Back to Journal Cover Page


Volume 3, Number 3
Submitted: January 5, 1998
Resubmitted: April 9, 1998
Accepted: April 28, 1998
Publication Date: May 15, 1998


Toby A. Ten Eyck
Louisiana State University


Studies of communication processes seem to focus either on interpersonal communications or mass-mediated communications. The dearth of studies which compare these two types of communication may stem from studies conducted in the 1940s and '50s which found that interpersonal communications had more influence on voting behaviors than mass-mediated communications. Whether or not generalizations from these studies have colored subsequent communications research, this area has been bypassed by most researchers. This study begins to fill this void by comparing and contrasting one type of mass-mediated communication (radio) to one type of interpersonal communication (a one-on-one telephone conversation) in terms of the third-person effect. Findings from this study show that individuals are just as likely to display a third-person effect towards interpersonal communication as mass-mediated communication.



There is currently an opportunity to extend our understandings of both human communication processing and the self in modern society, which seems to have been overlooked by researchers in the social psychological field. The opportunity consists of bringing together classic work on interpersonal vs. mass-mediated communication (Berelson, Lazarsfeld, and McPhee 1954; Katz 1957; Katz and Lazarsfeld 1955; Lazarsfeld, Berelson and Gaudet 1948; Merton 1949), contemporary research on the third-person effect (e.g., Cohen and Davis 1991; Davidson 1983; Fields and Schuman 1976; Gunther 1991), and theoretical insights on the self (e.g., Gergen 1991; Giddens 1991; Turner 1976). My purpose here is to challenge researchers to approach interpersonal and mass-mediated communications without a priori assumptions that these types of communication are processed differently. I begin my discussion by distinguishing between mass mediated and interpersonal communications.


Purcell (1997) has argued that continuing to dichotomize mass-mediated and interpersonal communication is archaic and downplays the dialectic relationship between them. Actually, one could argue that this was an insight made, in part, a number of years ago by proponents of the two-step flow of communication perspective (e.g., Katz 1957), who argued that mass-mediated messages were further disseminated and given credence by local opinion leaders. While I would agree that these types of communication are not polar opposites, I do not think we have reached a point where we can substitute one for the other. Still, distinguishing between them can be problematic. With the proliferation of electronic forms of interpersonal communication, such as telephones and computer networks, it is no longer useful to rely on distinctions based on medium. Audience size is also problematic, as conference calls and mass electronic mailings can be made at the interpersonal level, but audience size cannot be abandoned. Following Neuman (1991), I will argue that mass-mediated communication takes place when there are at least 500 receivers of the information. In addition, concerns of audience size must be accompanied by the nature of the relationship between the communicator and receiver involved in the communication process. The commodfication of mass-mediated messages and the expectations of mobilizing social capital can help further distinguish between these types of communication.


Thompson (1988:365) has argued that one difference between mass-mediated and interpersonal communication rests on the fact that mass-mediated "messages are incorporated into products which are sold [e.g., at the institutional level], or which are used to facilitate the sale of other goods [e.g., the consumer level]...." This translates into a specific purpose behind constructing these messages marketing a product. Interpersonal communication can rest on any number of purposes, including exchange value, and it is directed at specific people who are expected to act in specific ways. In other words, while the mobilization of social capital is expected to occur as a result of both types of communication, specific individuals are expected to act upon interpersonal communication. In addition mass-mediated communication is meant to mobilize economic capital, which is not a necessary outcome of interpersonal communication.

Social capital, borrowing from Coleman (1990) and Putnam (1993), is defined as relations between individuals which can be called upon to enable action. This pragmatic definition is important in that social capital is more than just linkages between people, but linkages that can be mobilized for purposes of reaching some kind of goal. In mass-mediated communication, the communicator, while expecting action, does not expect action from specific individuals, as the levels of trust and reciprocity between communicator and receivers are low due to the nature of mass-mediated communication. Baudrillard (1981:169-170) has argued for much the same approach, though to a greater extent, saying that "[t]he mass media are anti-mediatory and intransitive ... making all processes of exchange impossible." In addition, both communicator and receiver have little personal knowledge of each other. Interpersonal communication, on the other hand, is constructed around the belief that social capital will be mobilized by specific actors, and failure to do so could result in a breakdown in the relationship.

This combination of audience size, commodification, and expectations of the mobilization of social capital actors makes a strong enough distinction between mass-mediated and interpersonal communication to hypothesize that these types of communication will be processed differently at the individual level. That is, people may be more likely to act on interpersonal communications than mass-mediated communications. The proliferation of mass-mediated communication in present day society, though, calls this hypothesis into question.



Self, for the present study, is that aspect of an individual that is observable by the individual and all others (Schwalbe 1991). This definition of the self rests on Mead's (1934:135) idea that the self is constructed through "social experience and activity...." In the present study, the self decides to act or not act on information coming either from a mass communication source (radio) or interpersonal (telephone call from a good friend). While the decision to act on these types of information will have links to past experiences, in the present situation it will be the self which must decide on what action is to be taken.

There has been a great deal of attention given to the evolution of the self in modern society, some of which has focused on the part played by mass-mediated communication. Kenneth Gergen (1991) has put one approach forth. According to Gergen, the number and variety of relationships now available to the individual in mass-mediated form (radio, television, etc.) far exceeds the number of such relationships available at the interpersonal level. This gives impetus to the individual to focus on mass media when constructing the self and identity, as identities for every occasion are available from these channels of communication with very little expected from the viewing self.

Giddens (1991) has argued that society cannot be separated from the media through which language travels.  Electronic mass media, which is a prevalent part of the modern social landscape, offers instantaneous images of distant places, presenting the self a collage of "distant events into everyday consciousness..." (27). This has the effect of partially separating the self from the immediate space in which s/he occupies. This places mass media institutions in the role of producers of reality. The self, then, must build trust in these institutions, and will turn to them for information concerning life events.

These ideas of the modern self as a product of mass media are not new. Twenty years ago, Lasch (1979:268) lamented about how the mass media were taking over the socialization of children in our society. Turner (1976), while not speaking specifically about the media, contended that the traditional self, echoing Mead's work on the I and the me, was composed of both institutional and impulsive focuses. This traditional self has (d)evolved into a new self which is more inclined to follow impulsive forces. The impulsive self was an underlying factor in the many social movements of the 1960s. The tactics used by those involved in these movements -- violence and rage against authority figures -- were considered to be expressions of an impulsive self. This shift from the institutional-directed self to the impulsive-directed self was indicative of larger social changes in which the self lost specific characteristics, and instead was constantly modifying to fit the situation.


This framing of the modern self is not to say that modern individuals are cultural dupes, easily manipulated by mass media. The modern self is also attached to real others, and must engage in negotiations with these others if the self expects to have many life chances. Goffman's (1959, 1971) work on the presentation of self is still appropriate in understanding our actions and sense of self. Blumstein (1991) argues that close relationships, such as marriage, are basic in constructing concepts of self. In addition, any sense of community arises through conversation (White 1992), as well as personal experiences and meaning structures (Hogg 1992:74). Finally, not all needs can be met by mass-mediated communications, forcing people to leave their armchairs in search of specific others (Rubin and Rubin 1985).

The notion that communication is basic to the construction of self has been known by sociologists since the founding of the discipline. Communication, though, takes many forms, including interpersonal and mass-mediated. Given the differences in these specific types of communication, though, we would expect individuals to process them differently, and each would have a different impact on the self. At the same time, the quantity of electronic mass-mediated information available to the self in modern society rivals that of interpersonal communication, leaving us with a dilemma that cannot be easily answered. There are theoretical and methodological tools that are available which we can use to begin to understand how these types of information are processed. One such tool is the third-person effect.


The main thrust of research on the third-person effect is that people will tend to say that others are more affected by mass-mediated messages and images than themselves (Cohen and Davis 1991; Davidson 1983; Fields and Schuman 1976; Gunther 1991). Atwood (1994) has argued that this phenomenon is a consequence of our need to enhance our feelings of self-worth, which is accomplished by thinking that we have more control over our actions than do others. This is further supported by the findings that when the message is positive, people are more likely to say that they will be more affected by the message than others a reversal of the third-person effect (Duck and Mullin 1995). Yet to be determined is whether people tend to say that others are also more likely to be affected by interpersonal communications and images than themselves (as the self could gain status by saying that others are more controlled than themselves). Given the recent theoretical insights concerning the modern self mentioned above, and the significance of individualism in the U.S. (Hewitt 1989), there is reason to believe that the modern self will make little distinction between the control efforts embedded in mass-mediated and interpersonal communications. Using measures taken from research aimed at measuring the third-person effect, we can investigate the differences, if any, people make between these forms of communication.


The major drawback to this formulation of the problem is that evidence for this hypothesis would be revealed by a non-finding--that people will act towards interpersonal and mass-mediated communications in much the same way. Instead, I propose the following hypothesis:

When an individual is asked to act on information, a third-person effect will be stronger when that information is mass mediated than when communicated interpersonally.


Data were collected in the spring of 1996 from undergraduate students enrolled in sociology survey courses at a major southern university. A total of 89 students took part in the vignette study, of which 87 questionnaires were completed (55 by women and 32 by men). Students received extra credit for participation.

The vignette asked respondents to imagine they heard about an event that was taking place at the university's football stadium within the next hour or so. The event was absolutely free and was to be very exciting, "one of the most exciting events to ever take place." Respondents ostensibly heard the information in the evening, and only had half-an-hour to an hour to plan for the event (the event was to take place at 7 PM, and the information was heard between 6 and 6:30 PM). Respondents were then asked to choose between seven different possible responses to the message, ranging from "be at the event at 7 PM" to "pay no attention to the information." Respondents were also asked to give a brief explanation for their choice of action. These responses were collapsed into a three point scale (Go To Event = 1; Maybe Go To Event = 2; Do Not Go To Event = 3), so logical comparisons could be made with other responses which were asked on a three point scale. (See Appendix for the original distribution of these responses and how they were collapsed.)

To test for ordering effects, half the respondents were told that the information was in the form of a radio commercial, while the other half heard it from a good friend (the information was stated exactly the same in both vignettes). After giving a course of action for themselves, respondents were asked whether or not they thought (1) their friends and (2) most other people would go to the event. The respondents were then asked if they would be more or less likely (or no difference) to go to the event if they heard the information from the other source. Again, they were asked whether or not their friends and most other people would go upon hearing the information from the other source. Each of these five questions were asked on a three point scale (with "Don't Know" and "No Difference" serving as middle categories). Vignette questionnaires took approximately ten minutes to fill out, and respondents were debriefed after finishing.



Table 1 contains the mean scores for respondents’ likelihood of going to the event when the message was either mass communicated (via radio) or received interpersonally (from a friend). The origin of the information (radio or friend) made little difference in respondents’ reports. While people hearing about the event from the radio are more likely to feel that they would go to the event, people hearing about the event from a friend are more likely to say that significant others would act on the information. Only the latter difference is statistically significant, and can probably be explained by saying that we would expect a higher level of trust among our friends as compared to the mass media, though respondents themselves are still beyond this type of control.

While there is little difference in the probability of acting on information coming from a friend or the radio when comparing across categories of potential recipients (self, significant others, generalized others), a third-person effect does seem to be operating across these groups. Tests of mean differences between self and both categories of others were all significant at the .01 level.

Table 1: Means for Acting on Information by Source of Information (Standard Deviations in Parentheses)


Self Significant Others Generalized Others
Radio 1.99
Friend 1.88


To test for spurious relations of ordering and source effects, a multinomial logit analysis was constructed to test for any such effects in addition to the observed third person effect. Table 2 contains the parameter estimates of the main effects of ordering (hearing the information first from a friend vs. radio), source of information (friend vs. radio), and recipient of information (self, significant others, and generalized others). It was found that source had no main effect or interacted with any other variable. As this is the variable of interest in the hypothesis, I have kept it as a main effect, but have not included it as an interaction variable.

As with interactions in regression models, the main effects of order of information and recipient of the information are now dependent on other variables in the model (DeMaris 1992). This interaction model does show a strong ordering effect, with people hearing the information first from a friend more likely to say that generalized others will probably not act on the information. Respondents hearing the information first on the radio were more likely to say that self and significant others would probably not go to the event. The negative interaction estimates emphasize the fact that people hearing the information from a friend were more likely to say that self and significant others will act on the information. That there is no difference within groups with regards to the source of the information highlights the fact that the origin of the information makes little difference in how it was processed, which does not support the proposed hypothesis.

Table 2: Logit Model of Acting on Information as Function of Order, Source, and Recipient
(z-Scores in Parentheses) (n= 524)

Parameters Act on information to
"probably not go"
Act on information to
"maybe go"
Intercept -3.32*
Hearing From a Friend First (Order) 2.00*
Hearing From a Friend (Source) -0.07
Self as Recipient of Information (Self) 3.81*
Significant Others as Recipient of Information (Significant Others) 2.79*
Order X Self -2.50*
Order X Significant Other -2.46*

Pearsons's Chi-Square = 16.52 (p = 0.09) (df = 10)
Concentration = 0.06 (df = 24, 1022)
* p < 0.05
Note: This does not represent 524 independent cases, as each respondent was asked to respond to hearing the information from a friend and the radio.



The results I have reported differ from the communications studies reported in the 1940s and '50s. Instead of finding an attenuating third-person affect when moving from mass-mediated to interpersonal communications, there seemed to be little difference between the processing of these types of information. This might not come as a surprise to researchers who argue that mass media have become as powerful, if not more so, than interpersonal communication (Kellner 1990; Mitroff and Bennis 1989; Parenti 1993; Postman 1985; Schiller 1989). The modern self, possibly saturated by both types of communication, does not seem to differentiate between acting on interpersonal or mass-mediated communications. At the same time, this saturation of types of communication has made us impervious to control efforts embedded within them, though we feel that others are not so fortunate.

These findings call into question the expectation of social capital being mobilized in interpersonal communication, as the likelihood of acting on the information was similar for both sources. It could also be a consequence of the message, in that respondents treated interpersonal communication that was ambiguous and possibly misleading as a commodity that a friend was trying to sell. This highlights the importance of understanding not only the content of information, but context as well. In addition, the findings could be an outcome of the role the respondent was asked to play in the vignette. If the respondent had been asked to predict the action of others if contacted by the respondent, there may have been a difference between this type of communication (information originating from the respondent) and mass-mediated communication.

In considering other research aimed at investigating the control efforts of individuals, work on the illusion of control, which has found that people tend to see themselves as having more control over situations than they really do, can be cited (e.g., Alloy et al. 1981;Koenig et al. 1992). While the main focus of these studies has been to induce moods across subjects to measure feelings of control, in all cases subjects felt that they had some control over situations in which they were part. For example, Alloy et al. (1981) found that even depressed subjects who are supposed to have the greatest insight concerning self-efficacy thought they had nearly 20 percent control over a random experiment. The third person effect, much like the illusion of control, points to the fact that people want to believe that they have control over other forces the mass media in the case of the third person effect, and a contingent environment in the illusion of control experiments.


It is my contention that while the third-person effect is only a proxy for self-control, the fact that I found no difference between acting on information from a friend or an impersonal source points in the direction of a modern self engrossed in efforts at self-control. These findings are obviously tenuous, as my study was constructed differently than the earlier studies. They do raise possibilities for further research aimed at gaining further insights into the similarities and differences between how we approach different channels of communication and concepts of self control in a mass media-saturated social environment.


The original categories for the seven-point scale on whether or not a person would go to the event was as follows:

1. Be at Tiger Stadium by yourself at 7pm.
2. Pay no attention to (the commercial/your friend).
3. Call another friend/relative/neighbor to see if they want to go, and go with them if they want to go.
4. Call another friend/relative/neighbor to see if they want to go, but go even if they do not want to go.
5. Call another friend/relative/neighbor to see if they want to go, but stay home if they do not want to go.
6. Call another friend/relative/neighbor to see if they have any information concerning the event, and go if they can tell you more about it.
7. Call another friend/relative/neighbor to see if they have any information concerning this event, but stay home if they cannot tell you more about it.

Recode: Based on action and wording of response.
2,5,7 = Probably Not Go (n = 21)
3,6 = Maybe Go, Maybe Not Go (n = 46)
1,4 = Probably Go (n = 22)


Alloy, Lauren B., Lyn Y. Abramson, and Donald Viscusi (1981). "Induced Mood and the Illusion of Control," Journal of Social Psychology, 41:1129-1140.

Atwood, Erwin L (1994). "Illusions of Media Power: The third-person effect," Journalism Quarterly, 71:269-281.

Baudrillard, Jean (1981). For a Critique of the Political Economy of the Sign. St. Louis, MO: Telos.


Berelson, Bernard R., Paul F. Lazarsfeld, and William N. McPhee (1954). Voting. Chicago, IL: University of Chicago Press.

Blumstein, Philip (1991). "The Production of Selves in Personal Relationships," pp. 305-322 in The Self-Society Dynamic, Judith A. Howard and Peter L. Callero (Eds.). New York: Cambridge University Press.

Cohen, Jeremy and Robert G. Davis (1991). "Third-Person Effects and the Differential Impact on Negative Political Advertising," Journalism Quarterly, 68:680-688.

Coleman, James S (1990). Foundations of Social Theory. Cambridge, MA: Belknap/Harvard Press.

Davidson, Phillips W. (1983). "The Third Person-Effect in Communication," Public Opinion Quarterly, 47:1-15.

Demaris, Alfred. (1992). Logit Modeling. Newbury Park, CA: Sage.

Duck, Julie M. and Barbara-Ann Mullin (1995). "The Perceived Impact of the Mass Media: Reconsidering the third person effect," European Journal of Social Psychology, 25:77-93.

Fields, James M. and Howard Schuman (1976). "Public Beliefs about the Beliefs of the Public," Public Opinion Quarterly, 40:427-488.

Gergen, Kenneth (1991). The Saturated Self. New York: Basic Books.

Giddens, Anthony (1991). Modernity and Self-Identity. Stanford, CA: Stanford University Press.

Goffman, Erving (1971). Relations in Public. New York: Harper and Row.

Goffman, Erving (1959). The Presentation of Self in Everyday Life. Garden City, NY: Doubleday Anchor.

Gunther, Albert (1991). "What We Think Others Think: Causes and consequences in the third-person effect," Communication Research, 18:355-372.


Hewitt, John P. (1989). Dilemmas of the American Self. Philadelphia: Temple University Press.

Hogg, Michael A. (1992). The Social Psychology of Group Cohesiveness. Washington Square, NY: New York University Press.

Katz, Elihu. (1957). "The Two-Step Flow of Communication: An up-to-date report on an hypothesis," Public Opinion Quarterly, 21:61-78.

Katz, Elihu and Paul F. Lazarsfeld (1955). Personal Influence. Glencoe, IL: The Free Press.

Kellner, Douglas (1990). Television and the Crisis of Democracy. Boulder, CO: Westview Press.

Koenig, Linda J., Caroline M. Clement, and Lauren B. Alloy (1992). "Depression and the Illusion of Control: The role of esteem maintenance and impression management," Canadian Journal of Behavioural Science, 24:233-252.

Lasch, Christopher (1979). The Culture of Narcissism. New York: Warner.

Lazarsfeld, Paul F., Bernard R. Berelson, and Hazel Gaudet (1948). The People's Choice, (2nd ed.). New York: Columbia University Press.

Mead, George H. (1934). Mind, Self, and Society. Chicago: University of Chicago Press.

Merton, Robert K. (1949). "Patterns of Influence: A study of interpersonal influence and communications behavior in a local community," pp. 180-219 in Communications Research, 1948-9, Paul F. Lazarsfeld and Frank N. Stanton (Eds.). New York: Harper and Brothers.

Mitroff, Ian I. and Warren Bennis (1989). The Unreality Industry. New York: Oxford University Press.

Neuman, W. Russell (1991). The Future of the Mass Audience. New York: Cambridge University Press.


Parenti, Michael (1993). Inventing Reality (2nd ed.). New York: St. Martin's Press.

Postman, Neil (1985). Amusing Ourselves to Death. New York: Penguin.

Purcell, Kristen (1997). "Towards a Communication Dialectic: Embedded technology and the enhancement of place." Sociological Inquiry 67:101-112.

Putnam, Robert D. (1993). Making Democracy Work. Princeton, NJ: Princeton University Press.

Rubin, Alan M. and Rebecca B. Rubin (1985). "Interface of Personal and Mediated Communication: A research agenda," Critical Studies in Mass Communication, 2:36-53.

Schiller, H. (1989). Culture, Inc.. New York: Oxford University Press.

Schwalbe, Michael L. (1991). "Social Structure and the Moral Self," pp. 281-303 in The Self-Society Dynamic, Judith A. Howard and Peter L. Callero (Eds.). New York: Cambridge University Press.

Thompson, John B. (1988). "Mass Communication and Modern Culture: Contributions to a critical theory of ideology." Sociology, 22:359-383.

Turner, Ralph H. (1976). "The Real Self: From institution to impulse," American Journal of Sociology, 81:989-1016.

White, Harrison C. Identity and Control. Princeton, NJ: Princeton University Press.


Toby A. Ten Eyck is a Ph.D. candidate in the Department of Sociology at Louisiana State University. His current work is focused on newspaper arenas and the reporters and sources involved in the coverage on food irradiation. Email:


Back to Journal Cover Page