Sense-Making Home Page

Articles, Papers, Commentaries

 

 

 

[BACKGROUND PAPER]
WHAT WE KNOW ABOUT INFORMATION SEEKING AND USE

AND HOW RESEARCH DISCOURSE COMMUNITY

MAKES A DIFFERENCE IN OUR KNOWING

by

Brenda Dervin
Ohio State University
Columbus, OH, USA
dervin.1@osu.edu


CITATION AND COPYRIGHT INFORMATION
© Dervin, B. 2001. Cite as: Dervin, B. (2001). What we know about information seeking and use and how research discourse community makes a difference in our knowing. Background paper prepared for Health Information Programs Development, National Library of Medicine, Bethesda, MD. Available [online]: http://communication.sbs.ohio-state.edu/sense-making/art/artabsdervin01nlm.html

ABSTRACT

The purpose of this position paper is to describe the current state of empirical study into everyday health information seeking and use and what implications these research findings have for the design of information systems to better serve user needs.  Drawing on the two fields that give the most attention to the study of information seeking and use - - library and information science (LIS) and communication (COMM), this position paper addresses how the two fields study information seeking and use differently and what differences these differences make in results.  The paper then draws propositional themes out of each body of work showing where the propositions converge and diverge.  A major conclusion is that while the two fields differ in numerous ways, they have both converged on a conclusion that expertise-based transmission systems do not succeed communicatively.  Bottom-line, what is needed to better serve user needs are systems and procedures built on dialogic communication principles which allow users to access systems and expertise in terms of the time-space contexts of their needs.  The first draft of this paper was submitted October 22, 2001 as a background paper to Eliot R. Siegel, PhD, Associate Director for Health Information Programs Development, National Library of Medicine, Bethesda, MD.  This version, while still dated 2001, has undergone systematic checking and corrections for all citations and references as well as an editing.  The paper will be updated based on a new literature review in 2005 and that revised paper will be linked to this one.    

 

OVERVIEW OF PURPOSE
My purpose in this background paper is to summarize what I understand about everyday health information seeking and use as supported by empirical work, and to draw out the implications of this understanding for research and ultimately for information system design.  My guiding focus is NLM's desire to get a handle on how MEDLINEplus can be maximally useful and used by its mandated users, and how to conduct research to create useful input for website design.

 

I will implement my purpose with a propositional journey through my understanding of everyday citizen health information seeking and use in a way that allows me to illustrate why I approach research problems as I do. To this end, in this essay I focus extensively on two academic disciplines and the work they do that is relevant to health information seeking and use. It was not my intent to provide such an extensive background paper, but as I dug into the literatures it became apparent that I had to, in order to communicate well. The two literatures that pertain most directly to health information seeking and use include one from the communications field (COMM) focusing on information campaign design; the other from the library and information science field (LIS) focusing on information seeking and use.

 

The difficulty is that there is very little overlap between these two literatures and, in fact, the portrait that one gets of information seeking and use from the two seems on the surface entirely unrelated to each other. Since my goal is to present a propositional overview of what I think we know with reasonable evidential support, I had to come to grips with the seeming incommensurabilities of these two fields. In short, I had a nagging feeling that there was insight to be gained from the effort even though it took far more time than intended.

 

Throughout this paper, I will refer the members of the mandated publics of institutions by a variety of names -- patrons, patients, users, publics, audiences, clients, customers, and so on, depending on the research literature I am drawing from. In all cases, my intended focus is the designated persons whom an institution or system is mandated to serve even if only a small percentage is actually served. In addition, I will draw on research from beyond the health information context when I think it helps complete the portrait. I do this because research shows, as I emphasize in my propositional presentation below, that the substantive focus of information seeking generally accounts for less variance in information seeking and use than other factors and that many important aspects of information seeking and use are generalizable across substantive contexts.

 

I proceed in five sections:

 

1: A background commentary on two academic disciplines

 

2: On health information seeking and use: The public communication campaign literature

 

3: On information seeking and use in context: The library and information science literature

 

4: On online communication: More of the same

 

5: Implications for research

 

SECTION 1:

A BACKGROUND COMMENTARY ON TWO ACADEMIC DISCIPLINES

What seems like a simple task is in actuality not so because it requires I share as briefly as possible my understanding of what these two disparate literatures have to say that is pertinent. The difficulty is that these literatures say different things.  It isn't so much that the literatures conflict, but rather that they provide potential insights about this or that piece of what seems on the surface to be two different puzzles. The puzzle pieces rarely seem to touch, or even in many cases to cohere around the same end.  

 

Even for those places where the puzzle pieces do seem potentially relevant to each other, the large differences in interpretive frameworks which drive the two literatures makes developing a cohesive picture extraordinarily difficulty. The difficulty is that the interpretive frames within which research is presented necessarily drives how evidence is collected and interpreted, and it is here that these disparate research approaches seem -- at least on the surface -- to be most incommensurate.

 

This problem requires attention because it is otherwise hard to understand why the portrait of citizen information seeking and use that one gets from viable but disparate literatures seems so entirely different. Does demography and personality predict information seeking and use, as one research genre seems to suggest? Or not, as another contends? Does cultural context provide avenues for more effective health provision, or not? Are quantitative surveys an effective way to learn about users? Or, are there efficient and effective alternatives? And, so on.

 

None of these questions have simple answers except when answered within narrow literature confines. If one looks broadly, little is straightforward because, in fact, the social sciences are, in important ways, in disarray (Carter, 1990, 1991; Dervin, 1991, 1993). There are wide varieties of competing "paradigms," which have very little communication with each other. In psychology, for example, there are sub-fields -- clinical, social, ecological, and humanistic -- which have very little overlap with each other and present increasingly disparate findings even about the same phenomena. In sociology, there is a marked divide between quantitative and qualitative approaches even when they focus similarly. The same is true of anthropology with its recently emerging thrusts into, for example, critical and dialogic anthropology in contest with traditional anthropology. And, for a variety of reasons, the communication field is even more marked by polarities and disparate camps without communicative bridges between. Communication mirrors sociology in its division into qualitative and quantitative camps; and is further riddled by lack of robust dialogue between U.S. scholars and communication scholars in the rest of the world. This is a striking loss because in fact there is strong evidence that capacities to dialogue across competing paradigms are stronger in the academy outside the United States, particularly in Scandinavia, New Zealand, Australia, and Canada.

 

Traditionally, we would think that these divides could be transcended with quality state-of-the-art-reviews and meta-analyses.  Indeed some useful ones do exist within genres. However, there are few that focus between genres because the fundamental difficulty is not a lack of synthesis of evidence, but rather a lack of tools for comparing seemingly disparate research approaches (Carter, 1990, 1991, in press; Dervin, 1991, 1993).  The ways in which these disparate research genres ask questions and answer them are so radically different that they appear incommensurate.

 

It is too easy to say that the approaches differ methodologically because in fact there is very little explicit attention to methodology in social science research today (Dervin, 1999a). In quantitative studies, methodology is reduced to method as if the mapping of random error is sufficient assessment; while in qualitative studies, it is usually reduced to philosophic critique. Little research attends explicitly to building bridges between theoretic assumptions and broadly-defined methods -- that is, the methods of framing questions, collecting and analyzing data, and interpreting results. Comparison of why different research genres approach the same phenomena in such different ways with such different results requires methodological comparison; unfortunately, we are bereft of even a discourse that would allow such a comparison.

 

Moreover, our difficulty synthesizing across research perspectives is compounded because we lack sophisticated tools for meta-analyses of research that are useful across different research approaches. While once (in the 70s) there was a trend emphasizing that all quantitative research should report variance accounted for, now we see such figures only in very constrained genres of attention (e.g., public communication campaign penetration and effect rates is a notable exception as exemplified by Snyder, 2001). When we move from more straightforward persuasion and impact studies to studies attempting to test complex theories, however, the problem of comparing results becomes more difficult. Finally, when we move from quantitative studies to qualitative, we see in some quarters a rejection of any kind of meta-analysis across contexts and more generally a lack of frameworks for doing so.

 

Despite these difficulties, my own research trajectory has traveled across these two genres in good part because a central focus of its mission has been to build bridges between them. Further, my work has been applied to a wide variety of substantive attentions (e.g. library use, health information seeking, electronic system use) which on the surface makes its trajectory look capricious. In actuality, in all applications I have focused on how to conduct research to assist in the design of more useful and effective communication and information systems and procedures.  This is an important point for understanding what I write below. I take it as a given -- an assumption based on evidence -- that there are more commonalities in information seeking and use across substantive contexts than uniquenesses.  The trick is to find a conceptual framework which allows these commonalities to show while at the same time honoring the uniquenesses of substantively different terrains.  This has been one of my major goals -- the search, if you will, for what I call the universals of human sense-making and sense-unmaking.  It is a premise of my Sense-Making Methodology that expanding our research foci beyond substantive terrains provides better avenues for accounting for more variance in human information seeking than the highly compartmentalized approaches that currently dominate our work.

 

The Sense-Making Methodology, which I have been working on since 1972, has been developed as a tool for this alternative way of thinking about information seeking and use. Its development has been pursued in many contexts with the primary mission being to find better ways to think about and study users (i.e., audiences, mandated publics, patrons, patients, citizens, readers, and so on), thereby permitting more effective research that more usefully informs the design of communication/information systems and practices. Constant challenges in this work have included, for example, how to conceptualize needs to permit comparisons across applications contexts, and how to conceptualize users to increase research capacity to account for variance in predicting use.

 

Even though the Sense-Making Methodology has been applied in multiple, seemingly disparate contexts, the research has all had the same purpose as specified above and is brought to bear in this paper as one cumulative body, including studies by Dervin and by others who have used the methodology.  The application contexts include studies of:

 

*Health information seeking and use

(Baker, 1998; Brendlinger, Dervin, & Foreman-Wernet, 1999; Dervin & Clark, in preparation; Dervin, Harlock, Atwood & Garzona, 1980; Dervin, Harping, & Foreman-Wernet, 1999; Dervin, Jacobson, & Nilan, 1982; Dervin, Nilan, & Jacobson, 1981; Dervin, Nilan, Krenz, & Wittet, 1982; Frenette, 1999; Nelissen, van Eden, & Maas, 1999)

 

*Environmental information seeking and use

(Madden, 1999; Murphy, 1999)

 

*Users of libraries and information systems

(Dervin, 1992, 1997, 1999a, 2001e; Dervin & Clark, 1987; Dervin & Dewdney, 1986; Dervin & Fraser, 1985; Dervin & Nilan, 1986)

 

*Everyday citizen information seeking and use and sense-making

(Atwood & Dervin, 1981; Coco, 1999; Dervin, 1984, 1997, 1999a; Dervin, Zweizig, Hall, Kwan, & Lalley, 1977a; Dervin et al., 1976, 1977b; Nilan & Dervin, 1999; Palmour, Rathbun, Brown, Dervin, & Dowd, 1979)

 

*Information seeking and use of professionals

(Cheuk, 1998; Cheuk & Dervin, 1999; Teekman, 1999)

 

*Information use by targeted campaign audiences

(Dervin, 1981, 1989a, 1995, 1997; Dervin & Frenette, 2001; Dervin & Shields, 1999)

 

*Communication/information gaps between "haves" and "have nots"

(Dervin 1980, 1982, 1989b)

 

*Researching journalistic audiences and designing more communicative journalistic systems

(Dervin & Foreman-Wernet, in press; Dervin & Huesca, 2001; Dworkin, Foreman-Wernet & Dervin, 1999; Huesca & Dervin, in press; Spirek, Dervin, Nilan, & Martin, 1999)

 

*The uses and perceptions of users of telecommunication systems

(Dervin, 1995; Dervin & Shields, 1999)

 

*Participant dialogues (in person and online) in multiple settings (e.g., informal, organizational, governmental, citizen participation)

(Dervin, 1994b; Dervin & Clark, 1993; Dervin & Huesca, 1997; Dervin, Osborne, Jaikumar-Mahey, Huesca, & Higgins, 1993; Dervin & Schaefer, in press, 1999; Schaefer, 1999)

 

*Organizational knowledge management

(Cheuk, 2002; Dervin, 1998)

 

*Patrons and audiences of arts institutions, popular and elite

(Dervin, Wyszomirski, & Foreman-Wernet, 2000; Foreman-Wernet & Dervin, 2000)

 

*Media practitioner encoding and sense-making

(Higgins, 1999; Huesca, 1999)

 

Along with Carter (1990, 1991, in press), Thayer (1978, 1982, 1996) and Stephenson (1980, 1995/6), I share a belief that if the communication field can construct methodologies that focus on communication as communication, we would be more able to free our field from its reliance on other disciplines.  The difficulty with the approaches of other disciplines is that they necessarily truncate their foci to particular levels of analysis (i.e., psychology for the individual; sociology for the collective; anthropology for the spatially bound collective) rather than to communication processes which serve as linkages between levels (Dervin, 1991, 1993, 1994a, 1997; Dervin & Clark, 1989, 1993).

 

It is because of this background that I am bringing to bear on this position paper my understanding of what we know about everyday citizen health information seeking and use, drawing on two primary literatures: the communication (COMM) field's study of health information seeking in the context of public communication campaigns; and the library and information science (LIS) field's study of information seeking and use.

 

Each of these literatures is, of course, served by other literatures. For example: health information seeking and public communication campaigns overlaps the extensive work in health communication generally, and also the work using personality predictors of health behaviors (e.g., locus of control as predictor of responses to uncertainty in health situations), the persuasion literature (e.g., testing alternative ways of framing messages and attempting to map these alternatives to cognitive predispositions and other personality attributes of audience members), and a variety of literatures outside the COMM field which focus on physician-patient interactions.  Similarly, information seeking and use research in LIS overlaps with related genres, for example: work on information retrieval; database searching; classification; indexing; and so on.

 

In my judgment, the two literatures I have selected for primary focus this this paper are most pertinent to the propositional narrative I want to present below and provide a sufficient portrait of relevant related literatures. Adding the related bodies of work adds detail but does not change the direction of the narrative.

 

I do occasionally add ancillary literatures from what could best be termed the "cutting edge" of health studies.  For example, I include work from medical anthropology, ecological psychology, and contextual psychology, as well as qualitative approaches to health communication, nursing, and public health. An interesting aspect of the difficulty one faces trying to get a handle on what we know in these terrains is that in fact the ancillary literatures rarely inform the primary COMM and LIS literatures.  This seems true even when their general sensibilities are in agreement and when the ancillary literatures add important new understandings. Another interesting aspect is that these cutting edge literatures tend to rise from the human support systems of health care delivery rather than its core expertise systems. It is generally fair to say that knowledge experts (e.g., subject specialists) focus on transmission more than communication while their support systems (e.g., librarians, nurses) focus more on dialogue.

 

In building my portrait of what we understand about health information seeking and use, it is important to understand that although my own work has a significant presence in both the COMM and LIS research genres, it does not have a straightforward relationship to either. Sense-Making Methodology (Dervin, 1999a) draws primarily for its methodological viewpoints on the work of theorists and methodologians in the COMM field but not the sectors of the COMM field that informs most of those doing research focusing on information transmission and campaigns.  Likewise, it could be said that my work is informed by interpretive and qualitative scholarship even though I apply the approach more often quantitatively while my doctoral students have applied it more often qualitatively.  In the COMM field when it comes to compilations that focus on information seeking and use, you usually see numerous citations to my work, and I am generally identified as in the top 5% of communication academics in terms of productivity and citations. But clearly, I am not a player in the COMM field's approach to campaign studies, which is also the primary terrain where the COMM field focuses on information seeking and use. 

 

On the other hand, in LIS many would label my work as the major force. Its impact has been measurably significant. Vakkari (1997) noted that Dervin & Nilan (1986) set new directions for studying information seeking and use in LIS and that ". . .Dervin has today the most marked metatheoretical influence on the studies of INS [information seeking] (p. 457). Williamson (1997) commented in the same volume that the theoretical contribution of Dervin ". . .has been profound. . ." (p. 337). Others in LIS have written review articles on Dervin's work (Morris, 1994; Neill, 1987; Savolainen, 1993) and there is currently a LIS doctoral dissertation in preparation in Australia focusing on the constructing of meaning and significance of Dervin's work in the LIS field. A citation analysis by McKechnie, Pettigrew, & Joyce (2001) named Dervin along with Kuhlthau as the two most cited theorists in human information behavior research. The differences in the impacts of my work between COMM and LIS are another illustration of how disparate these two fields seem.

 

However, both the COMM and LIS literatures offer us some understanding of health information seeking and use. Both fields focus on non-expert information seeking and use. From the field of COMM with its focus on health communication and communication campaign research, the traditional framework has focused on assuring that appropriate health care is delivered to citizens. The central theme of this work has been gaining compliance -- organizing informational messages and health care delivery in such a way that there is an increased probability that patients will comply with prescribed medical treatment or citizens will attend to their health in mandated ways. From the field of LIS with its emphases on information seeking and use, the traditional framework has focused on serving needs -- assuring that a seeker goes away from the information system with informational help.

 

Despite the differences in their mandates, up until the mid 1980s at a foundational level there were marked similarities in the literatures of these two fields. For both, the task at hand was one of getting the right information to laypersons. For the communication field this was conceptualized most often as a task of enticing attention, persuading viewpoint, and encouraging behavior, all by using appropriate information. For the library and information science field, this was conceptualized most often as a task of facilitating access and mediating system structures to link the user with a need to appropriate information.

 

The societal challenges that became increasingly evident as we moved into the 1980s impacted both fields. Both faced growing recognition of the diversities of their mandated publics, particularly in terms of gender, culture, and class identities. Both faced the exponentially growing information stockpile and the understanding that their "information" -- the stuff they transmit to their publics -- was but a tiny point in an increasingly chaotic, disorganized, uncontrolled, and rapidly growing information environment. And, both faced the cynicism and malaise increasingly associated with the idea of information (or knowledge) as having uncontested and neutral expert foundations.

 

For both fields what was formerly thought of as primarily a task of well-orchestrated information transmission became infinitely more complex. Given the logics of their original mandates it is not surprising that the two fields responded by pursuing different paths.

 

The COMM field saw itself facing growing evidence of the limited effects possible from health communication campaigns, and low compliance rates with medical prescriptions. Researchers set themselves for the most part to the task of finding ways to get the best results they could but with lower expectations. The result of this journey is exemplified by the excellent compilation by Rice & Atkin (2001) in which a wide variety of strategies are reviewed. In essence, the communication approach to health information remained one of focusing on persuasion and compliance but with more modest expectations and with greatly complexified roads for getting there. Research began to pursue, for example, multivariate models for predicting campaign impact (McGuire, 2001).

 

Many arguments could be advanced of why LIS traveled a much different road because in fact fundamentally both COMM and LIS as fields of practice rest their ultimate measures of accountability on their communicative success with individual members of their mandated publics. But fact is that the COMM fields had never relied conceptually on an assumption of effective communication to each individual member of an audience. In contrast, LIS as a practice measures itself on such personalized and effective services. Hence, the forces of now apparent diversity and complexity in society hit LIS practitioners in ways that only became visible much later to the COMM fields. The result in the LIS field was a period of deep reflexive evaluation of purpose and some of the first robust calls for "user-oriented" approaches to research.

 

The result was a marked turn in how LIS studied its users. The turning point is usually punctuated with the Dervin & Nilan (1986) review (Morris, 1994; Vakkari, 1997) focusing on information needs and uses in which they called for a turn from transmission theories of users to sense-making or interpretive theories. There are still transmission studies of information seeking and use in LIS but they co-exist side-by-side with interpretive studies, and discussions in LIS engage the contests and disagreements between approaches. These debates inform LIS research and have allowed LIS research to move in deliberate albeit struggling steps toward a more holistic approach and to more heuristic theorizing. In my judgment, this is why in general LIS is producing more useful research for understanding information seeking and use in an electronic information environment than is COMM.

 

As background it is important to understand that the same kinds of calls for an interpretive turn were made in the communication field (e.g., Carter, 1990; Dervin, 1991; Stephenson, 1980; Thayer, 1982) but in communication, for myriad reasons, a bifurcation resulted with the transmission orientation to communication and the interpretative orientation to communication ending up for the most part in separate literatures with relatively little overlap between. Hence, to this day one can readily identify in the field of communication those scholars who take transmission orientations to their studies and those who take interpretive orientations. The former are usually quantitative and the latter qualitative. The former most often rely on persuasion oriented theories to guide their research, focusing on what produces greater attention, attitude change, behavior change, and so on. The latter focus on understanding how information users make sense of the health related life situations or understanding the forces of society and culture that constrain such sense-making. Both poles argue methodological issues but usually only within their genres. For the most part even departments of communication in Class I research universities increasingly show the same bifurcation. And, for the most part, the literatures from the two poles do not inform each other. Hence, the meaning-making thrust in health communication literature exists quite independently of the focus on introducing change in the health communication campaign focus. In even a single journal devoted to health studies (i.e., Journal of Health Communication), one finds the same disparities.

 

Further, in actuality, explicit study of user information seeking and use has not been an explicit focus in COMM studies since the late 70s and early 80s. At the more quantitative end of the COMM field with its focus on transmission and persuasion, information seeking and use has been mostly hidden in studies of knowledge gain resulting from message transmission -- that is, do patients get the doctor's information; do campaign audiences learn the intended facts. At the more qualitative end of the COMM field with its focus on interpretation, studies have often marginalized the concept of information as outside its theoretic purview and focused instead of meaning-making. Or, they have examined how what society calls information is inscribed by and in societal forces.

 

The situation in LIS is entirely different. The same paradigmatic and methodological tensions exist but they are muted and remain in dialogue. The result has been the emergence of an identified literature genre now labeled "information seeking in context" established formally in 1997 with the publication of the first conference proceedings (Vakkari, Savolainen, & Dervin, 1997) and followed by, in turn: Wilson & Allen (1999); Hšglund & Wilson (2000, 2001).

 

An interesting impact of the LIS versus COMM field difference is that the LIS field is not as constrained intellectually by the current emphasis on cultural differences as the answer to communication dilemmas. Both LIS and COMM moved from transmission theories of communication to cultural relativistic theories and then to constructivist theories (which can be conceptualized as a more detailed relativism, now applied at the individual rather than cultural level). Both tumbled from there into the postmodern crisis (Dervin, 1994). What happened next differs because the COMM field bifurcated into divisions that represent the different steps of the journey -- transmission, cultural relativism, constructivism, and postmodernism. In contrast, LIS's studies of information seeking and use took all four perspectives onto one road. For this reason, in LIS studies of information seeking and use cultural differences have taken their place among a battery of alternative predictors. It is fair to say that for the COMM field the primary research work that is being applied to understanding information seeking and use and compliance still focuses on demographic, personality, and lifestyle differences as the primary means of differentiating users with cultural identification being a much-emphasized current approach. In contrast, in LIS the primary means of differentiation has moved to descriptors of context, situation, and practices. In an important way, in the COMM field the unit of analysis remains the person; in LIS, it has become the moment of intersection between user and system. As Zweizig & Dervin (1977) termed it, the focus has involved a move from looking at use to users to uses.

 

The discussion I have presented above rests, of course, on stereotypes. In between the cracks of the main thrusts of the portrait I present above we find ourselves in the best and the worst of times for social science scholarship. It is both exciting and confusing because the dust of the postmodernism crisis (e.g., objectivity vs. subjectivity) has settled a bit so researchers are searching for viable mediating alternatives, and because the new technologies offer flexibilities in literature review as well as data collection and processing not before possible. Young scholars in particular are abrading their boundaries, and wherever one looks at the cutting edge one sees calls for genuine dialogue across fields and at the same time calls for approaches that break out of old restraints. The propositional review below absorbs these innovations to the extent that cumulative evidence warrants it.

 

Finally, before proceeding, it is important to distinguish between the rhetorical needs of a political climate and what evidence says about human behavior. Two examples will suffice.

 

One concerns expertise, particularly medical expertise. Despite considerable evidence about the elusiveness and changeability of medical knowledge and the struggle of patients as well as doctors coping with overwhelming contradictions in the medical information marketplace, medical knowledge systems still rest in practice on transmission assumptions. Any recommendations regarding system design and practice must be sensitive to this gap between evidence and normative practice.

 

The second concerns cultural differences, the by-word of the current moment for accounting for differences between people. Evidence shows that in fact class, education, and geographic locale are usually far stronger predictors than ethnic identity per se; and further that when people who are culturally different see their worlds similarly they are more like each other in terms of sense-making and sense-unmaking than if they are culturally the same but see their worlds differently. Given this evidence -- which I will present below -- if we speak based on evidence alone we might suggest that there are more efficient ways to conceptualize differences than culture. In fact, cultural differences when reduced to ethnic and demographic attributes fast generate an astonishing number of different ways one must design and implement communication. Take, for example, only ten identity attributes with two values each. Intersecting them yields more than one thousand possible communication designs if the intent is to appeal to each identity sub-group (e.g., a group of older Hispanics, recently immigrated, living rurally, and so on).

However, the more fundamental issue for our societal moment is not the impracticality of what I will call below a "noun" framework for understanding difference. The more fundamental issue is that it is simply not rhetorically appropriate to de-emphasize cultural categories even if they do not predict well. Further, it is also true that evidentially there are conditions when culture is highly predictive. Therefore, the more important goal is to identify the conditions under which cultural differences are important to communication design. I will speak to the issue of how to navigate these contradictions below.

 

In proceeding below, I generally focus on the COMM literature in the first section of the presentation and the LIS literature in the second. Occasionally, I cite work from one area to support the other and vice versa. My format for the presentation of propositions is to present the propositions as a continuing coherent narrative illustrating under each proposition with relevant citations from the literature: studies, as well as theoretical essays and state-of-the-art-reviews. I do attempt to provide some degree of detail in reporting studies because I am attempting to find connections between work ordinarily thought of as disparate in focus, so I want to make the connections I see more visible

 

SECTION 2: ON HEALTH INFORMATION SEEKING AND USE: THE PUBLIC COMMUNICATION CAMPAIGN LITERATURE

2.1: SOURCE-ORIENTED: Traditionally, the COMM field has not focused on information seeking and use as a focal phenomenon for study. Rather the phenomenon of interest has been compliance gaining via message construction and campaign design. Traditionally, the approach has been source-oriented. The view is based on institutional maps of the world, not the maps of people-struggling-in-situations. The information presented is framed within a worldview as defined by the expert institution, which has often focused on victim-blaming and been ignorant of the experiential conditions and societal constraints within which laypersons make sense of the world. This impact of this difference on possible effectiveness seems fundamental given our basic understandings of how communication works; yet, this is still the major criticism to be launched against most public communication campaigns.

 

*Myrick (1998), in his assessment of the effectiveness of a state-run HIV prevention campaign in Alabama targeted to African Americans, noted that the target audience saw the campaign as involving a White power structure targeting a minority audience as in need of repair.

 

*Tardy & Hale (1998) reported from their participant observation study of mothers' conversations relating to health and illness that the conversations served to "crack the code" of institutional messages and practices and at the same time allowed the women to bond to each other. Their data led these researchers to say that ". . .what is labeled noncompliance in the institution could be labeled strength or survival outside the institution; not survival of the body, perhaps, but survival of identity and spirit" (Tardy & Hale, p. 153).

 

*Baer (1996) concluded from his study of perceptions of health -- physical and mental -- that the biomedical categories of health institutions may be inappropriate not only to ethnic minorities but also to the mainstream population. He said of his informants that ". . .their ideas and formulations of this domain of existence were so different from the ways these issues were categorized by the NCHS [National Center for Health Statistics] that the questions seemed irrelevant to their lives and very 'distant' from their daily concerns" (Baer, p. 63).

 

*Brendlinger, Dervin, & Foreman-Wernet (1999) concluded that a traditional health department survey of knowledge resulting from an HIV/AIDS awareness campaign tapped how well audience members had memorized the institution's story but was missing altogether how respondents constructed their own pictures.

 

*Sobo, Zimet, Zimmerman, & Cecil (1997) came to the same conclusion in their review of AIDS campaigns. They suggested that in their study of adolescent runaways, high AIDS "knowledge" scores reflected rote memorization rather than comprehension and application.

 

*Nelissen, van Eden, & Maas (1999) in their Sense-Making study of practitioners and patients in a comprehensive cancer center found that while health practitioners focused on medical information, patient question-asking focused more on non-medical questions.

 

*Dervin, Jacobson, & Nilan (1982) found much the same in their study of information seeking and use by blood donors -- much of the question asking focused on issues which information systems traditionally term "subjective." In this context, many donor questions focused on blood donors' struggles with themselves.

 

*Dejong & Wallack (1999) evaluated the drug czar's antidrug media campaign as ignoring the best of campaign research in its development. In particular, the campaign neglected issues of audience involvement and efficacy.

 

*Any number of campaign studies still take a blame-the-victim stance toward campaign failures. Often, these unwittingly take on racist and stereotyped stances. For example, Donnelly, Mowery, & McCarver (1998) judged their inner-city African American mothers as having inadequate knowledge about substance abuse. Yet, knowledge levels were above average when compared across a large number of campaigns in a variety of communities (Snyder, 2001). Or, Stephenson & Witte (2001) concluded that ". . .most people are incapable of accurately assessing their own vulnerabilities to risky health behaviors" (p. 98). Yet, Klingemann (1995) challenged risk related research as being obsessively oriented toward defining risk taking as negative when in fact voluntary risk taking is part of both human life-facing and even human pleasure-seeking. And, Dervin, Harpring, & Foreman-Wernet (1999) emphasize on interviewing approaches can easily yield distorted and shallow pictures of what is really going on in how people make sense of their health struggles.

 

*Hillier, Harrison, & Warr (1998) evaluated safe sex campaigns to adolescents and concluded that they neglected teenager social contexts and the realities of their lives and paid virtually no attention to assisting teens with the barriers they saw to implementing safe sex alternatives.

 

2.2: BAD GUESSES: Because of their focus on institutional frameworks, message and campaign designers often make wildly wrong guesses about what publics do and do not want. These guesses are usually erroneously confirmed by the usual approaches to survey research because of the ways in which researchers unwittingly constrain their data collection to yield the very answers they expect.

 

*Carlson, Siegel, Wang, Jichaun, & Faick (1996) in their ethnographic and survey studies found evidence that contradicted the common myth in public health literature suggesting that drug users find needle-sharing a desirable behavior.

 

*When the California Public Utilities Commission mandated informed consent for its citizenry before the introduction of caller-ID, a marketing firm proposed an education campaign based on the assumption that the public for the most part would not care. Dervin (1995) noted in her assessment of the proposed campaign that research increasingly showed the general public to be very attentive to privacy issues.

 

*From qualitative audience and policy studies of health mandates have come insights on how health campaigns ignore social realities. Dievler & Pappas (1999), for example, noted how the common focus on race and ethnicity as a predictor of health behaviors and susceptibilities ignores the fact that social class and economic issues are as likely to offer better explanations. The problem is made even more difficult because popular media sometimes offer highly distorted views of medical realities, as for example in television's treatment of HIV/AIDS as a deviant and stigmatizing disease (Tulloch & Lupton, 1997).

 

*In a Sense-Making study of information seeking in leisure time situations, Spirek, Dervin, Nilan & Martin (1999) compared information seeking as it occurred while reading newspaper leisure time coverage versus as it occurred in natural everyday settings. The questions asked were significantly different. Further, a comparison of question-asking in natural situations by college students versus those asked by a general population sample showed few significant differences.

 

2.3: COMPLIANCE IS AN INSTITUTIONAL CONCEPT: Compliance as a construct is derived from an institutional map of the world. Human beings struggling with situations as they move through time-space do not refer to their actions as compliant or non-compliant. They naturally refer to struggles, useful information, useless information, constraints, and so on.

 

*The struggle with "who is the blame" for the relative lack of success of compliance gaining activities has been, of course, a continuing theme for attention since Bauer's (1964) germinal piece on "the obstinate audience." Dervin (1981, 1989a) summarized this work by suggesting the operating metaphor was one of conceptualizing audience members as if they were empty buckets. The obstinate audience includes buckets that are either leaky (with holes in the bottom), or stubborn (with lids put on tightly).

 

2.4: DECREASING TRUST: Decreasing public confidence in expert information and the institutions mandated to supply it has a major impact on communication effectiveness making both attention and impact harder to achieve. This phenomenon is generalizable across the U.S. population.

 

*A Pew-funded study (http://www.pewtrust.org/, 1997) documented that citizen evaluations of governmental institutions were lower than their evaluations of trades people and local news -- at the bottom of the institutional credibility grid.

 

*Sobo, Zimet, Zimmerman, & Cecil (1997) cited a series of studies showing "considerable mistrust" of statements regarding HIV/AIDS made by experts. Some studies showed as many as 20% of African American men as seeing AIDS as a governmentally sponsored form of genocide. In one study, the figure was 65% for African American religious leaders. Other studies showed that as many as 40% of adolescents did not trust government information on AIDS.

 

*Siegel & Raveis (1997) found their minority (African American and Puerto Rican) male informants both judged social service organizations as disinterested in them.

 

*Guttman, Boccher-Lattimore, & Salmon (1998) found a widespread mistrust in expert information regarding the probability of AIDS transmission.

 

*Tardy & Hale (1998) found their informants -- young mothers in a toddlers' playgroup -- distrustful of physicians.

 

2.5: IGNORING SOCIAL FORCES AND ETHICS: A recent trend in campaigns research -- an important one in the context of this background paper -- is the turning of attention to placing the campaign in a larger political, social, and economic context. It has been widely argued in recent work that this lack of attention has been a major weakness in both campaign research and design. Ignoring these larger contexts, some charge, is at best simply bad top-down, transmission-oriented communication. At worst, it treats target audiences as dupes without awareness of their larger environments.

 

*Proctor (2001) warned that the prescriptive and eliminative aspirations of public communication campaigns could have an unfortunate similarity to the same kinds of aspirations manifested in the public education goals of the Nazi regime. He called for reflexive examination of purpose particularly vis-â-vis how goals may marginalize and stigmatize in ways parallel to negative forces in society. He has also called for attention to how politics shapes the information that is the focus of communication campaigns relating to cancer (Proctor, 1995).

 

*Myrick (1998) noted how important it was that the HIV/AIDS campaign in Alabama targeted at African Americans be aware of the adversarial history that had been established in the past when Tuskegee stigmatized and targeted blacks with their syphilis experiments.

 

*Guttman (1997), in an important review article, discussed the ethical dilemmas faced by health campaigns and often ignored in campaign design and research. Issues she raised included: the ethics of persuasion and coercion; the possibility of infringing on individual rights; the problem of targeting certain sub-populations as particularly needy and the resulting stigmatization of population sub-groups; the potential exploitation of family, community, and cultural groups as vehicles for implementing campaigns; the de-emphasis on a broader vision of social values.

 

*DeJong & Wallack (1999) criticized the drug czar's antidrug media campaign for its lack of attention to societal infrastructures (e.g., treatment programs) that could support the campaign's objectives.

 

*Dervin & Frenette (2001) challenged that campaigns and campaign research as must focus on campaign and compliance efforts by attending to issues of social power and ethics, how social class is often implicated in expert knowledge, how audiences are increasingly weary of being judged wanting, and how audiences are increasingly savvy about contradictions in the information marketplace.

 

*In noting that there is in actuality little agreement about what constitutes an information campaign effect, Salmon & Murray-Johnson (2001) noted that the conceptualization of effectiveness has been more an accident of different researcher's rhetoric than coherently adopted conceptualizations. They offered a six category conceptualization as a potential organizing tool for research: definitional (getting a phenomenon defined as a social problem); ideological (implementing a definition of the problem as either individual and/or social); political (using the campaign to garner political support); contextual (assessing the match and interaction of selected mechanisms for change -- engineering vs. enforcement vs. education); cost (campaign costs not only in terms of penetration, but in terms of savings to society and population); and programmatic (the specific goals as defined for a given information campaign). Their categories clearly made room for a specific focus on societal forces.

 

*This trend is supported by an increase in campaign studies focusing not on understanding audience members but on understanding the societal mechanisms that promote particular behaviors. Alcohol use is a prime example where attention has turned in good part to eliminating sources of media messages that promote alcohol (Casswel, 1995, 1997).

 

*Another example of a discourse analytic approach to studying health communication is Jones (1998), who analyzed the way in which media, government, public, and celebrity each brought their own theories of self and communication to bear on the construction of AIDS celebrity in Hong Kong.

 

*Another example comes from Harrison & Cantor (1997) focusing on understanding how media presentations of thinness contribute to eating disorders.

 

*Others calling for campaign research and design to include emphasis on social contexts, societal power structures, and ethics include: Lupton (1994); Lyons & Rittner (1998), Rakow (1989), Wallack (1989); Weiss & Tschirhart (1994).

 

*One frequent criticism is that campaigns in attempting to avoid controversy leave behind reality as living people know it. One often cited example is the reference in HIV/AIDS campaigns to bodily fluids instead of tears, sweat, salvia, vaginal fluids, and so on (Ratzan, Payne, & Massett, 1994).

 

*Singer (1998) provides an interesting counterpoint to the above arguments in his book on the political economy of AIDS. He argues that AIDS policy is implicated in societal racism, classism, homophobia, and so on, but he notes that changing entrenched societal conditions is not an easier challenge than changing individual behavior. The point the above authors make, however, is that ignoring societal conditions introduces a communication gap between institution and intended audience.

 

2.6: LOW IMPACT: Given everything we know about what makes communication work more effectively, it is not surprising therefore that in general attention to mass mediated campaign messages ranges from usually dismal to moderate; and compliance with medical expertise is rarely higher than 50% and often as low as 1-2%. A general caveat is that public communication campaigns work at best modestly well and do so at great cost. In general, the larger the campaign budget the greater the exposure and the redundancy of exposure and the greater the impacts. Other factors relating to higher impacts include legal sanctions and a chance intersection with some issue that has independently captured public concern.

 

*A host of authors has commented on the limits of campaign potential: Brendlinger, Dervin, & Foreman-Wernet (1999); Freimuth & Mettger (1990); Myrick (1998); Ratzan, Payne, & Massett (1994).

 

*Atkin (2001) summarized the relatively modest effects from campaigns, noting that effects varied depending on the behavior change mandated, the receptivity of the audience, and the quality and quantity of messages.

 

*In the campaigns he reviewed, Snyder's (2001) meta-analysis of campaign effects found exposure rates ranged from a low of 13% to a high of 99% with an average of 40%; and behavior change ranged from a low of 1% to a high of 41%; with an average of 7-10%. Impacts on behavior were higher when the behavior involved legal sanctions (17%); and when introducing new behaviors (12%) rather than changing old ones (5%) or preventing behaviors deemed negative (6%).

 

*Exposure to campaign messages is a necessary condition for behavior change, of course, and most frequently described as a linear positive relationship. Kane, Gueye, Speizer, Pacdquemargolis, & Baron (1998), for example, found the impact of a family planning campaign in Mali was directly related to intensity of exposure.

 

*Occasionally a campaign has large impacts and usually this means the public had a readiness already established by other means. A recent example outside the health communication terrain illustrates the point: B. Givens (personal communication, June 11, 1996) reported on the public education campaign on caller-ID mandated by the state of California. The essential design of the campaign was mandated by Dervin (1995) in her evaluation of the initially proposed public education campaign, which she deemed seriously inadequate. Dervin concluded that because of increases in privacy intrusions resulting from advances in telecommunications technology, the public was already alert to privacy issues. When a campaign was built around the slogan "your phone, your privacy, your choice," in a 3-month period penetration rates were an exceptionally high: 74% general awareness; 69% knowledge of how to prevent exposure of phone numbers; and 50% planning to choose maximum security protection.

 

2.7: RICOCHET EFFECT: A phenomenon that is not uncommon in public communication campaigns is the ricochet effect where the campaign produces effects opposite to those intended.

 

*Hafstad & Aaro (1997) in their study of a mass-mediated anti-smoking campaign targeted at adolescents found that a relatively high proportion of female smokers decided not to quit as a reaction to the campaign.

 

*Herrmann & McWhirter (1997), in their review of efforts to impact young people's smoking, substance abuse, and sexual activity, found that some interventions inspired the very behavior they were designed to prevent, and others inspired refusal and resistance not to the targeted negative behavior but to the desired campaign effect.

 

*McGuire (2001) in reviewing the campaign literature specifically emphasized the ricochet effect as being possible depending on the particular confluence of mediating factors along any step in what he described as the multi-step campaign influence chain.

 

2.8: MULTI-STEP CAMPAIGN MODELS: The dominant response to the above within the compliance-gaining and campaign literature has been to develop complex predictive models incorporating measures of audience members' prior experience, beliefs, prior knowledge, attitudes, and a host of other factors. While this has increased predictive capacity, its important impact has been to change how planners focus on message and campaign design -- moving slowly, and sometimes reluctantly, from more institutional oriented frameworks to more receiver-oriented frameworks.

 

*McGuire (1989, 2001) has provided some of the best overviews of this body of work presenting what he terms a communication persuasion matrix composed of groups of factors in multiple domains: source, message, channel, receiver, and target goal. He also conceptualizes the persuasion process in a series of steps: tuning in, attending, liking, comprehending, acquiring relevant skills, agreeing, storing in memory, retrieving from memory, deciding to act, acting, cognitive integrating, persuading others. He describes the process as a very complex multi-step chain which can easily ricochet to opposite impacts depending on the intersections of mediating factors.

 

2.9: MORE LIKE US EFFECT: In general, studies show that those most impacted by compliance-gaining messages and campaigns are more likely to be more like the very persons designing those campaigns in terms of demography, personality, and lifestyle. Those "more like us" were also more likely to use sources, and channels, and evaluate source credibility as we do.

 

*Guttman, Boccher-Lattimore, & Salmon (1998) found in their survey that those more likely to be informed about AIDS and expert sources of AIDS information were the more educated. They also saw the experts as more credible.

 

*Rimal, Flora, & Schooler (1999) concluded that in general education was positively associated with most behavioral impacts from campaigns.

 

*Dervin (1980, 1981, 1989b) reviews literature documenting the ways in which audience members most impacted by campaigns are more like system designers demographically and otherwise. She documents the same phenomena for LIS systems as well. She terms the research genre that focuses on "knowledge gaps" as an institutionally-inscribed view of the world. Those least like the experts are most likely to be ignorant of expert knowledge. The difficulty with this conceptualization, however, is that it ignores context and life conditions. Most health experts, for example, would be ignorant of most aspects of migrant worker life and survival, or of inner-city life and survival. In this context, the experts would be in "knowledge gap." Dervin concludes the knowledge gap hypothesis, while useful in some social planning contexts, must be applied cautiously and must be examined reflexively so as to avoid insofar as possible culturally and economically insensitive stigmatizations. Further, she emphasized that designing systems based on knowledge gap hypotheses is doomed to failure.

 

*O'Keefe, Boyd, & Brown (1998) found in a large Midwestern survey of exposure to health information that their more educated respondents were more likely to learn from print media.

 

*Trotter et al. (1999) showed beliefs about AIDS in four culturally diverse Latino communities were better predicted by education and income. The higher these indicators were the more likely respondent's beliefs were closer to biomedical beliefs. This work is an example of a medical anthropology project that has concluded that the important gaps in communication in health care settings are not between ethnic groups but between experts and laypersons (Baer, 1996).

 

2.10: INFORMATION IS NOT ENOUGH: A general caveat of the campaign approach to understanding information seeking is that information is usually necessary but rarely sufficient. Across studies, the stability of this finding seems to hold. It is, however, confounded by usual survey study emphasis on highly detailed attributes of medical knowledge -- the kind that average citizens deem unnecessary to their own health-facing approaches except for when their situations warrant it. Hence, there is some counter-evidence that when people come to information in a state of readiness information may be enough. This issue is discussed below.

 

*Bakker (1999) reviewed literature documenting that knowledge about AIDS was not sufficient to cause behavioral risk reduction. Other mediating variables accounted for the change. Among others making the same point: McGuire (1989, 2001); Rimal, Flora, & Schooler (1999).

 

*Davis, Noel, Chan, & Wing (1998) showed that while their large sample of Hong Kong adolescents had a lot of misinformation about HIV/AIDS, as judged by expert standards, in fact few reported that they engaged in at-risk behaviors.

 

*Brendlinger, Dervin, & Foreman-Wernet (1999) found that while the health department wanted assessments of the knowledge levels of respondents on a battery of detailed HIV/AIDS facts (e.g., can a baby get the virus from a nursing mom? does it take only a few days to test positive? can you get infected by a mosquito? and so on), their Sense-Making survey respondents focused on these items only if their situational conditions warranted it.

 

*Myrick (1998) in his review of HIV/AIDS prevention campaigns directed at rural African Americans concluded that while most campaigns do successfully disseminate information about the presence and danger of HIV/AIDS they rarely provided contextualized information relating to specific behaviors and to skills for modifying behaviors. Myrick saw this emphasis on behavioral efficacy as vitally important for campaigns targeted at ethnic minorities.

 

2.11: NOVELTY IS NEEDED: There is some evidence that the information that is presented even if it does not directly impact change does need to be "new" in some way. For some, this involves linking information with societal approved stimulators such as entertainment and celebrity endorsement. For others, it involves simply a need for new information because old information has already been processed.

 

*Donohew, Lorch, & Palmgreen (1998) tested message strategies based on presenting information in novel and sensation-arousing ways and found that in general they got significant impacts but their results were stronger for those with compatible cognitive needs.

 

*Petty & Cacioppo (1996) presented evidence lending some support to the idea that in fact depending on the audience involvement with the message the kind of novelty used may or may not have impact. They described what they called a "peripheral route" focusing on supplementary cues such as celebrity endorsement and a "central" route focusing on cognitive engagement and decision-making.

 

*Snyder (2001) in his meta-review of campaign studies noted that campaigns offering new information had higher behavior change impacts (14% compared to the 7-10% on average).

 

2.12: TREAT PEOPLE AS IF THEY ARE HUMAN: In general research agrees that recipients of health information are more receptive when they feel respected and are treated as capable and able agents rather than as recipients of behavioral modification theory. The most general form of this understanding is manifested in a call to turn from transmission approaches to communication approaches, focusing on audiences as active and not passive, focusing on process and not outcomes. In effect, it is fair to say that this call has become a general rhetorical stance in campaign and information seeking literature. Even those implementing top-down transmission campaigns still talk about them communicatively. But, bottom line, the call to being communicative has slowly meant changes in either campaign design and/or how campaign and information behavior research is approached. This is more true for LIS research than for COMM research, although here I focus primarily on the latter.

 

*Clatts (1994) challenged in his conclusions after 10 years of AIDS ethnography that ". . .AIDS prevention has become an exercise in behavioral modification theory, an enterprise that seems to me to have more to do with social control than with the prevention of a disease" (pp. 94-95).

 

*In a review of health campaigns in general, Rimal, Ratzan, Arntson, & Freimuth (1997) called for involving target audiences in genuine dialogue and in changing public policy.

 

*Ruben (1993) did a content analysis of critical incidents in health care focusing on what patients remembered and concluded, as others have, that more than anything else patients want to feel listened to, respected, and cared for.

 

*Tones (1996) offered a deep critique of the social marketing approach to drug use education, prevention, and policy.

 

*Myrick (1998) called for attention to modeling campaign and health information efforts as involving mutual responsibility.

 

*A variety of other authors warn about using this or that technique to approximate dialogic communication when in fact the intent is to ensure compliance (Guttman, 1997; Opie, 1998; Ratzan, Payne & Massett, 1994).

 

2.13: ADDING COMMUNICATION INTERVENTIONS: The most common version of the implementation of this call to focus on campaigns communicatively has been to actually add communication interventions to campaign design in the form of interpersonal, group, family, and/or community activities. The underlying assumption here is that it is not possible to package messages in such a way to promote this kind of communicative impact from within the message itself. Therefore, communication activities must become carrier. This change has consistently been shown to increase campaign impacts, improving exposure, awareness and behavioral change, but within limits and with the possible ricochet effects identified above.

 

*Rimal, Flora, & Schooler (1999) documented that exposure-induced information seeking and discussion accounted for four times more variance in health behavior change than exposure alone.

 

*Rimal, Ratzan, Arnston, & Freimuth (1997) in their review of campaigns concluded that the difference between the usually low impacts from campaigns and those that had gained moderate successes were due primarily to explicit interventions focused on increasing involvement.

 

*When Valois, Adams, & Kammerman (1996) combined interpersonal communication (live phone-in support sessions) with mass communication (televised program) on smoking cessation, results showed a success rate on behavior change at the end of one year of 17% compared to 5-15% for previously televised only efforts.

 

*Dervin (1995) in her review of campaigns emphasized the importance of adding interpersonal, family, and community interventions as a substantial part of campaign design.

 

*As part of the Stanford Five-City Project, Rimal & Flora (1998) emphasized the need to focus on families, as opposed to individuals, to change dietary behaviors. Their campaigns were more successful when they conceptualized both children and adults as potential sources of influence. In another report, Rimal, Flora, & Schooler (1999) emphasized the importance of interpersonal communication in campaign impacts.

 

2.14: THE CULTURAL ROUTE TO