a contract basis for Deloitte & Touche, Inc. <app-e description of production>. To the investigator's knowledge, the DaTa system is the largest and most complex hypertext product currently published, with the exception of the massive Oxford English Dictionary hypertext product. Moreover, DaTa is a serial hypertext publication, updated four to six times annually, and thus requires a sophisticated maintenance and updating schedule. The December, 1990, edition was eighty megabytes in size. It contained approximately 200 interconnected hierarchical hypertext networks, with 20-25,000 ASCII files, 60,000 display screens, and an estimated 150,000 individual hypertext links. 3. The MaxThink system was additionally attractive as a case study subject because of the publishing activity of the developer. MaxThink, Inc. combined software developer status with active management of a large hypertext publishing operation. In addition to the DaTa product, these principals have undertaken publication of numerous other hypertext information products. 4. The information organizational features of the MaxThink product reflect the strong personal views of the software developer about the importance of organization of a knowledge base (Fersko-Weiss 1991 <refs -fersko>, Perez 1991 <refs -perez>, Urr 1991 <refs 23 8>). This focus is evident throughout his documentation, one oft-repeated phrase carrying even to the logo screens of his programs: "The value of information lies in its organization" (TransText 1990, [logo screen]) <refs -transtext>. The Information Access Model (IAM) The investigator defined a conceptual Information Access Model (IAM), as a representation of the features in traditional information access systems. This model was expressed in a brief outline format (See Appendix A). <app-a> The IAM was designed primarily to present or identify the different general approaches or access points to information. It does not deal with details of system design or implementations of access and control systems. To create the IAM, the investigator studied the library and information science sources cited in the first part of the Review of the Literature chapter. During this detailed review, he paid particular attention to the sources dealing with planning for design of indexing and subject access systems. He extracted both general ideas and details of recommended information system design, and of access point planning and control (Borko and Bernier 1978 <refs -borko>, Cleveland and Cleveland 1990 <refs -cleveland>, Foskett 1982 <refs -foskett>, Jonker 1964 <refs -jonker>, Metcalfe 1957 <refs -metcalfe>, Milstead 1984 <refs -milstead>, Rowley 1987 <refs -rowley>, Taylor 1986 <refs -taylor>). After thorough comparison of the categories and descriptions of information access methodology presented by the various sources, he synthesized the final IAM conceptual model outline <app-a>. This offers a very generalized statement covering the important characteristics to be considered in the design of an effective information storage and retrieval system. Dr. Adeline Wilkes, the dissertation committee member specializing in the cataloging and classification area, reviewed and approved the model prior to further work. Part I of the IAM, "Information archive transport system," <app-a 1 8> briefly lists basic components of archival or storage systems. It notes the need for an archival or collection assembly and maintenance mechanism, the necessity for a logical direct (basic) file arrangement, for interface design considerations, and the necessity for robust production and organization/control tools. The IAM's Part II, "Information access systems," <app-a 1 21> covers the major areas of information access. The three major subsections are "B.1. Types of access points," "B.2. Type of access approaches," and "B.3. Control mechanisms." Section B.1. <app-a 1 23> covers the main access points possible in information access systems. These include the major points generally taken as distinctive approaches to selecting document pointers within an information system. Examples are: Author, Title, Keywords, Subject, Date, Geographic, etc. Section B.2. <app-a 2 24>covers the types of classification and indexing approaches generally available for designing information access systems. These are "types of classification systems," "types of indexes," or other access mechanisms. Examples are: hierarchical or enumerative classifications, alphabetical indexes, permuted indexes, post-coordinate or Boolean retrieval approaches, and full-text searching. Section B lists these access types in detail and is a prelude to the more detailed coverage in the interview schedule. The complete IAM model was prepared in an outline format, and is attached to this study as Appendix A. Notes on various lines of the model refer to discussions of the particular item in Cleveland and Cleveland (1990) <refs -cleveland>, a recent and complete overview of the index access topic. The Interview Schedule The writer used the IAM as the basis for an expanded interview schedule or questionnaire <app-b>. This instrument was designed for administration to the principals of the MaxThink/Pondlife Systems operation. These individuals are the producers of the DaTa hypertext product. Two members of the Faculty Committee reviewed and approved the schedule prior to the field study visit to Berkeley, California. The schedule closely paralleled the IAM, but was worded in a predominantly layperson or natural language style of usage. It included a categorical answer scale and space for brief notes. The investigator discussed the schedule with both interviewees on the day prior to the interview and left a copy with them for examination. Immediately prior to the interview session, he explained and discussed with them any areas which they had not found to be clear in meaning. The investigator felt this sequence of presentation to be necessary since he was discussing details of technical library or information science topics with a software developer and a domain expert who were laypersons in this area. He therefore used this pre-explanation and discussion approach, to clarify the questions to the interviewees. A copy of the complete Interview Schedule is attached to this study as Appendix B. Case Study Protocol The investigator prepared a detailed Case Study Protocol <app-c>, as a formal expression of the details of the proposed study. The document was prepared in an outline format based upon Yin's protocol design recommendations (Yin 1984) <refs -yin>. The protocol presented details on case study purpose, procedures, and final analytical strategy. This was presented in a summarized format, so that the details of the study were stated in four single-spaced pages. With this requirement for clear, formal expression of procedures, the protocol document therefore mandated detailed and careful planning. It was thus an additional project-planning aid. Two faculty members of the Dissertation Advisory Committee reviewed and approved the protocol document prior to the field study trip. The complete protocol document is attached to this study as Appendix C <app-c>. Procedures The investigator accomplished the study using the following general progression of steps: 1. Initial study planning, selection of methodology, and selection of the case study subject. 2. Formal study proposal and committee approval. 3. Preparation of and committee approval of the Information Access Model (IAM) and Interview Schedule. 4. Working Protocol preparation and committee approval. 5. Scheduling of case visit to Berkeley, California, during week of March 11-15, 1991. 6. Field study visit, accomplishment of interviews and background observations. Interviews were conducted using detailed note-taking, plus tape-recording of the interview session. 7. Upon return, transfer of interview tape and notes detail over to interview summary reports. These two interview reports, not verbatim transcripts, are attached to this study as Appendixes D and E. 8. Preparation of final study report. Methodology Section Conclusion The next chapter reports the findings of the study. It also contains preliminary discussion on the implications of these findings.