Information Technology For The Social Scientist

Ray Lee University of London., Ray Lee University of London.

• Publisher: Routledge
• Year: 2014
• ISBN:
  9781134218219
• Edition: 1
• Topic: Social Sciences
• Subtopic: Sociology
• Index: Social Sciences

Accessible and practical overview to help social reseachers make the most of information technology in relation to research design and selection, management and analysis of research data. The book pinpoints current and future trends in computer-assisted methods.; This book is intended for postgraduate and undergraduate social research methods courses and professional social researchers in sociology, social policy and administration, social psychology and geography. Particular appeal to courses in computer applications for social scientists and researchers.

CHAPTER 1

Information technology for the social scientist: an introduction

Raymond M. Lee

The work of social scientists is being transformed by the increasing use of new technologies. For thirty years, the computer in the social sciences was traditionally a tool for analyzing quantitative data. Now, however, massive increases in computing power at declining cost put onto the desks of users ways of dealing with topics and problems thought traditionally to lie beyond the scope of computers (Brent 1993, Blank 1991). Indeed, few areas of research, teaching or scholarship now remain untouched by developments in information technology. This book seeks to examine some of these new developments, to bring them to a wider audience, including potential new users, and to assess their possible impact, good or bad, on how social scientists go about their work.

Transformative technologies of social research refer to new means of representing and manipulating information, in particular the use of computer technology to provide (a) improved tools for the acquisition, storage and management of data, (b) means for representing conventional media forms — numbers, text, audio, video, pictures — into a single (digital) medium, and (c) computer networks which facilitate communication with other individuals and within teams, and which also permit access to information resources that can be widely dispersed. Such technologies would include, in no particular order: computer assisted data collection, computer analysis of text and of visual materials, large scale database systems, new developments in statistical analysis, modelling and simulation, artificial intelligence techniques, large scale computer networks, data archives, hypertext, multimedia and visualization techniques (see, e.g., Blank 1991, Fischer 1994).

Each chapter of this book contains a state-of-the-art review of a recent and interesting development in the application of information technology to the work of social scientists. The attempt has been made to present this information in a way which is both comprehensive and comprehensible. Although it is impossible to eradicate technical language completely, contributors have tried as far as possible to keep their contributions free of jargon and “technospeak”. Hopefully, therefore, the book will appeal to those who are relative computer novices as well as those who wish to widen their knowledge of current and possible future developments. It may also be useful on undergraduate and postgraduate courses on information technology taken by students in the social sciences. Increasingly such students are expected to have a basic familiarity with information technology and to be familiar with the basic computer tools in their discipline. To date, however, there has been a dearth of suitable texts which look broadly at computing developments in the social sciences.

The chapter by Mike Fisher of the University of Bristol is designed to provide an overview of desktop tools for the social scientist. In particular, he identifies a number of key areas where computers are useful: research management, maintaining contact with research participants, writing, and in the presentation of material both in print and before an audience. For each, Fisher describes appropriate software tools and some of the pitfalls in their use. Extending the craftwork metaphor with which he starts the chapter, Fisher closes, very appropriately, with a section on how computer tools may best be used wisely, safely and productively. Fisher’s chapter has much in it which will be of use to social scientists whatever their level of computer literacy. Those who are new to computing or perhaps a little intimidated by it will also find a reassuring introduction to the possibility of harnessing computer technology to enhance individual productivity.

A number of contributors to this book make reference to the usefulness of electronic mail and to the ability to access by electronic means information held at remote sites. Indeed, this book itself provides one example of the use of such technologies. A good deal of its initiation, organization and preparation was carried out electronically. A number of contributors responded to a request for papers put out on an electronic bulletin board and much of the editorial work of negotiating deadlines, reviewing the manuscripts and maintaining contact with authors was done by electronic mail. To take another example, students on a course run by Noel Heather (coauthor of the chapter on expert systems in this volume) analyze a text by Emile Durkheim as an exercise, the machine-readable version of which was originally retrieved by ftp (file transfer protocol) from a site in Canada.

Many social scientists appear intrigued and puzzled in equal measure by such developments, especially since the “information super-highway” has become a buzz term for media pundits and politicians alike. In this context the chapter by Ray Thomas is a useful one. Thomas looks at the kinds of information available to researchers through on-line databases and on CD-ROM, and examines the ways in which computers increasingly mediate communication between social scientists. After briefly recounting the development of computer based methods of information retrieval, Thomas reviews a major use of such technologies, the retrieval of bibliographical and text based material and the increasing access to library resources through worldwide computer networks. He then looks at the availability of statistical and administrative data in electronic form. (See also David Gilbert’s chapter on geographical information systems.) Thomas concludes his chapter with a discussion of how electronic mail, electronic bulletin boards and list servers are used by social scientists to communicate and exchange information with one another. Given the rapid pace of change such a chapter can only be indicative of the range of material available through forms of electronic dissemination and communication. For those unfamiliar with such developments, however, Thomas provides a basis for exploring the possibilities opened up by on-line retrieval and computer mediated communication.

In his chapter on desktop tools Mike Fisher gives an instance of how he used a database to track participants in a research study over a period of time. Databases have many uses in social research. They are best-adapted to situations where the researcher needs to store, order, update, retrieve and display information for a relatively large number of structured records. Databases are often used, therefore, to deal with bibliographic references, contact records and the like where, on the one hand, the data are, for example, not amenable to detailed statistical analysis but are more ordered than unstructured text on the other. There are various approaches to the design of databases depending on how data are structured and manipulated within them. Nick Rossiter describes the various types of database currently available and evaluates the strengths and weaknesses of each approach. Although the details of Rossiter’s account are sometimes intricate, he shows with examples from a social historical study and from political science research how databases can handle large amounts of disparate data to answer complex research questions.

The idea that a computer is predominantly a tool for the analysis of data dies hard. Increasingly, however, researchers are exploring ways in which the computer can be used to collect data, particularly in survey research. Jean Martin and Tony Manners have been at the forefront of these developments in Britain. In their chapter Martin & Manners explore the development of computer assisted personal interviewing (CAPI). Increasingly survey researchers no longer go out into the field armed with a clipboard. Instead they carry portable computers loaded with special software. This software allows them to enter directly responses to questions displayed on the computer’s screen. The program also automatically manages question-skips and edit-checks, making the process of data collection less prone to error.

Martin & Manners begin their article with a brief recapitulation of existing work on CATI, computer assisted telephone interviewing. They note that CATI was first embraced in the United States by commercial survey organizations, encouraged no doubt by high levels of telephone coverage and the geographically dispersed nature of the population in the US. By contrast, CAPI systems were developed first in Europe and by national census offices. Martin & Manners describe the main features of CAPI systems and discuss in detail their advantages and disadvantages. Moving to CAPI involves balancing the advantages of time and cost savings and a marked improvement in data quality against the capital cost of the equipment required, various constraints imposed by available software and the need for technical support. Martin & Manners also remind us that computers are not neutral in terms of their social impact. As they put it, “From the start of CAPI development it was apparent that there would be major organizational implications.” Based on their own extensive experience of implementing CAPI systems, Martin & Manners point up some of the organizational contexts and constraints which surround the introduction of new technologies into the data collection workplace. In a concluding section Martin & Manners look forward to improvements in hardware and software over the next few years and to the development of increasingly modular systems. They close their chapter, however, with a reminder that, despite the intervention of technology, the interview is still a social situation and, as such, remains dependent on the social skills of interviewers.

There are substantial academic and commercial markets for statistical software packages. Although there are clear market leaders, deciding on which particular package to use is not always an easy task. Moreover, the choice is becoming more difficult in some respects as packages spread across different hardware platforms, acquiring as they do subtle, and sometimes not so subtle, modifications in interface design. In their chapter devoted to statistical software, Ward and Dale provide a sure-footed guide to this treacherous terrain. For the purposes of their chapter they distinguish between packages mainly designed for the management of data and those primarily useful for data analysis. The latter they further distinguish into those designed for exploratory analysis and those for confirmatory analysis. In their final part they discuss a number of other aids to survey analysis in the form of spreadsheets and databases as well as a number of specialist programs. In each case they look at the major features of packages, their ease of use and the level of expertise required of the user.

Based on a long involvement with computer technology and his own extensive experience of anthropological fieldwork, Michael Fischer, of the Centre for Social Anthropology and Computing at the University of Kent, considers how computers can be used in ethnographic fieldwork. As he points out, although traditionally associated with research in anthropology, ethnographic, qualitative or field research is now common in many disciplines. Two particular aspects of this kind of research are relevant to computer use. The first has to do with the size, portability and reliability of hardware. If you are going to use a computer for this kind of research, you need to be able to take it with you and it has to be unobtrusive. Moreover, for some kinds of research there are also climatic and environmental considerations to be borne in mind. Fortunately, this is one area where commercial and academic imperatives for once align. If the executive flying in business class from Kuala Lumpur, say, to Copenhagen is not to be denied the benefits of portable computing, then they potentially also become available to the academic squashed into economy class on the same flight who is en route to or from a field site. The second aspect of using computers in ethnographic research is the use of information technology to handle and maintain ethnographic data. The stuff of ethnographic research is messy. Field research typically produces an assemblage of data which is multi-stranded, multi-sourced and takes a multiplicity of forms (Lee 1993). In other words, a complex set of themes and topics is woven through the research material, which itself may take many different forms depending on the nature of the study: fieldnotes, genealogies, maps, diagrams, audio-visual materials and so on. In a way, Fischer extends here some of the themes already developed by his near namesake Mike Fisher in the chapter on desktop tools, by discussing tools appropriate to handling the richness of ethnographic data. Those interested in the use of visual materials, particularly videotape, in ethnographic research will find his discussion particularly useful as it provides an introduction to relevant software and hardware for video-based research.

One area which has recently attracted a good deal of attention is the use of computers to analyze qualitative data. In this context the term qualitative data refers to unstructured data such as fieldnotes or transcriptions of depth interviews or focus groups. Developments in this field are discussed in the chapter by Wilma Mangabeira, a sociologist at the Federal University of Rio de Janeiro and someone who began early on to explore the implications of using computers to analyze qualitative data. Although clearly written, Mangabeira’s chapter may be less accessible to readers new to computing than some of the other chapters in this collection. In a sense this reflects the field itself, which has developed rather rapidly within a relatively short period of time. It may be appropriate therefore to review, very briefly, developments to date. (Further information can be found in Pfaffenberger 1988, Tesch 1990, Fielding & Lee 1991, Huber 1992.)

The use of computer software to analyze textual data is not entirely new. Humanities researchers, especially biblical scholars and the like who have to deal with very large volumes of textual material, have long been interested in the use of computers for analyzing non-numeric data (Lee & Fielding 1991). Traditionally, in disciplines like sociology, qualitative researchers used manual methods to analyze data from field research, perhaps marking up fieldnotes or interview transcripts with coloured pens, or making multiple xerox copies of relevant segments of field material. Only in the 1980s did computer packages specifically designed to replace these cumbersome and tedious procedures begin to emerge. These early programs which were mainframe-based were replaced quite soon by microcomputer programs which allowed researchers to delineate and collate patterns or themes appearing in, say, an interview transcript through a process of attaching codes to segments of text. As Mangabeira points out, however, this mimicking of traditional “cut and paste” methods gave way to programs which aim to facilitate the development of sophisticated interrelated systems for categorizing data and, in some cases, for hypothesis testing. These “third generation” programs, as they have been called, form the subject of Mangabeira’s chapter, which is made more valuable by being based on a close involvement, in the course of a Fullbright fellowship, with the developers of two of the more intriguing packages in the genre. Mangabeira explains the thinking behind third generation software and seeks to relate the underlying models to new substantive and methodological concerns.

There is a sense in which one can regard Mangabeira’s contribution as a “third generation” paper. Just as the software has evolved, so has commentary upon it. Much of the early literature reported, with an air of faint amazement, that, yes, computers could be useful in qualitative research. Later, users and developers of the software began to speculate on the possible impact CAQDAS programs would have on the craft of qualitative research. Only now are we beginning to see papers like Mangabeira’s which place computer assisted qualitative data analysis in context by going beyond speculation to material based on empirical analysis and/or explicit evaluation.

Hansen’s chapter, along with a number of others in this collection, reflects a trend in which the computer is no longer seen as a device, quite literally, for computing, for handling numeric data, but one which handles many different forms of information, numeric, visual and textual. In effect what the computer allows is the translation of many media into one medium. Specifically, Hansen describes ways of harnessing information technology to the analysis of newspaper text. While newspapers are an extremely important source of data on social life, their analysis has traditionally been hampered by a number of factors. The dross rate for mass media data is high. In the case of newspapers, the physical characteristics of the data source in terms of size, bulk and, over time, friability additionally complicate the analyst’s task. Hansen describes recent developments which make newspaper data available, at a price, in electronic form. He then shows how the computer can be used to address, manipulate and analyze the information so provided. He also points to the way in which the computer affects the analysis process, in this case in the ease with which it is now possible to counterpose a quantitative approach of the kind traditionally associated with content analysis with a more qualitative analysis based on themes, images, narrative and so on. How far it might be possible in the rather distant future to extend such methods to other forms of mass media remains to be seen.

As David Gilbert points out in his chapter on geographical computing, maps were imp...