25.3 CHAPTER OVERVIEW

Each of the three areas of research in information access are now examined in depth: users, access, and information.

• Research and methodological issues are discussed.
• Analyses of significant studies are provided.
• Directions for future research are considered.

In addition, three longitudinal research projects that integrate users, access, and information are reviewed at length. The final section of the chapter summarizes LIS research and briefly discusses implications of access issues in library and information centers for research in educational communications and technology.

This chapter has three limitations for readers to consider: subject content, time frame, and scope of literature reviewed.

The subject content is the literature and research of library and information science related to information access. By focusing on one subject content domain, it is possible to identify patterns in the research. These can be matched with research in compatible areas such as educational communication and technology. The time frame is approximately a 15-year span beginning in 1980. Some earlier work is examined, and reviews of research that cover earlier periods are included. The scope is reviews of research as the first focus, and specific studies secondarily. Reviews of research provide foundations for exploring patterns. Specific studies provide exemplars and point to new research areas.

25.3.1 RESEARCH ON USERS

User research has undergone significant changes in the past 25 years. It has moved from a field that primarily studied identifiable groups and focused on system-centered questions to a field that looks at individuals and focuses on user-centered behavior and questions. The current foundations encompass communications theory, social psychology, and cognitive psychology. Methodology is moving from sociological surveys toward qualitative techniques such as case studies, ethnographic research, and grounded theory. Quantitative methods from communications theory and cognitive psychology are also being applied.

Research about users has formed a significant part of the research in library and information science (Krikelas, 1983). Emphasis on user groups predominated the early research efforts (Hewins, 1990) and still accounts for a large portion of research. Because of the specific characteristics and problems of groups in different content areas, the research literature on user groups is scattered among many content areas from medical science to agronomy research. A research trend is to explore characteristics of individual users rather than focus on specific content groups. Since the late 1970s, suggestions for focusing on individuals in practical contexts has become prevalent (Wersig & Windel, 1985; Wilson, 1981, 1984). Research on how people, whether collectively or individually, fit into information access can be divided into three areas:

• User studies, which include characteristics of users, information on how different user groups function, and identification of user needs
• Information seeking, which includes cognitive processes used to access information, user search strategies, and differences between end users and search intermediaries
• Information skills, which include creating models of information skills functions and developing research on the information search process

25.3.1.1. User Studies. The most important body of literature on user studies is the review series appearing in the Annual Review of Information Science and Technology

(ARIST) since 1966. These reviews provide an overview of a research area over an extended period of time.

They show the development of a field study over more than 25 years. Earlier reviews detail how user studies began. Later reviews build on earlier reviews by evaluating their ideas and analysis schemes. In the beginning, these reviews appeared every year, then every 2 years, and most recently on an unscheduled basis with 4 to 8 years between reviews. Krikelas (1983) believes this may be due to erratic and all-encompassing definitions of what user studies constitute. Bates (197 1), for example, includes catalog use, use of reference services, use of library materials, and general library attendance, in addition to information-gathering habits of scientists. To examine cur-rent trends in user studies and definitions, this section elaborates on the two latest review articles (Dervin & Nilan, 1986; Hewins, 1990). An analysis of issues of concern related to user studies completes this section.

Dervin and Nilan (1986) review 1978 to approximately 1986, almost 8 years. They include all user studies since the previous review (Crawford, 197.8). They approximate 300 overall citations related to users but narrow their focus to conceptualizations; that drive the research. Past reviews have shown that the field may have conceptual limitations. Menzel (1968) explains that theory from the social sciences should drive user research. Crane (1971), Crawford (1978), Lin & Garvey (1972), and Paisley (1968) discuss the development of theory or conceptual frameworks within the context of user studies. Lipetz (1970) focuses on a methodological issue: improving predictive value. Allen (1969), Herner and Herner (1967), and Martyn (1974) ask for methodological choice and attitudinal perspectives.

Dervin and Nilan (1986) say the current literature, 1978-86, has a number of critical essays calling for reassessments of the type of research and a particular concern for conceptualization. They see two distinct areas: need for theory and need for better definitions and premises. They believe that definition is preliminary to theory and should be developed first. They state that innovative work in the time frame of the review does concentrate on definition.

Four research changes for user studies are identified. The first is to increase the match between information systems and users to serve the client better. This would result in greater accountability and fewer undeserved clients. Their second research change would focus on user needs and uses. This would result in the user as a central issue for research, not the system or the technology. Their third research change would redesign the system to look more closely at person-machine links. Their fourth research change would focus on the technology of user access.

Dervin and Nilan (1986) also discuss their beliefs about changes in user studies research. Much of the work is the same as before (looking at users in terms of systems), but some research seems to represent a different direction (looking specifically at users). A common systems study looks at the degree that a user has actually used a system or systems, identifies barriers to the use of the system, or identifies satisfaction with attributes of the system. This research has also looked at demographics, sociological components, and life-style tasks. Problems with these types of studies have included ill-defined terminology and terms used interchangeably. They suggest that this systemic emphasis results in six approaches to needs assessment:

1. How much do users use the system?
2. How aware are users of the system?
3. What do users like and dislike?
4. What are the users priorities for information?
5. What are users interests, activities, and group involvement?

Researchers trying to change user studies suggest an alternative approach. Dervin and Nilan (1986, pp. 13-16) develop seven categories to contrast differences between the traditional and alternative approaches they believe are beginning to take hold in user research:

1. Objective vs. subjective information
2. Mechanistic, passive vs. constructivist, active users
3. Trans-situationality vs. situationality
4. Atomistic vs. wholistic views of experience
5. External behavior vs. internal cognitions
6. Chaotic vs. systematic individuality
7. Quantitative vs. qualitative research

Hewins's (1990) review updates Dervin and Nilan (1986) and reasserts the belief that research has shifted from sociological to psychological. Hewins's review states that Dervin and Nilan proposed a change from traditional empirical studies to user studies that provides rational frameworks and systematic bases for methods and definitions. Hewins restates Dervin and Nilan's belief that three new approaches are emerging: (1) user values, (2) sense making, and (3) anomalous states of knowledge (ASK). These new approaches focus on identifying user characteristics rather than on measuring system performance. In addition, Hewins says studies from the past have a common framework in that they are sociological or system based. The study of groups and the design of appropriate systems formed the foundation for the earlier work in user studies.

Hewins (1990) believes that few innovative methods have been developed, because questionnaires and surveys are still the norm. Dervin and Dewdney's (1986) time-line-critical-events approach is described as a new and interesting methodology. The National Library of Medicine's (1988) solution development record, a variation on critical-incident methods, is also discussed. In summary, Hewin (1990) states that new cognitively based approaches include categorization techniques, long- and short-term memory, learning

styles, motivation, personality types, and semantic factors. The need for more interdisciplinary is discussed.

After reading both Hewins (1990) and Dervin and Nilan (1986), it is apparent that a change is taking place in the study of users, systems, and their interaction. This seems to reflect general changes taking place in the way educational researchers and scholars view research. There has been more emphasis on qualitative methodology, a swing from behavioral to cognitive psychology, and a greater tolerance for deriving theory from a broad spectrum of research areas.

User surveys, the most common method of data collection in user studies, are reviewed by Verhoeven (1990) and reflect some of the same issues discussed by Dervin and Nilan. Both methodological and conceptual problems with the user survey are discussed. Issues such as incomparable results, lack of generalizability, and confounding methods are examined. In a general review of research findings from surveys over time, Verhoeven reports the following:

• In public libraries, adult users are well educated; 20% of the population accounts for most of the adult use.
• ounger library users used public libraries almost exclusively for course work.
• On-line catalogs are popular.
• Microfilm has been dissatisfying to people who wear glasses.
• People who do not use libraries have no need or no time.
• Friends and colleagues are preferable to libraries for seeking information.
• Considerable variability exists in information-seeking behavior. One persons' easy search can be most difficult for another (p. 391).

New approaches to the user survey can be divided into two general areas: (1) improving surveying techniques, (2) finding new directions (Verhoeven, 1990). The first type of researcher turns to surveys as a research area. The attempt is to find out why traditional surveys do not work. The next step would be to use this knowledge to design better surveys. The second approach looks at a different aspect. Traditional surveys were predominantly questionnaires that examined user satisfaction and user interaction with specific researcher-generated services. A new research focus is to identify the situations that prompted the user need. The information problem is examined from the user rather than the system perspective. Verhoeven defines the first approach as positivistic and the second approach as phenomenological. The ongoing conflict between the two methods is summarized in this way:

Some commentators believe that, although the phenomenologists have destroyed positivist underpinnings, they have proffered little to put in their place.... On the other hand, there are those who find it not unreasonable to continue asking library users about their experience with libraries ... (p. 394).

Applegate (1993) is one of the researchers suggesting a new method for determining user needs. Three models of user satisfaction are evaluated in the 1993 article: (1) material satisfaction model, (2) emotional satisfaction model-simple path, and (3) emotional satisfaction model-multiple path. In material satisfaction models, measures of product performance such as recall and precision are the determiners of user satisfaction. This can often lead to false positive conclusions where the product performs well but users are not necessarily satisfied. Simple-path emotional models are built on measuring users levels of satisfaction through direct and indirect questions to the user such as, "Are you satisfied with the results of your search?" Again, a false positive can result. Users express satisfaction but have received poor results from the search. The multiple-path model uses both user satisfaction responses and additional factors such as setting and expectations.

These three models have implications for research in user satisfaction. Applegate (1993) suggests that researchers should clearly specify what definition of satisfaction they are measuring. Secondly, more standardized instruments for data collection are recommended. Finally, multipath models such as that described above are suggested as explaining more complexity in user satisfaction.

There also appear to be other areas where user studies research could direct greater effort. One of the issues Hewins (1990) addresses is that studies which are site, system, or service specific and that do not contribute to method, theory, or model building are not covered in the review. On the surface this seems reasonable. On the other hand, this may not fully represent local knowledge and the development of grounded theory. Informal and formal case studies can provide valuable information about site-specific user characteristics and needs. From the description, it is difficult to know what was not included in this review. However, further use of techniques that gather and develop theory from local knowledge, such as action research (see 9.7.5), might prove useful for research in user studies.

User-need studies look at the user and what she or he brings to the situation. The studies also need to examine social constraints, cultural influences, and the effects of demographic variables such as age, gender, and class. Chen and Herndon (1982) is one of the landmark research studies in user needs. One issue from the Chen and Herndon study that is still relevant today is underserved populations. They found two especially relevant groups to consider: (1) users who are not using library and information center services, and (2) users who are not able to use library and information center services. For example, in their survey many elderly users were not able to answer the questions asked, so their information needs were not known. A more recent study by Metoyer-Duran (1993b) demonstrates that for minority populations, the source of information can serve as gatekeeper to access for both the individual and the community.

Reneker (1993) sums up the basic problems with user studies in this way:

[This research] leaves one with little more than a sense of what kinds of materials are used by specific groups of scholars, researchers, or students. Most of these studies attempt to relate specific variables to particular pieces of information-seeking tasks or behaviors (p. 488).

The previous research of user studies was focused on systems and founded in sociological traditions both theoretical and methodological. The reason ARIST reviews in this area are becoming fewer and farther between may be that a new area of research is emerging which combines aspects of users with other research issues. Research focusing on the individual user rather than on groups is discussed in the "information seeking" section that follows.

Information seeking is an active process: "... it begins when someone perceives that the current state of possessed knowledge is less than that needed.... [It] ends when that perception no longer exists" (Krikelas, 1983, p. 7). Researchers who study information seeking are looking at more than the specific user variables that relate to the system as described in the user studies above. Information-seeking researchers want to know the individual strategies and techniques users employ to look for and find answers. They are asking complex questions and using theories and methods that can provide breadth and depth to the answers. Two theoretical positions are foundational in information-seeking research: communication theory (Dervin, 1983a, 1983b, 1989b) and cognitive psychology (Belkin, 1990; Ingwersen, 1992; Reneker, 1993). In addition, two types of users are commonly considered by the information-seeking researcher: the end user and the search intermediary. End users are individuals who want to answer their own questions and use the information resources independently. Intermediaries are trained librarians who interpret users' questions and problems. Information seeking and search strategies can be different depending on who is seeking and searching.

The sense-making theory of Brenda Dervin and colleagues is founded in communication research and builds on constructivist principles (Savolainen, 1993). Its strength is that it is a programmatic research effort conducted over time, systematically examining problems and issues in information seeking and use (Dervin, 1983b, 1989a, 1989b). It focuses on a user-centered rather than a system-centered approach. In sense-making theory, sense-making behavior is considered to be communicating behavior. The theory has been tested over a series of 40 empirical case studies of information seeking and use. (A detailed discussion of Dervin's research is found in a later section of this chapter.)

A new trend in information-seeking research uses qualitative approaches, specifically ethnography (see 40.2; Reneker, 1993) and grounded theory (see 40.2. 1; Ellis, 1993; Ellis, Cox & Hall, 1993). This research attempts to examine complex issues and provide information on the nature and effects of information seeking. These researchers want to go beyond superficial or surface layers. Reneker (1993), in a study of information seeking among academics, used a set of 2,050 information-seeking incidents as a foundation for analysis. These activities were drawn from personal, social, scholarly, academic, work-related, political, and other arenas. Conclusions from the study showed that information seeking is a part of daily tasks and relationships. It can result both from an articulation of a need and from available information.

Ellis (1993) and Ellis et al. (1993) used grounded theory to examine, respectively, the information-seeking patterns of academic researchers and to compare researchers in the physical and social sciences. The results of their research are the outlining of different categories to represent information-seeking patterns for four content specialists: social scientists, physicists, chemists, and English literature researchers. After analysis, the following six categories were developed for social scientists: starting, changing, browsing, differentiating, monitoring, and extracting. For physicists, five categories were identified: initial familiarization, chasing, source prioritization, maintaining awareness, and locating. The study of chemists identified eight categories: starting, changing, browsing, differentiating, monitoring, and extracting (the same as the social scientists), plus verifying and ending. The English literature researchers resulted in six categories also: starting, surveying, chaining, selection and sifting, assembly, and dissemination. Ellis states that even though different terminology and slightly differing numbers of categories were identified, they can be considered to represent fundamentally the same activities.

Another information-seeking research area focuses on information seeking within the context of electronic technologies. Marchionini, Dwiggins, Katz, and Lin (1993) exemplify studies conducted in this area. They review four previous studies and discuss implications for various types of users. They describe information seeking as a problem-solving activity and look at both content area experts and search intermediaries. They focus on five factors: an information seeker, an information problem/setting, a knowledge domain, a search system, and a set of outcomes. They conclude that content experts focused on search strategies that emphasized content, while search intermediaries were problem driven.

New studies (Reneker, 1993; Ellis, 1993) in information seeking are quoting the Dervin and Nilan (1986) review of user studies literature to verify that a paradigm shift has occurred. To date, no one has actually validated the assertion with an analysis of the literature or articulated the common components of the new paradigm. It seems that one statement has been influential in directing the beliefs of other researchers. These beliefs need to be confirmed through further analysis and study. With the exception of the Dervin sense-making studies, there are few longitudinal studies creating a body of sustained knowledge. Each researcher creates unique categories of the seeking process. Attempts to integrate. the various points of view or to discuss the commonalties and differences between the results would be a valuable addition to the information-seeking literature. Research studies based on these common categories or focussing on the impact of theoretical differences could help extend the knowledge base of information-seeking research.

Information skills research is primarily undertaken within the context of school library media centers. It seeks to "... define the nature and scope of information processes" (Eisenberg & Spitzer, 1991, p. 265). Scholarly activity in this arena can be divided between model development and information skills process research. Cutlip (1988), Eisenberg and Berkowitz (1990), Irving (1985), Loertscher (1988), and Stripling and Pitts (1988) fall into the first category. Kuhlthau (1983, 1988, 1993b) is an example of a researcher with a systematic program in information skills: the information search process.

Kuhlthau has explored stages of the information search process through use of qualitative research methods to build grounded theory. Both affective and cognitive aspects of the process were developed based on longitudinal observations and interviews with high school students, public library users, and others. (A detailed discussion of Kuhlthau's work can be found in a later section of this chapter.)

Bartolo and Smith (1993) used Kuhlthau's information search process (ISP) theoretical model in a study of manual and on-line search methods in university journalism classes. The focus was on interdisplinarity, and the method was an eclectic mix of techniques. They found that: "the superior performance of the on-line group over the manual group supports the authors' hypothesis that on-line search methods are more effective than manual search methods in helping researchers handle the challenges of interdisciplinary work ..." (p. 35 1).

Work in information skills and the development of research studies and a research agenda is emerging. Treasure Mountain Research Retreat 5 (Barron, Grover & Loertscher, 1994) shows future directions for information skills research including: further model development, longitudinal studies of integrating information skills with content areas, and alternative assessment measures.

Many of the early studies within user research are descriptive. They delineate characteristics of users and their needs. The populations of focus were groups of users such as chemists, doctors, and agronomists. The subject of focus was how the systems were used by these groups. Over time the emphasis shifted to look at more complex issues and began to examine the cognitive strategies of individuals. Models of user search strategies were created. Currently, the emphasis has shifted again to a qualitative focus. There are attempts to derive models from behavior rather than creating models to fit behavior. The greatest needs in user research are to integrate the results from studies and to develop long-term research agendas and projects. The work of Brenda Dervin and Carol Kuhlthau provide exemplars of longitudinal research based in strong theory and methodology.

There are also problem areas and questions that need to be more fully addressed in user research. Issues of use of libraries and information centers by social class (see 6.4.2), ethnicity, gender (see 9.5.4), and age are still underrepresented in the research. Characteristics of these users and issues of effectiveness, access, and satisfaction could use more concentrated study. Younger children are also underrepresented as subjects of research. Emphasis seems to be predominantly in academic settings or with professional people. Studying users of library and information centers may focus on the available pool rather than on the possible pool. Extending the populations studied to nonlibrary users and nontraditional information users, and focusing on people as well as institutions and systems, could increase the range and usefulness of user needs research.

Ongoing research issues in user studies include identifying what is user research, defining terms, and differentiating between a system in which users are a component and research that concentrates on users. Researchers who focus on specific user characteristics such as information search strategies or the information skills process often incorporate access issues into their research. This section has identified the specific issues related to users who influence the research process. In the next section, patterns related to the access of information are explored. This chapter identifies common components in each area first. It examines how theory and methods are being applied to solving specific problems related to users, access, and information. Comparing similarities, differences, and potential for future studies can isolate trends and commonalities for application. to library and information center problems.