Communication Institute for Online Scholarship
Communication Institute for Online
Scholarship Continous online service and innovation
since 1986
Site index
ComAbstracts Visual Communication Concept Explorer Tables of Contents Electronic Journal of Communication ComVista

Article from ejc/rec Electronic Journal of Communication
EJC logo
The Electronic Journal of Communication / La Revue Electronique de Communication

Volume 9 Numbers 2, 3, 4 1999

Sense-Making and Television News


Mark Dworkin
Moving Images Video Project
Whidbey Island, Washington USA

Lois Foreman-Wernet
The Ohio State University
Columbus, Ohio USA

Brenda Dervin
The Ohio State University
Columbus, Ohio USA

Abstract.  This article reviews David Morley's well-known Nationwide audience study as a way of considering some of the issues in the field of cultural studies, particularly regarding audience reception analysis. One overriding concern is the search for evidence of subordinated decodings (i.e., decodings of media messages that accept societal patterns and relationships as given) in view of the current trend of documenting interpretive resistance. We present an alternative way of studying audiences using the meta-theoretic premises and methods of Sense-Making Methodology and offer an empirical example of an audience reception study executed using Sense-Making's qualitative sensibilities and access to quantitative analytics. Important findings include: documenting that a majority of a general population random sample of TV news viewers exhibited subordinated decodings; and, at the same time, finding that a Sense-Making derived measure of societal involvement state, which tapped situational anchorings and phenomenological interpretings proved to be the only powerful predictor of social criticism decoding. An interesting clue pointing to further research showed that in fact under some conditions labor class audience members exhibited less negotiated decodings of media messages than not-labor class audience members (consisting of managers and professionals). The paper concludes with a review of what methodological aspects of Sense-Making were seen as facilitating the search for evidence of both resistant and subordinated decodings.
The Problematic of the Audience

Although it has been nearly two decades since its initial publication, David Morley's (1980) audience study of the BBC television news magazine Nationwide remains an important benchmark in the field of cultural studies and, in particular, in the area of audience reception studies. In addition to its continued presence through comparisons and citations in the cultural studies literature 1, the work has been given new energy through publication of a book that includes reprints of the two Nationwide studies, combining for the first time Morley's audience study and the slightly earlier detailed textual reading of the program by Charlotte Brunsdon and Morley (Morley & Brunsdon, 1999). 2

Morley's audience study broke new ground by empirically testing viewers' differential decodings of television news. The study, which was developed within the context of the Birmingham Centre for the Study of Contemporary Culture at a time of industrial conflict and racial unrest in Great Britain, focused on the ideological relationship between the media and the state. As such, the Nationwide study represented the first empirical application of Stuart Hall's (1980) encoding/decoding model of the circulation of meaning in television discourse. 3

Hall's model included a conceptualization of three hypothetical ideological positions that audience members might take vis-a-vis the broadcast message. The first was the dominant-hegemonic position, which represented the ideological perspective of the dominant elite and was the framework in which the broadcast message was encoded as well as decoded, or interpreted, by the viewer. The negotiated position was conceptualized as occurring when the viewer privileged the dominant ideological definitions in the abstract while reserving the right to negotiate the meaning of specific situations. In the third case, viewers who took an oppositional position dismantled the broadcast message in the dominant code in order to recast it within an alternative frame of reference.

Morley's study was designed to look at the relationship between the dominant or preferred reading of a news text and viewers' social class. To collect his data, Morley showed one of two Nationwide programs to different groups of students at different levels in the educational system, in different parts of the country, and from different social and cultural backgrounds. On the basis of brief (about 30 minutes) free-ranging group discussions after viewing, Morley attempted to distinguish preferred, negotiated, and oppositional decodings of the program based upon the social class of viewers (Morley, 1980; Morley & Brunsdon, 1999).

The thinking at the time for Morley and kindred scholars (e.g., Hall, 1977; Parkin, 1973) was that news broadcasts might be differently interpreted by members of varied social groups because divergent social experiences lead to alternative social outlooks. These differing opinions and philosophies are conceptualized as supplying the perceptual and conceptual bases that people use when they make sense of mass media broadcasts. The transcribed responses of discussions in Morley's study evince immense variability in people's responses to the news program. Morley expressed an inability to conduct a quantitative analysis of his data, however, and in his qualitative analysis he found no systematic relationship between class and decoding.

In a comprehensive "Postscript" (1981), Morley largely faulted his operationalization of social class and the criteria he used to establish a standard "preferred reading" to which alternative decodings were compared. But Morley did not attempt to ascertain the conceptual frameworks employed by his respondents in making sense of the program he showed them; rather, he sought simply to associate differential decodings directly with the social class of audience members.

Despite its failure to confirm Morley's hypothesis, the Nationwide study brought to bear a number of important issues and tensions in cultural and media studies that remain of interest to scholars in the field. 4 First, Morley's focus on class as a predictor of audience decoding didn't hold up empirically, which brought to the fore concerns about the interpretation and generalizability of such research. Second, Morley found problematic the conceptualization of textual readings of viewers as "preferred," "negotiated," or "oppositional," which called into question the usefulness of the dominant ideology thesis of critical scholars. This, in turn, reinvigorated the debate regarding the role of social structure versus individual agency in reception of mass media by audience members and hence, by implication, for any theorizing about media effects.

Current Issues in Audience Research

In large part as a response to the Nationwide study, many researchers have shifted focus from a dominant or preferred reading stance toward the notion of a polysemic text, or one that can be (and is) interpreted in many different, often subversive, ways 5. his active audience perspective is part of a larger movement from a more sociological, structural orientation toward a more focused situational or ethnographic approach in audience studies. Indeed, Morley's own work since Nationwide has paid much more attention to context, especially to the domestic setting, where most media consumption occurs (Morley, 1992).

Of course, these research trends are not without controversy. For example, while Morley (1993) acknowledges that his Nationwide study presented evidence against the dominant ideology thesis, he disagrees with what he perceives to be an undocumented presumption within active audience theory that interpretive resistance is more widespread than subordinated meanings. The shift in level of analysis also poses continued difficulties for Morley and others. 6 In one of the stronger indictments of the ethnographic trend, for example, Osterud (in press) argues that the current focus on concrete practices of individuals has resulted in a tendency to jettison structural considerations altogether. 7

A related research issue in cultural and other interpretive studies is the perceived inadmissibility of quantification and the disdain for statistics, even when such methods would strengthen and complement basically qualitative approaches (Murdock, 1997). And finally, a number of scholars recently have noted that there is still a relative dearth of empirical studies from which to draw. 8

Sense-Making Methodology as an Alternative

An alternative path, which we believe has the potential to overcome many of the difficulties outlined above, is the use of the Sense-Making Methodology developed by Dervin 9 and her colleagues. This approach has in the past been useful for the study of human sense-making (and unmaking) in a wide variety of contexts. 10

Sense-Making begins with the assumption that reality, our personal, social and physical world, is orderly in part and chaotic in part and cannot be known completely. What we understand is always itself subject to change as well as subject to multiple interpretations. Further, interpretations themselves have the potential for constant change. To conceptualize in this way does not mean that Sense-Making denies structure or constancy. Rather, structure and constancy as well as chaos and fluidity are placed within a framework of movement through time-space. It is a central premise of Sense-Making's methodological mandate that neither stability nor change, structural constraint nor resistance and innovation can be adequately studied within a framework that assumes constancy. In essence, it is assumed that it is only by implementing a framework that invites attention to change that constancy across time-space can emerge systematically for empirical observation.

Sense-Making encapsulates its emphasis on movement in a central metaphor picturing humans journeying through time and space engaging in day-to-day activities, encountering gaps related to assumptions about reality and the nature of knowing, bridging these gaps and moving on. This metaphor points to a highly abstract set of metatheoretic assumptions, not to any limiting framework suggesting that human sense-making is only linear or purposive or even orderly. Rather the assumption is that gap-bridging is a mandate of the human condition and takes on myriad forms, often repeating past practices (which maintain structural conditions) but sometimes innovating, inventing, resisting, colliding, and so on.

Following Carter (1972), Sense-Making mandates methodological attention to individual steps of meaning-making: sense is made and unmade incrementally, moment by moment, often and always. Thus, communication is seen as a "verbing" process (Dervin, 1993; in press) whereby the focus is on the movement (or constraints) in ongoing gap-bridging efforts to make sense of the world.

In addition to mandating attention to movement through time-space, Sense-Making mandates a conceptualization of the person as embodied, as having thoughts and feelings, confusions and questions anchored in material conditions. In this way, Sense-Making pays attention to both the inner and outer worlds of individuals and links human agency and societal structures. The link, Sense-Making posits, is in the step-takings -- the sense-makings and sense-unmakings that create, maintain, challenge, and resist structures.

With its attention to the in-between, to the verbings by which humans enact their journeys, Sense-Making presents a methodological mandate to incorporate attention to sources of power, both those forces that facilitate movement and those that constrain it. In this way, Sense-Making acknowledges, for example, that people sometimes are controlled by hegemony but that they themselves also are sites of power with the capacity to create, transform, or resist social forces.

In data collection, researchers using Sense-Making are mandated to share with those they are researching (e.g., respondents, informants, participants) the power of theorizing about themselves and about communication, society, media, and life experience. To every extent possible, the Sense-Making interview does not name the world according to the researcher's views but invites respondents to describe their own worlds, in their own terms: conceptualizing situations, making connections, hypothesizing outcomes and causes, and identifying gaps. This process is conceptualized in Sense-Making as dialogic. The researcher is mandated to a new kind of listening while respondents are asked questions which do not name their worlds substantively but which invite them to speak without forcing them to fit their speech to the dominant discourse. 11

Central to Sense-Making is the idea that this dialogic exchange involves bringing the unsaid or the unarticulated into consciousness, a process that Freire (1970) labeled as conscientizing. In this framework, the Sense-Making interview invites respondents to theorize their worlds, to consider relations of power, connections among past and present experiences, and so forth. Sense-Making implements this process in its interviews both in individual and group settings by asking a set of seemingly simple and universally human questions: what happened? what did you think, feel, conclude about what happened? what confused you? how did these thoughts, feelings, conclusions, confusions relate to your past life? to your expected future? to forces and constraints in your world?

While Sense-Making's questions (Dervin & Dewdney, 1986) seem simple, Sense-Making mandates that they be asked by focusing on specific micro-moments in time-space and that they be asked by an interviewer-researcher who absents from mind insofar as possible any predetermined map that might constrain listening. In essence, Sense-Making invites the researcher to a new framework for listening, a framework based on comprehending how another made a special kind of journey, a sense-making and sense-unmaking journey.

Also central to Sense-Making as a methodology is the assumption that both quantitative and qualitative approaches to research are required and that the polarization between these that has marked the past two-plus decades in the social sciences has been clouded with misrepresentations and misunderstandings. >From one perspective it is fair to say that Sense-Making is inherently a qualitative approach to research, focusing as it does on qualities of sense-making and sense-unmaking. But the body of work using Sense-Making has represented approaches both highly quantitative (e.g., multivariate multiple regressions, systematic content analyses) and highly qualitative (e.g., participant observations, in-depth interviews, case studies, thematic text readings).

The study we report on here applied Sense-Making to audience interpretations of the news in two ways. The first was by implementing the Sense-Making analytic in the framing of research hypotheses. The second was by implementing the Sense-Making approach to the interview with an interviewing instrument specifically adapted for these purposes. Data collection involved random sampling with the use of open-ended qualitative phone interviews. The specific set of analytic tools applied were quantitative -- content analyses and statistical tests of relationships.

Making Sense of the News: Study Hypotheses

The study reported here specifically set out to reconsider the relationship between viewers' interpretations and broadcast news reports. 12 The study retained Morley's premise that news broadcasts may be interpreted differently by different groups of viewers. It tested once again Morley's question of social class as a predictor of decoding, and it suggested an alternative predictor that was guided in its development by Sense-Making's conjoint emphasis on the situational and the phenomenological. The study retained Morley's definitional distinction between preferred readings (those which favor the dominant social order) and alternative readings (those that are critical of the political and social status quo).

Primary Hypotheses

This study incorporated two predictor variables and one criterion. In line with Morley, social class was one predictor, assumed to relate to audience decoding of the news. This study incorporated a second alternative predictor also assumed to relate to audience decoding, this one drawn from a Sense-Making conceptualization. These two predictors were analyzed in terms of their statistical capacities to relate significantly to a criterion measure tapping audience decoding, which for this study was called social criticism decoding. Specific operationalizations are presented in the next section of this article. The intent of this section is to present the study's guiding hypotheses.

As a complement to Morley's use of social class as a determinant of audience decoding, this study used its alternative indicator -- societal involvement state -- to account for individuals' varied situational perspectives as they viewed television news. Societal involvement state was defined operationally as one's orientation to social-political consciousness in society and, in particular, whether the viewer described self (or people cared about) as a victim or not. The measure was intended to be at one and the same time phenomenological (focusing on agency) and structural (focusing on constraints), thus positioning the respondent as an agent vis-a-vis structure. Comparison of social class with societal involvement state as a predictor of social criticism decoding was structured in terms of three major hypotheses. The first expressed Morley's perspective:

H1: For class-related stories (politics and economics, crime, labor/management disputes), social class will be significantly related to social criticism decoding.
In contrast, the Sense-Making perspective necessitates the use of measures that attend to movement across time and space as well as to the material conditions of respondents. From this point-of-view, then, a situational/phenomenological variable is assumed to be more likely than a demographic measure to capture the uniqueness of the moment and the balance of contradictory forces experienced by the viewer. This reasoning led to these additional hypotheses drawn from Sense-Making's meta-theoretic assumptions:
H2: Societal involvement state and social criticism decoding will be significantly related.

H3: Societal involvement state will show a stronger relationship with social criticism decoding than social class.

Control Variables

To subject the three hypotheses above to the sternest possible test, this study used statistical tools to determine whether a set of four alternative predictors would, when introduced as control variables, change the statistical relationship between the primary predictors and social criticism decoding. The selected set of four control variables all represent strong research traditions verifying their roles as significant predictors in a variety of ways of audience-media relationships. The selected control variables were: 1) cognitive intensity; 2) credibility of news coverage; 3) frequency of news viewing; and 4) education.

1) Cognitive Intensity. The idea that there is a relationship between how intensely a person thinks or feels about something and how he or she processes messages has a long tradition in the social sciences and communication in particular. The variable of involvement has been used fairly extensively in psychologically based communication research to measure, in one way or another, the strength of a relationship between a message and the individual. For Sherif and Hovland (1961), authors of the first major work in the area of Social Judgment Theory, "the greater the individual's ego-involvement, the more resistant the individual will be to a change of attitude which is critical to the individual's identity and ego." A similar concept was used by Petty and Cacioppo (1986) in Elaboration Likelihood Theory, which suggests that the more personal relevance an argument has to an individual, the more thoughtfully involved he or she is likely to be in considering its merits. 13

2) Credibility of News Coverage. Another extensively used variable in communication has been media credibility, because trust in the news media has been shown to be a factor that sometimes makes a difference in audience use and interpretation of mass media. There have been numerous studies that have examined media credibility, especially in light of recent public opinion polls indicating that public confidence in the media is declining. 14 The notion of media credibility or believability also has been foundational to a branch of uses-and-gratifications theory developed by Palmgreen (1984) and others, which is based on expectancy value theory. According to this line of research, there is a relationship between one's evaluations of and beliefs about a given segment of media content and one's expectations of the gratifications it will provide. Similarly, Wanta and Hu (1994) provided support for the argument that if the news media are perceived to be credible, then people will rely on them for information, which will increase their exposure to the media and in turn make them more susceptible to such effects as agenda-setting.

3) Frequency of News Viewing. The amount of one's regular media usage is another variable that has often been employed in mass media research. 15 This underlying assumption is derived in part from the uses-and-gratifications principle that people differ in their media use behaviors, i.e., in what they want from and how they use the media. Research in this area has indicated that regular news users may decode differently from infrequent users. Similarly, a cultivation analysis approach to media research has documented that regular news viewers may be more likely to interpret the news consistent with its "preferred readings" than infrequent viewers. 16

4) Education. The final variable chosen for use here is education. Viewer education was considered to merit analytic control given the known relationship between education and social class and the lessons of past research that audience interpretations may vary based on a receiver's cultural knowledge and capabilities. 17

Two final hypotheses were framed to structure the analytic control for cognitive intensity, credibility of news coverage, frequency of news viewing, and education:
H4: The hypothesized significant relationship of societal involvement state with social criticism decoding will be maintained despite controls for cognitive intensity, credibility of news coverage, frequency of news viewing, education, or all control variables taken together.

H5: The superiority of society involvement state over social class for predicting social criticism decoding will not be significantly weakened by control for cognitive intensity, credibility of coverage, frequency of news viewing, education, or all control variables taken together.



The questionnaire used Dervin's (1983) Abbreviated Micro-Moment Time-Line interviewing approach with telephone interviews ranging from 10 to 20 minutes in length. The interviewing structure was one where the respondent was asked a series of questions that progressively narrowed the focus of inquiry to reception behaviors related to the research interests here.

To begin, viewers were asked to recall a recent encounter with a television news program. Selecting one news item from that program, based on importance to each respondent at the time, viewers were requested to list the different elements seen as contained in the story chosen and then to identify whatever thoughts or questions came to mind while viewing each element named. This pool of viewers' thoughts and questions became the data source for the criterion variable social criticism decoding.

The interviewees were asked to specify the one thought considered most important to self at the time of viewing, and these single most important thoughts were conceptualized as the unit of analysis. The unit then is not reduced to a person but rather to a person-in-a-material- situation-with-an-interpretation. Respondents were then asked to rate the level of importance of their most important thought on a five-point scale. This provided a measure of cognitive intensity.

Then each viewer was asked to focus on his or her life situations (past and present) and how that one most important thought during the news story might have been prompted by life events. This generated a pool of open-ended responses that became the data source for the situational/phenomenological variable societal involvement state.

A data source for determining social class was developed by eliciting the occupation of the respondent, occupation of the main income-producer in the respondent's household, occupation of the main income-producer while respondent was a child, and current household income. Credibility of coverage was measured by asking respondents to rate the accuracy of their chosen news story on a five-point scale. Similar self-reports were utilized for the remaining control variables, frequency of news viewing and education. Specific operationalizations of variables are presented below.

Sampling and Fielding

Telephone interviews were conducted using a systematic random sample drawn from the Seattle telephone directory. Out of 227 participants (each 18 years or older) who were interviewed, 69% (n=158) reported on a news story that involved political and social issues. The analyses were confined to these 158 respondents. Table 1 compares key demographic characteristics of these 158 people with those of the entire population of the city studied. 18


Table 1. Demographic comparison of respondents with general population of Seattle.



The data sources for the two predictors (social class and societal involvement state) and one criterion (social criticism decoding) were open-ended responses elicited as described above. These were quantified using content analytic procedures. The senior author served as primary coder. A randomly selected 20 percent sub-sample was content-analyzed again by one of the original interviewers. Intercoder judgment reliability was evaluated using Stempel's (1952) percentage agreement index and Scott's (1955) pi, which corrects for expected agreements by chance. Results showed that interjudge coding reliabilities were all above 90% using the more lenient Stemple's agreement index and from 80-86% using the conservative Scott's index. The specific coding operations are described below. Table 2 presents the results of the content analytic procedures in terms of the percentages of respondents coded into each of the categories for each of the three primary variables.


Table 2. Percentages of respondents coded into each of the categories of the three primary study variables -- social criticism decoding, social class, and societal involvement state.


Social Criticism Decoding. As discussed above, viewers' news decodings were compared with one another rather than against the external standard of a societal observer's assessments of the "preferred readings" of the news stories. Thus the values for social criticism decoding were based entirely upon the content of thoughts and questions reported by viewers to have occurred while viewing a chosen news report. For the research reported here, these were coded into two categories: a) societal acceptance where respondent did not criticize or challenge general social relationships and patterns; and b) societal negotiation where respondent criticized or challenged general social relationships and patterns. As indicated in Table 2, 37.3% of participants were coded as accepting society, while 62.7% did not. Figure 1 presents examples of verbatim responses from respondents that were coded as accepting versus negotiating society.


Figure 1. Examples of open-ended responses as coded for social criticism decoding -- accept society and negotiate society.

Social Class. Class status was based on the occupation of the main income-producer of the respondent's household. Where that person was a student, respondent's class was coded the same as his or her parents' social class. A simple dichotomous coding was used which assessed each respondent as labor class or not-labor class. Of course, much finer distinctions can and ought to be made, but this would have required a larger sample than resources allowed. Table 2 shows the coding results: 58.9% of the 158 respondents were coded as labor and 41.9% as not-labor.

Societal Involvement State. What was required here was a nominal measure for viewers' involvement with television news stories, incorporating the emphasis of British cultural studies on political and social power and domination. It was decided to define this situational predictor in terms of the political or social consciousness of self in society. The data source for the variable came from respondent descriptions of the factors in their lives that they saw as leading them to their most important thoughts at the time of viewing a news story. Responses were coded into two categories: a) societal victims, those who described themselves (or those to whom they felt closest) as generally injured or powerless in terms of the issues at hand; and b) not societal victims, those who did not describe themselves (or those to whom they felt closest) as generally injured or powerless in terms of the issues at hand. An example of a response coded as societal victim came from a woman discussing sexism who expressed concern about the lack of women executives. An example of a response coded as not societal victim came from a woman discussing childcare who described her backup plans if the primary care-giver was not available.

Figure 2 presents examples of viewers' comments and their codings for societal involvement state. As indicated in Table 2 above, 47.5% identified as societal victims while 52.5% did not.


Figure 2. Examples of open-ended responses as coded for societal involvement state -- societal victims and not societal victims.


Control Variables

As discussed above in the section on hypotheses, four control variables were selected -- cognitive intensity, credibility of news coverage, frequency of news viewing, and education -- each of which was derived from a different tradition of research relating to media audiences. Operationalizations were as follows:

Cognitive Intensity. Cognitive arousal in the media encounter was measured by asking viewers to rate on a five-point scale the importance to them of their most important thought at the time of viewing. As Table 3 shows, 67.7% of respondents rated that thought as "highly important" or "extremely important."

Credibility of News Coverage. To measure each viewer's level of confidence in the news coverage during the viewing situation reported, respondents were asked to rate the accuracy of their reported news stories on a five-point scale of increasing magnitude. The mean score for credibility of coverage was high at 3.99, with 58.9% of the 158 respondents judging their focal news story as very or extremely credible.

Frequency of News Viewing. Respondents were asked how often, on a five-point scale, they exercised the opportunity to watch television news. As shown in Table 3, frequency of news viewing was high, with 61.5% watching "every day or almost every day."

Education. The median number of years on the self-reported education variable was just under 15 years. Although the full range of responses was used in the covariate analysis, for ease of presentation Table 3 groups education into four categories: 3.8% of respondents reported education of 8 years or less; 21.5% reported 9-12 years; 49.5% reported 13-16 years; and 25.3% reported 16+ years.

Table 3 presents the results of these operationalizations in terms of the percentages of the 158 respondents in each measurement category of each control variable.


Table 3. Percentages of respondents coded into each of the categories of the four control variables -- cognitive intensity, credibility of news coverage, frequency of news viewing, and education.


Statistical Analyses

The statistical analyses involved the use of different forms of analysis of variance. Hypotheses 1, 2, and 3 were tested by employing social class and societal involvement state as alternative predictors of social criticism decoding in a factorial analysis of variance. Specific procedures used made no assumptions about causal ordering, but prioritized main effects over interaction effects. Examination of within-table cell mean differences was performed using Duncan's comparison for paired means. The control variable hypotheses were executed with a series of analyses of covariance in which the question was whether the covariates could account for enough variance in social criticism to diminish any obtained significances resulting from Hypotheses 1, 2 and 3.


Table 4 presents the results of the factorial analysis of variance that addressed Hypotheses 1, 2, and 3. Examining sources of variance revealed only one significant main effect, for the predictor societal involvement state (p < .001). This finding supported Hypotheses 2 and 3. Absent a significant relationship between social criticism decoding and social class, Hypothesis 1 (for Morley's position) was unsupported. The interaction was not significant.


Table 4. Factorial analysis of variance, with social criticism decoding predicted by societal involvement state and social class.


Results showed, thus, that even though 41.9% of the 93 labor respondents exhibited social criticism decodings (i.e., decodings that negotiated rather than accepted society) compared to 30.8% of the not-labor respondents, this difference was not significant. In contrast, 61.3% of the 75 respondents coded as exhibiting a consciousness focusing on societal victims exhibited negotiated decoding compared to only 15.7% of those whose societal involvement state was coded as not-victim.

An examination of the within-cell means in Table 4 shows that the main effect results maintained themselves. Thus, for labor respondents, 58.5% of those whose societal involvement was coded as victim exhibited social criticism decoding, compared to only 20.0% of those coded as not-victim. The comparable figures for the not-labor respondents were 68.2% and 11.6%, respectively.

While societal involvement as a victim versus not-victim was the over-ridingly significant comparison in this analysis, the size of this difference was greater for those respondents not coded as in the labor class than for those coded in the labor class. For the not-labor respondents this differential was 56%, the difference between the 68.2% of not-labor victims who exhibited social criticism decoding and the 11.6% of not-labor not-victims. For the labor respondents this difference dropped to 38%, the difference between the 58.5% of labor victims who exhibited social criticism decoding and the 20.0% of labor not-victims. This comparison shows most clearly in Figure 3, which plots the cell means from Table 4. Here we see an interaction which, although not statistically significant, bears attention. In contradiction to the expectation that the laboring class should exhibit the most negotiated decoding, in this data there was a tendency for the most social criticism decoding to come not from labor but from managerial and professional groups and others coded as not-labor.


Figure 3. Plot of the cell means comparing the levels of social criticism decoding of respondents classified as not-labor and labor victims and not-labor and labor not-victims.


Table 5 presents the results of the analyses of covariance, which tested the relationships obtained while controlling for cognitive intensity, credibility of news coverage, frequency of news viewing, and education, both individually and together.


Table 5. Summary of statistical findings for social criticism decoding as predicted by social class and societal involvement state controlled by cognitive intensity, credibility of news coverage, frequency of news viewing, and education.


Results showed that:

1) Cognitive Intensity. Control for cognitive intensity made no significant difference, with no shift in the marginal means and only a slight alteration in the multiple regression coefficient.

2) Credibility of News Coverage. Control for credibility of news coverage was itself significantly related to social criticism decoding (p < .05), and control for credibility of coverage helped the interaction between social class and societal involvement state to achieve near statistical significance (p < .10). Control for credibility of coverage altered the social criticism rate for each social class so as to bring the rate for not-labor to a not significantly different higher value than that for labor rather than a not significantly different lower value. This covariate also reduced the difference in rates of social criticism between societal victims and non-victims, although not enough to undercut the level of significance for societal involvement state as a predictor of social criticism decoding.

3) Frequency of News Viewing. This variable was, by itself, a significant predictor of social criticism decoding (p < .05). But control for this covariate had no measurable effect on the level of significance for the main effects previously discussed.

4) Education. Control for education made no difference in the significance of the findings from the baseline factorial analysis.

Finally, the four covariates taken together were not, in themselves, significant predictors of social criticism decoding. Simultaneous control for all four covariates failed to undercut the level of significance for societal involvement state as a predictor of social criticism decoding, although it did bring marginal means for victims and non-victims slightly closer together (57% to 18% compared with 61% to 16%, reported in Table 4).

Control for all covariates at once brought the social criticism rates for labor and not-labor respondents to an equivalent 37%. Simultaneous control for all covariates brought statistical significance to the interaction between social class and societal involvement state. But as none of the control variables in the covariate analysis altered the level of significance for the main effects obtained in the original analysis of variance, Hypotheses 4 and 5 were both supported.

Summary, Conclusions, and Implications

This study set out to demonstrate the utility of the Sense-Making Methodology for researching people's differential interpretations of mass media and specifically for interrogating questions left unanswered and/or contested as a result of the Morley (1980) audience study of a BBC television news program. The Morley study expected to find that the laboring classes would exhibit more criticism of society in their decodings of news stories than the not-laboring classes. Results did not support the expectation. This study builds on Morley's essential interests by pitting two different predictors of social criticism decoding against each other. The first of these was social class, as mandated by Morley. The second predictor was derived from the meta-theoretic premises of the Sense-Making Methodology, namely that the imposing of societal markers on interpretive actors from the outside is too far removed from material lived conditions and struggles to tap the essence of what would drive the audience member to social criticism decoding. Instead, this study used an alternative predictor tapping the consciousness of the news viewer regarding societal victims. The core hypothesis was that those oriented toward a consciousness of societal victimization would be more likely to exhibit social criticism decoding. Results showed, in fact, that it was consciousness of societal victimization -- what was called societal involvement state -- which predicted social criticism decoding and continued to do so despite a series of attempts to diminish the relationship with statistical controls.

This study, then, had one major conclusion. Social criticism decoders for a given news story were more likely to be members of the audience who saw the relationship between their own lives and the content of a news study as representing in some way the manner in which society victimizes them or their loved ones.

This study purposively used the analytics of quantification as an approach to analysis despite the fact that the work of Morley and his colleagues has been qualitative and among some which have been resistant to quantification. The present study has taken a different stance -- that in the long-term step-takings of a research program there is a role both for studies traditionally labeled qualitative and those traditionally labeled quantitative. One of the more interesting results of this study is the way in which a series of statistical controls provided some subtle cues for future research questions.

Thus, on the one hand we get a strong mandate from these data to look at what we have called situational/phenomenological measures of audience orientations toward news stories. In the present study there is little doubt that this was where the important activity emerged. A close examination suggested some avenues to pursue. For one, these results showed an interaction between social class and societal involvement state that became significant after all statistical controls. What was notable about this was that it was counter to the Morley hypotheses. Thus, among those audience members with a consciousness of victimization it was not the laborers who exhibited more social criticism decoding but the not-laborers (managers, professionals, etc.). In contrast, among those without a consciousness of victimization the opposite trend emerged -- the laborers, in line with Morley's expectation, exhibited slightly more negotiated decoding.

These results point to fruitful avenues to pursue. For example, might it be concluded that under conditions where there is a disjuncture of some kind between one's phenomenological worlds and one's material circumstances one engages in more cognitive work and this is what leads the not-laborers with a victimization consciousness to exhibit somewhat more social criticism decoding. 19 The fact that the statistical findings showed this anomalous finding became stronger with statistical controls is a further suggestion of a potentially useful avenue of exploration.

Of course, the findings open the door to a host of questions of the conditions that lead members of the audience to social criticism decodings. Morley (1993) emphasized his continued belief in the face of contradictory implications that subordinated meanings (i.e., decodings that accept societal conditions as they are) will be more widespread than interpretive resistance. In fact, in the present study this was confirmed -- 62.7% of these 158 respondents did not criticize or challenge societal relationships in their decodings of a recent news story that they selected as important to them. This makes the presence of the 37.3% of respondents who did engage in societal criticism even more interesting, particularly when we see that that figure rose to 61.3% for those with consciousness of societal victimization and to 68.2% among the not-laborer sub-group with this consciousness. The very fact that these social criticisms are being made against considerable hegemonic forces makes their presence in these data especially noteworthy.

There was much that could have been done with the data from this study to unravel some of these mysteries -- qualitative comparisons, for example, of the decodings of respondents with differing consciousness of societal victimization relating to one particular news story; or, comparisons of the decodings of laborers or not-laborers that differed radically from each other. 20 Our purpose, here, however, has been to present one highly constrained but explicit test in as powerful a way as possible.

We started by assuming that Sense-Making Methodology has within it some meta-theoretic assumptions and related methods that could usefully address some of the conundrums that have faced audience reception analysis. Our conclusions support this expectation in two ways. One, the data presented here found high levels of preferred decodings (i.e., decodings that accept societal structures) where others have had difficulty. Two, the data isolated a significant predictor of social criticism decoding where others have had difficulty.

We do not offer Sense-Making as a panacea to reception analysis. But we do point out that there are some deliberately theorized aspects to the approach that are of methodological merit and which we believe made achieving these results possible. We review them here briefly in terms of how they were implemented in this study.

1. In this study, respondents named news stories that were important to them. It was these stories that were used as the focus for assessments of the presence of social criticism decoding, not stories imposed on respondents that might or might not have had any relevance to them.

2. The interviewing format imposed no nouns of substance on respondents but rather focused entirely on asking them how they made and unmade sense as they moved through time-space vis-a-vis particular news stories and the relationships of these stories to their lives.

3. Respondents were invited to be social theorists -- to describe connections between news stories and their lives. They were additionally invited with Sense-Making's imposed disciplined listening by interviewers to speak their stories on their own terms without having to navigate an imposed discourse.

4. Respondents were freed from being placed in social marker boxes (e.g., social class and other demographic characteristics) with no room for expression of their material lives that rarely fit so tidily. In particular, respondents were conceptualized as embodied in material conditions and moving through an evolving, albeit elusive, time-space.

5. This study specifically applied from Sense-Making an analytical interest in interrogating how respondents are characterized and the conditions under which, for example, across-time-space characterizations (such as social class) have an impact on audience interpretations and the conditions under which it is more time-space bound characterizations that do.

6. Finally, this study entered the middle terrain by proposing a characterization of audience members, which was at one and the same time structural and phenomenological and, thus, attended to the time-space where agency mediates structure and vice-versa.


1.  Testimony to the continued importance of the Nationwide study can be seen in the references to it in such recent literature as Dickerson (1996), Hoijer (in press), Hunt (1997), Lewis (1997), Shields (1996), and Storey (1996).

2.  It is worth noting here that while the first Nationwide study was received within the context of an ongoing debate over the significance of textual analysis, publication of the second study was greeted with "deafening silence" (Morley & Brunsdon, 1999) because of its novel approach. Only with the subsequent publication of work by Hobson, Ang, Radway, and others did the area of audience reception begin to draw the attention of media scholars.

3.  Stuart Hall's (1980) encoding/decoding model was initially presented in a 1973 paper. The model conceptualized the process of communication in terms of a complex structure of relations produced and sustained through the articulation of distinctive moments: production, circulation, distribution/consumption, and reproduction. Informed by Marx's notion of commodity production, Hall suggested that it would be useful to consider how the continuous circuit is maintained as well as the forms in which the product of process appears at each moment.

4.  Recent articles that focus on current issues in cultural and media studies include Barker & Brooks (1998), Kellner (1997, 1995), and Murdock (1997), and Osterud (in press).

5.  For examples of studies that take an active audience approach, see Tufte (in press), Fingerson (1999), Fiske (1986, 1987), and Sampedro (1998).

6.  Morley and his colleagues (see, e.g., Morley & Silverstone, 1990; Silverstone, 1998) themselves note the importance of television's "double articulation": the distinction between television viewing as ritualistic activity and as an active mode of consumption and the production of meaning.

7.  Others who have discussed the tensions between micro- and macro-level analyses include Drotner (in press), Kellner (1997, 1995), Maxwell (in press), and Murdock (in press).

8.  Among those who have commented on the relatively few empirical studies are Kellner (1995), Lewis (1997), Murdock (1997), and Morrow (1994).

9.  For a comprehensive overview of the Sense-Making Methodology, see Dervin (in press).

10.  Sense-Making has been used in studies focusing on information seeking and use, health communication, public communication campaigns, audience research, alternative media, development communication, telecommunications policy, and media pedagogy, to name a few. For overviews see Dervin (1992, 1989). The most recent empirical study is Dervin and Shields (1999). For a list of those who have used Sense-Making, other than as represented in this issue, contact

11.  Sense-Making has documented extraordinary differences between the typical questioning methods that position respondents as reactors and questioning in which respondents' perceptions are freely submitted.

12.  This study is drawn from the senior author's unpublished doctoral dissertation (Dworkin, 1987). Data were collected in 1985. While some 15 years have intervened, the research questions raised are even more pertinent today and the results speak to still unresolved issues. While the news stories of that period differed, an examination of respondents' narratives about these news stories sound as fresh today as then. There is no reason to expect that the process of audience interpretings has changed in ways that would impact the results reported here.

13.  Recent examples of such cognitive research can be seen in Bauserman (1998), Dinoff and Kowalski (1999),
Duck, Terry, & Hogg (1998), Frewer et al. (1999), Kim and Rubin (1997), Kirby et al. (1998), and Slater (1997).

14.  Examples of recent media credibility studies include Johnson and Kaye (1998), Major and Atwood (1997), Austin and Dong (1994), Almakaty, Boyd, & Vantubergen (1994), Stamm and Dube (1994), and Johnson (1993).

15.  Studies that have considered frequency of viewing include Finn (1997), Roe & Vandebosch (1996), and Vincent & Basil (1997).

16.  For more on cultivation analysis, see Gerbner (1998), Gerbner et al. (1994), and Signorielli & Morgan (1990).

17.  Among recent studies that have included measures of education are Dupagne (1999), Gaziano (1997), Long & Traudt (1999), and Rothenbuhler et al. (1996). Education, as it is translated into symbolic and cultural capital, also is very important in the work of Bourdieu (1993, 1984).

18.  Table 1 demographic comparison of sample and population relied on 1980 census figures for the city of Seattle.

19.  Recent work applying Sense-Making to policy issues regarding the "missing user" in policy discourse specifically attends to the issue of how to conceptualize users in work focusing on prediction and ultimately in policy that makes assumptions regarding users. See Dervin & Shields (1999) and Shields, Dervin, Richter, & Soller (1993).

20.  Several possibilities come to mind, for example, why not-labor respondents who saw themselves as societal victims were even more likely to exhibit social criticism decoding than their labor counterparts. Perhaps they were angry and resented being victims, having concluded from their own experience that the prescribed social levelers of education and profession still cannot completely overcome the imposed constraints of a class/race/gender-structured society. Alternatively, perhaps in the course of their education or upbringing or professional life they acquired more tools for social analysis and were encouraged to challenge the received wisdom presented in the dominant-hegemonic interpretation of news. Or maybe they participate in social circles and arenas of discourse (such as the university) where social criticism and thinking for one's self are more highly valued. We should mention again that the number of participants in the study was relatively low, so it would take just a few critical intellectuals to push the results in the critical direction.


Almakaty, S. S., Boyd, D., & Vantubergen, G. N. (1994). Source credibility during the Gulf War: A Q-study of rural and urban Saudi-Arabian citizens. Journalism Quarterly, 71(1), 90-98.

Austin, E. W., & Dong, Q. W. (1994). Source versus content effects on judgments of news believability. Journalism Quarterly, 71(4), 973-983.

Barker, M. & Brooks, K. (1988). On looking into Bourdieu's black box. In R. Dickinson, R. Harinaranath, & O. Linne (Eds.), Approaches to audiences: A reader (pp. 218-232). London: Arnold.

Bauserman, R. (1998). Egalitarian, sexist, and aggressive sexual material: attitude effects and viewer responses. Journal of Sex Research, 35(3), 244-253.

Bourdieu, P. (1993). The field of cultural production: Essays on art and literature. New York: Columbia University Press.

Bourdieu, P. (1984). Distinction: A social critique of the judgement of taste.  Cambridge: Harvard University Press.

Carter, R. F. (1972). Application of signaled stopping technique to communication research. In P. Clark (Ed), New models for communication research . Beverly Hills, CA: Sage.

Cameron, G. (1993). Spreading activation and involvement: An experimental test of a cognitive model of involvement. Journalism Quarterly, 70(4), 854-867.

Dervin, B. (in press). On studying information needs and seeking methodologically: The implications of connecting metatheory and method. Information Processing and Management.

Dervin, B. (1993). Verbing communication: A mandate for disciplinary invention. Journal of Communication, 43(3), 45-54.

Dervin, B. (1992). From the mind's eye of the user:  The sense-making qualitative-quantitative methodology. In J. D. Glazier, & R. R. Powell (Eds.), Qualitative research in information management (pp. 61-84). Englewood, CO: Libraries Unlimited.

Dervin, B. (1989). Audience as listener and learner, teacher and confidante:  The sense-making approach  . In R. Rice, & C. Atkin (Eds.), Public communication campaigns (2nd ed., pp. 67-86). Newbury Park, CA: Sage.

Dervin, B. (1983). An overview of sense-making research: concepts, methods, and results. Presented at International Communication Association meeting.

Dervin, B., & Dewdney, P. (1986). Neutral questioning: A new approach to the reference interview. RQ, 25(4), 503-513.

Dervin, B., & Shields, P. (1999). Adding the missing user to policy discourse: Understanding US user telephone privacy concerns. Telecommunications Policy, 23, 403-435.

Dickerson, P. (1996). Let me tell us who I am. European Journal of Communication, 11(1), 57-82.

Dinoff, B. L., & Kowalski, R. M. (1999). Reducing AIDS risk behavior: The combined efficacy of protection motivation theory and the elaboration likelihood model. Journal of Social and Clinical Psychology, 18(2), 223-239.

Drotner, K. (In press). Less is more: Media ethnography and its limits. In I. Hagen, & J. Wasko (Eds.), Consuming audiences? Production and reception in media research . Cresskill, NJ: Hampton Press.

Duck, J. M., Terry, D. J., & Hogg, M. A. (1998). Perceptions of a media campaign: The role of social identity and the changing intergroup context. Personality and Social Psychology Bulletin, 24(1), 3-16.

Dupagne, M. (1999). Exploring the characteristics of potential high-definition television adopters. Journal of Media Economics, 12(1), 35-50.

Dworkin, M. A. (1987). Making sense with television news:  Situations, context, and psychology of the audience experience. Unpublished doctoral dissertation, University of Washington, Seattle.

Fingerson, L. (1999). Active viewing: Girls' interpretations of family television programs. Journal of Contemporary Ethnography, 28(4), 389-418.

Finn, S. (1997). Origins of media exposure: Linking personality traits to TV, Radio, Print, and Film Use. Communication Research, 24(5), 507-529.

Fiske, J. (1998). Audiencing: Cultural practice and cultural studies. In N. K. Denzin, & Y. S. Lincoln (Eds.), The landscape of qualitative research: theories and issues (pp. 359-378). Thousand Oaks, CA: Sage.

Fiske, J. (1987). Television culture. London: Routledge.

Fiske, J. (1986). Television: Polysemy and popularity. Critical Studies in Mass Communication, 3(2), 2.

Freire, P. (1970). Pedagogy of the oppressed. New York: Herder and Herder.

Frewer, L. J., Howard, C., Hedderley, D., & Shepherd, R. (1999). Reactions to information about genetic engineering: impact of source characteristics, perceived personal relevance, and persuasiveness. Public Understanding of Science, 8(1), 35-50.

Gaziano, C. (1997). Forecast 2000: Widening knowledge gaps. Journalism and Mass Communication Quarterly, 74(2), 237-264.

Gerbner, G. (1998). Cultivation analysis: An overview. Mass Communication & Society, 1(3/4), 175-194.

Gerbner, G., Gross, L., Morgan, M., & Signorielli, N. (1994). Growing up with television: The cultivation perspective. In J. Bryant, & D. Zillmann (Eds.), Media effects: Advances in theory and research (pp. 17-41). Hillsdale, NJ: Lawrence Erlbaum.

Hagen, I. (In Press). Modern dilemmas: TV audiences' time use and moral evaluation. In I. Hagen, & J. Wasko (Eds.), Consuming audiences? Production and reception in media research. Cresskill, NJ: Hampton Press.

Hall, S. (1980). Encoding/decoding. Culture, media, language: Working papers in cultural studies, 1972-79 (pp. 128-138). London: Hutchinson.

Hall, S. (1977). Culture, media, and the ideological effect. J. Curran, M. Gurevitch, & J. Woollacott (Eds.), Mass Communication and Society  (pp. 315-348). London: Arnold.

Hoijer, B. (In Press). Audiences' expectations on and interpretations of different television genres: A socio-cognitive approach. In I. Hagen, & J. Wasko (Eds.), Consuming Audiences? Production and Reception in Media Research. Cresskill, NJ: Hampton Press.

Hunt, D. (1997). Screening the Los Angeles "riots". Cambridge: Cambridge University Press.

Johnson, T. J. (1993). Exploring media credibility: How media and non-media workers judged media performance in Iran/Contra. Journalism Quarterly, 70(1), 87-97.

Johnson, T. J., & Kaye, B. K. (1998). Cruising is believing? Comparing internet and traditional sources on media credibility measures. Journalism & Mass Communication Quarterly, 75(2), 325-340.

Kellner, D. (1997). Critical theory and cultural studies: The missed articulation. In J. McGuigan (Ed.), Critical methodologies (pp. 12-41). London: Sage.

Kellner, D. (1995). Media communications vs. cultural studies: Overcoming the divide. Communication Theory, 5(2), 162-177.

Kim, J., & Rubin, A. M. (1997). The variable influence of audience activity on media effects. Communication Research, 24(2), 107-135.

Kirby, S. D., Ureda, J. R., Rose, R. L., & Hussey, J. (1998). Peripheral cues and involvement level: Influences on acceptance of a mammography message. Journal of Health Communication, 3(2), 119-135.

Lewis, J. (1997). What counts in cultural studies. Media, Culture & Society, 19, 83-97.

Long, J. F., & Traudt, P. J. (1999). Discriminating audience donor factors for public radio: A tale of two cities. Journal of Media Economics, 12(1), 51-66.

Major, A. M., & Atwood, L. E. (1997). Changes in media credibility when a predicted disaster doesn't happen. Journalism & Mass Communication Quarterly, 74(4), 797-813.

Maxwell, R. (In Press). Surveillance and other consuming encounters in the informational marketplace. In I. Hagen, & J. Wasko (Eds.), Consuming audiences? Production and reception in media research. Cresskill, NJ: Hampton Press.

Morley, D. (1993). Active audience theory: Pendulums and pitfalls. Journal of Communication, 43(3), 13-18.

Morley, D. (1992). Television, Audiences, and Cultural Studies. London: Routledge.

Morley, D. (1981). The Nationwide audience: A critical postscript. Screen Education, 39.

Morley, D. (1980). The "Nationwide" audience: Structure and decoding. London: British Film Institute.

Morley, D., & Brunsdon, C. (1999). The `Nationwide' television studies. London and New York: Routledge.

Morley, D., & Silverstone, R. (1990). Domestic communication -- technologies and meanings. Media, Culture & Society, 12, 31-55.

Morrow, R. A. (1994). Critical theory and methodology. Thousand Oaks, CA.: Sage Publications.

Murdock, G. (In Press). Peculiar commodities: Audiences at large in the world of goods. In I. Hagen, & J. Wasko (Eds.), Consuming audiences? Production and reception in media research. Cresskill, NJ: Hampton Press.

Murdock, G. (1997). Thin descriptions: Questions of method in cultural analysis. In J. McGuigan (Ed.), Cultural methodologies (pp. 178-192). London: Sage.

Osterud, S. (In Press). How can audience research overcome the divide between macro- and microanalysis, between social structure and action? In I. Hagen, & J. Wasko (Eds.), Consuming audiences? Production and reception in media research. Cresskill, NJ: Hampton Press.

Palmgreen, P. (1984). Uses and gratifications: A theoretical perspective. In R.N. Bostrom (Ed.), Communication Yearbook, 8, 20-55. Beverly Hills, CA: Sage.

Parkin, F. (1973). Class inequality and political order. London: Paladin.

Perse, E. M. (1990). Involvement with local television news. Human Communication Research, 16(4).

Petty, R. E., & Cacioppo, J. T. (1986). Communication and persuasion: Central and peripheral routes to attitude change. New York: Springer-Verlag.

Roe, K., & Vandebosch, H. (1996). Weather to view or not: That is the question. European Journal of Communication, 11(2), 201-216.

Rothenbuhler, E. W., Mullen, L. J., DeLaurell, R., & Ryu, C. R. (1996). Communication, community attachment, and involvement. Journalism & Mass Communication Quarterly, 73(2), 445-466.

Sampedro, V. (1998). Grounding the displaced: Local media reception in a transnational context. Journal of Communication, 48(2), 125-143.

Scott, W.A. (1955). Reliability in content analysis: The case of nominal scale coding. Public Opinion Quarterly, 19, 321-325.

Schwarz, N. (1998). Warmer and more social: recent developments in social psychology. Annual Review of Sociology, 24, 239-264.

Sherif, M., & Hovland, C. I. (1961). Social Judgment. New Haven, CT: Yale University Press.

Shields, V. R. (1996). Selling the sex that sells: Mapping the evolution of gender advertising research across three decades. In B. Burleson (Ed.) Communication Yearbook, 20, pp. 70-109. Thousand Oaks, CA: Sage.

Shields, P., Dervin, B., Richter, C., & Soller, R. E. (1993). Who needs "POTS-plus" services?  A comparison of residential user needs along the rural-urban continuum.  Telecommunications Policy, 17(8), 563-587.

Signorielli, N., & Morgan, M. (Eds.). (1990). Cultivation analysis: New directions in media effects.  Newbury Park, CA: Sage.

Silverstone, R. (1998). Television and everyday life: towards an anthropology of the television audience. In R. Dickinson, R. Harindranath, & O. Linne (Eds.), Approaches to audiences: A reader. London: Arnold.

Slater, M. D. (1997). Persuasion processes across receiver goals and message genres. Communication Theory, 7(2), 125-148.

Stamm. K., & Dube, R. (1994). The relationship of attitudinal components to trust in media. Communication Research, 21(1), 105-123.

Stempel, G. H. (1952). Increasing reliability in content analysis. Journalism Quarterly, 19, 449-455.

Storey, J. (1996). Cultural studies and the study of popular culture: theories and methods. Athens: University of Georgia Press.

Tufte, T. (In Press). The popular forms of hope -- television fiction, audience consumption and social change: a media ethnographical approach. In I. Hagens, & J. Wasko (Eds.), Consuming audiences? Production and reception in media research. Cresskill, NJ: Hampton Press.

Vincent, R., & Basil, M. (1997). College students' news gratifications, media use, and current events knowledge. Journal of Broadcasting and Electronic Media, 41(3), 380-392.

Wanta, W., & Hu, Y. W. (1994). The effects of credibility, reliance, and exposure on media agenda-setting: A path analysis model. Journalism Quarterly, 71(1), 90-98.

Copyright 1999 Communication Institute for Online Scholarship, Inc.

This file may not be publicly distributed or reproduced without written permission of the Communication Institute for Online Scholarship, P.O. Box 57, Rotterdam Jct., NY 12150 USA (phone: 518-887-2443).