broadcasting and narrowcasting: how audience size …

“Broadcasting and Narrowcasting: How Audience Size Impacts What People Share” © 2013 Alixandra Barasch and Jonah Berger; Report Summary © 2013 Marketing Science Institute MSI working papers are distributed for the benefit of MSI corporate and academic members and the general public. Reports are not to be reproduced or published in any form or by any means, electronic or mechanical, without written permission.

Marketing Science Institute Working Paper Series 2013 Report No. 13-116 Broadcasting and Narrowcasting: How Audience Size Impacts What People Share Alixandra Barasch and Jonah Berger

Report Summary Consumers communicate with people every day. These exchanges influence consumer behavior in important ways: word-of-mouth affects everything from the products people buy to the websites they visit. Most research on word-of-mouth has focused on these impacts. Less attention has been focused on how the audience itself might affect word-of-mouth. In this report, Alixandra Barasch and Jonah Berger investigate how audience size impacts what people talk about and share. They suggest that speaking to a large audience (“broadcasting”) leads people to share more self-presentational content, while speaking to one person (“narrowcasting”) leads people to share more useful content. They further suggest that these effects are driven by shifts in sharer focus from self-focus to other-focus. The authors test and support these hypotheses across five experimental studies. They find that broadcasting leads people to share things that make them look better, use more positive language, and reframe negative events to make themselves look less bad. In contrast, narrowcasting decreases the tendency to share self-presentational content and leads people to share more things that are useful to their conversation partner. Their studies demonstrate how shifts in sharer focus drive these effects across a wide range of manipulations and outcomes.

Managerial implications Given consumers’ ability to communicate with many people at once through social media, understanding how audience size impacts communication is increasingly important. While models of communication often focus on whether messages produce attitude change in the recipient, this study analyzes the sender’s choice of what message to share in the first place, with important implications for interpersonal communication and word-of-mouth strategies. At an individual level, these findings can help people manage how they are perceived by others. At the firm level, they suggest scenarios where “narrowcasting” or “broadcasting” are more effective for marketing communication. For example, companies that sell useful products (e.g., healthcare) may generate greater word-of-mouth by providing web forms that allow for narrow, personalized messages. Conversely, companies that sell products related to self-presentation (e.g., designer clothing) may increase sharing by facilitating easy broadcasting (e.g., one-click posting on social media). Alixandra Barasch is a doctoral candidate in Marketing, The Wharton School, University of Pennsylvania. Jonah Berger is the James G. Campbell Jr. Associate Professor of Marketing, The Wharton School, University of Pennsylvania. Acknowledgments The authors thank Ezgi Akpinar, Amit Bhattacharjee, Cindy Chan, Zoey Chen, Deborah Small, and seminar participants at the University of Pennsylvania for their insightful comments. This research was supported by the Dean’s research fund at Wharton and an Alex Panos research grant.

Marketing Science Institute Working Paper Series 1

Consumers communicate with dozens of people every day. They talk to friends, chat

with neighbors, and gossip with co-workers. These social exchanges have an important impact

on consumer behavior, and word-of-mouth affects everything from the products people buy to

the websites they visit (e.g., Chevalier and Mayzlin 2006; Trusov, Bucklin, and Pauwels 2009).

Communication, however, requires more than one party; indeed, people cannot share

things without an audience (real or implied). One fundamental aspect of the audience is its size.

Sometimes communication involves talking to many people, or broadcasting. In other instances,

it involves talking to just one person, or narrowcasting. At a party, for example, consumers can

find themselves talking to a group of friends or just one companion. Similarly, people may be

responding to an email chain that involves a crowd of co-workers or just one. Might these

differences in audience size affect what people talk about? And if so, how?

This paper investigates how mere audience size impacts what people share. We suggest

that, relative to narrowcasting, broadcasting leads people to share more self-presentational

content. Narrowcasting, however, encourages people to share more useful content. Further, we

show that these effects are driven by where communicators focus their attention. People naturally

tend to focus on the self, but communicating with just one person heightens other-focus, which

in turn impacts what people pass on.

This paper makes two primary contributions. First, although communication always

involves an audience (either real or imagined), no work has examined how mere audience size

affects what people share. While models of communication (e.g., McGuire 1985) often focus on

whether messages will produce attitude change in the recipient, we focus on the other side of the

equation, analyzing the sender’s decision of what message to share in the first place, and

considering how the mere number of recipients might play a role. Given consumers’ newfound

ability to communicate with many people at once through social media, understanding how

audience size impacts communication has become increasingly important.

Second, we contribute to the larger discussion in the word of mouth literature on when

the sender or the receiver is more important in what people share (Berger 2013; Dichter 1966).

While communication involves multiple parties, little is known about when and why what people

pass on is driven more by the sender versus the recipient of the communication.

We address both these points, illustrating how audience size shapes both what people talk

about and whether sharing is driven relatively more by themselves versus others.


Word of Mouth

Most research on word-of-mouth has focused on its consequences, or how it influences

choice, diffusion, and sales. Word-of-mouth can affect consumer behavior by generating

awareness, encouraging people to update their beliefs, or producing normative pressures (Van

den Bulte and Wuyts 2009). Marketing scientists have studied the causal impact of word-of-

mouth in a variety of domains, ranging from book and movie sales (Chintagunta, Gopinath, and

Venkataraman 2010; Chevalier and Mayzlin 2006) to adoption of websites, pharmaceutical

drugs, and television shows (Godes and Mayzlin 2004; Trusov, Bucklin, and Pauwels 2009;

Iyengar, Van den Bulte, and Valente 2011).

There has been less attention, however, to how the audience impacts what people talk

about and share.

Research on audience effects has mainly considered how tie strength affects

communication (Brown and Reingen 1987; Frenzen and Nakamoto 1993). Strong ties are close

others like family members and good friends, while weak ties are more distant others, such as

acquaintances (Granovetter 1973). Research finds that people share with anyone when the value

of information is low, but are more hesitant to transmit valuable information to weak ties

(Frenzen and Nakamoto 1993). Similarly, in networks of referral behavior, strong ties are

perceived as more influential than weak ties, and are more likely to be used as sources of

information for related goods (Brown and Reingen 1987).

But while such research has considered how audience type (i.e., tie strength) affects

communication, might mere audience size impact what people share, and if so, how?

The Current Research

We suggest that audience size affects the type of content people share by altering the

sharer’s focus, or the degree to which the sharer focuses on himself versus the communication

recipient(s) when deciding what to share.


Audience size and sharer focus

People have a strong default tendency to focus on the self. Individuals have privileged

access to their own thoughts and feelings and assume that others will see things the same way

(Ross and Ward 1996). Decades of research on egocentrism show that people disproportionately

attend to their own opinions and interests (Kruger 1999; Chambers and Windschitl 2004).

Whether comparing abilities (Kruger 1999) or predicting others’ preferences (LeRouge and

Warlop 2006) people tend to focus on themselves and inadequately take others’ attitudes and

values into account. Moreover, people have difficulty taking others’ perspective (Dunning, Van

Boven, and Loewenstein 2001; Van Boven, Loewenstein, and Dunning. 2005), and tend to focus

on the self in part because self-relevant information is more accessible (Ross and Sicoly 1979).

This natural propensity toward egocentrism can also be seen in what people share.

Much of our daily communication is fixated on the self. Studies of human conversation report

that self-disclosure is the most common topic (Emler 1990) and that over 60% of everyday

speech consists of one’s personal experiences and relationships (Dunbar, Marriott, and Duncan

1997; Landis and Burtt 1924). That percentage is even higher in social media, where 80% of

users focus on the self (Naaman, Boase, and Lai 2010). Neuroscientific evidence suggests that

self-disclosure is intrinsically rewarding, activating regions of the brain associated with primary

reinforcers like food and attractive members of the opposite sex (Tamir and Mitchell 2012).

We suggest that broadcasting should do little to move people from this natural tendency

for self-focus. Considering others is a deliberate process that requires substantial time, mental

effort, and motivation (Epley et al. 2004; Apperly et al. 2006). People do not consider others’

beliefs and knowledge unless something in their environment encourages them to do so (Zhang

and Epley 2012). There is little reason to believe that broadcasting would encourage such effort;

in fact, it may even increase the effort necessary to take others’ perspectives given that there are

many others to consider. Because groups are not seen as unified or possessing stable

dispositions, perceivers often do not waste time and cognitive energy trying to focus on a group

in the moment of interaction (Hamilton and Sherman 1996).

Narrowcasting, in contrast, should encourage other-focus. Having people think about a

specific other mitigates egocentrism because it makes others more concrete (Alicke and Govorun

2005). Seeing a single person’s name, for example, promotes individuation, or recognition of

that person’s distinct identity, which reduces the above-average effect (Alicke et al. 1995).


Similarly, charity appeals featuring singular targets lead people to help others more because they

make the victim more vivid (Jenni and Loewenstein 1997). Along these lines, sharing with just

one other person should make the audience more concrete and vivid, which should increase the

attention they receive (Taylor and Thompson 1982). Individual targets are also assumed to be

stable, consistent, and coherent (Hamilton and Sherman 1996), which causes people to focus on

those individuals and make on-line judgments about them (Hastie and Park 1986).

Audience Size and What People Share

By shifting sharer focus, we suggest that audience size impacts two fundamental word-

of-mouth drivers: self-presentation and helping others (Engel, Blackwell, and Kegerreis 1969;

Hennig-Thurau et. al. 2004; Dichter 1966).

Self-Presentation. People often share things to present themselves in a positive, rather

than negative, light. Social interactions can be seen as a performance where people promote

favorable impressions of themselves rather than unfavorable ones (Goffman 1959). Indeed, the

tendency to self-enhance, or bolster the self-concept, is one of the most central human

motivations (Fiske 2001). It is also one of the most studied drivers of word-of-mouth (Engel et

al. 1969; Hennig-Thurau, et. al. 2004; Packard and Wooten 2012; Wojnicki and Godes 2011).

People can present themselves positively in communication by distancing themselves

from negative personal outcomes (Sedikides 1993; Sedikides and Strube 1995) or negative

experiences (Richins 1984). Negative content is less viral than positive content (Berger and

Milkman 2012), potentially because it reflects negatively on the sender (i.e., people don’t want to

be known for sharing depressing stories). Further, people may be more likely to talk about novel,

interesting, or surprising products (Berger and Milkman 2012; Moldovan, Goldenberg, and

Chattopadhyay 2011) because doing so reflects well on the person sharing the information,

making him look more interesting and in-the-know (Berger and Schwartz 2011).

Compared to narrowcasting, we suggest that broadcasting should lead people to share

more self-presentational content because it does little to shift people away from their default self-

focus. When a person is focused on the self, he automatically associates himself with favorable

attributes more than with unfavorable ones (Paulhus, Graf, and Selst 1989; Paulhus and Levitt

1987). We predict that broadcasting will not discourage this natural tendency to view and present


the self in a non-negative light. Narrowcasting, however, should shift people’s focus away from

the self and towards the audience, thus reducing the sharing of self-presentational content.

H1a: Relative to narrowcasting, broadcasting will lead people to share more self-

presentational content.

H1b: This will be driven by shifts in sharer focus.

Helping others. Another major reason people share is to help others. Interview data

suggest that over 20% of word-of-mouth conversations are motivated by “altruistic” desires to

guide people towards good consumption experiences (Sundaram, Mitra, and Webster 1998).

When interacting with each other, people are always trying to be relevant and say things that are

pertinent to the discussion (Grice 1975). To do this, people often tune messages to their

audience, tailoring what they say to suit the audience’s knowledge or attitudes (Schau and Gilly

2003; Clark and Schaefer 1989; Fussell and Krauss 1989). People will be more likely to bring up

golf tips, for example, when talking to a golfer than a theater buff.

One way people help others is by sharing useful information (e.g., discounted products or

good restaurants, Dichter 1966; Hennig-Thurau et al. 2004).1 People are more likely to share

marketing messages that have more utilitarian value (Chiu et al. 2007), and more practically

useful news articles are more likely to go viral (Berger and Milkman 2012).

We suggest that narrowcasting will encourage people to share useful information by

boosting other-focus. Increased other-focus should lead sharers to see others as having their own

theory of mind, with unique mental states, thoughts, and knowledge (Pylyshyn 1978; Wellman

1988). Just like a child who develops to understand that others have beliefs and desires that are

different from one’s own (Piaget 1926), narrowcasting should discourage egocentrism and

encourage consideration of the audience’s point of view. This other-focus, in turn, should

facilitate the process of “audience tuning” (Higgins 1999) and lead people to share things that are

more useful or relevant to their audience.

H2a: Relative to broadcasting, narrowcasting will lead people to share more useful

content.

H2b: This will be driven by shifts in sharer focus.

1 Though one could argue that sharing useful information is simply another instantiation of self-presentation, this is

not always the case. In many instances, people share useful information that doesn’t make them look particularly

good (e.g., I just bought a horrible camera, and I don’t want you to make the same mistake). Thus while self-

presentation may lead people to share some useful information, helping others is a separate key driver.


In sum, we suggest that mere audience size impacts whether people share two different

types of content. Compared to narrowcasting, broadcasting should lead people to share more

self-presentational content because talking to a group does not shift people from their natural

self-focus. Narrowcasting, however, should increase other-focus and thus encourage people to

share more useful content. This cognitive mechanism builds on prior literature to explain how

audience size influences what communicators talk about and why.

Importantly, we are not explicitly comparing the sharing of self-presentational content to

useful content, as these are not always competing motives. Rather, we test how audience size

impacts the sharing of each type of content separately through shifts in sharer focus.

We test these predictions in five experiments. Study 1 investigates how audience size

impacts the types of events that people share. Study 2 examines how audience size impacts real,

face-to-face conversations, while also providing an initial test of the underlying process (i.e.,

sharer focus). Studies 3a and 3b more directly test this mechanism by manipulating sharer focus

and investigating how it shapes sharing of self-presentational and useful content. Finally, Study 4

further tests the process and demonstrates its robustness in a new domain. By directly

manipulating audience size, we demonstrate how it impacts what people share and illustrate the

underlying psychological process driving these effects.

Study 1: Talking about your Day and Self-presentation

In our first study, we simply manipulate audience size and examine how it affects what

people share. We gave participants a list of events that supposedly happened to them on an

imaginary day, some of which make the self look bad and some of which make the self look

good. Then, we asked them to write a short description of that day to share with either one

person (narrowcasting) or a group of people (broadcasting).

We predicted that compared to narrowcasting, broadcasting would lead to greater sharing

of self-presentational content. We test how audience size impacts which events people share (i.e.,

events that make the self look good versus bad), as well as whether they reframe negative events

to make the self look less bad.

Methods


One-hundred ninety-two participants were asked to imagine describing their day in an

email. To provide a level of control, we presented everyone with the same imaginary day. It

included five events that made the self look good (e.g., “Your friend complimented you on your

new shirt, which is one of your favorite brands”) and five events that made the self look bad

(e.g., “You overslept and realized you missed your favorite morning show”).2 The events all

centered around product or consumption experiences, such that talking about them could

generate word of mouth for products or brands. Participants selected which events to discuss and

were encouraged to add details and elaborate beyond what was described in the text.

The only difference between conditions was the size of the group they communicated

with. In the narrowcasting (broadcasting) condition they were asked to think about one friend (a

group of friends) they often talk to, and to imagine they were talking to that friend (those friends)

in an email. A manipulation check confirmed that people in the narrowcasting condition

imagined writing to fewer others than in the broadcasting condition (M = 1.0 vs. 5.1, t(190) = -

8.14, p < .001). See Appendix for an example of what participants wrote in each condition.

Our key dependent variable was how many events participants mentioned that made the

self look good versus bad (i.e., number of events that make the self look good minus the number

that make the self look bad).

We also tested whether participants reframed the events that made them look bad to find

a silver lining. For example, an individual might say that it is okay he slept through his alarm and

missed his favorite show because he needed to catch up on sleep or because he knew the show

would be replayed later. Two independent coders were given a short description of reframing

and then rated each participant’s passage based on how much the sharer “reframed negative

things to make the self look less bad” (1 = “not at all” to 5 = “a great deal”). Coders’ ratings

were highly correlated (r = 0.81) and averaged to form a reframing score.

2A pretest of self-presentation (N = 52, how good or bad the content made the sharer look) shows that the 10 events

selected made the self look good and bad, respectively. The 5 events that made the self look good were rated as

making the self look better than the scale midpoint (M = 5.33, t(51) = 15.52, p < .001), and the 5 events that made

the self look bad were rated as making the self look worse than the scale midpoint (M = 2.58, t(51) = 18.16, p <

.001).


Results

As predicted, compared to narrowcasting, broadcasting led participants to engage in more

self-presentation (t(190) = 2.11, p = .04). While narrowcasters mentioned slightly more negative

events (M = -0.14), broadcasters mentioned slightly more positive events (M = 0.37).

Broadcasting also led to participants reframing negative events more to make the self

look less bad (Mbroad = 3.13 vs. Mnarrow = 2.79, t(190) = -2.54, p = .01).3

Discussion

Study 1 provides preliminary support for our prediction that audience size impacts what

people share. Compared to narrowcasting, broadcasting encouraged self-presentation, increasing

the number of events participants mentioned that made them look good rather than bad, and

leading them to reframe the negative events they did mention to make the self look less bad.

Study 2: Real Conversation and Text Analysis

Study 2 has five goals. First, we examine whether the effect of audience size on self-

presentation persists in real interactions where people communicate with one or many

conversation partners.

Second, we control for audience closeness. The results of Study 1 are supportive, but one

could argue that narrowcasters might have imagined closer others, which made them more

comfortable sharing events that make them look bad. To rule out this possibility, all participants

in Study 2 interacted with complete strangers. This allows us to test whether our effects persist

even in a situation where audience closeness is held constant across audience size conditions.

Third, we provide an initial test of the process, looking at whether audience size impacts

sharer focus. We ask participants to report the extent to which they are thinking about themselves

versus others and examine whether it mediates the effects of audience size on what people share

(i.e., self-presentational content).

Fourth, we also include two control conditions to test our theorizing that self-focus and

self-presentation is the default tendency. If we are correct that broadcasting does little to move

people from their natural tendency to focus on the self, then the broadcasting condition should be

3Conditions did not differ in the number of words written (Mbroad = 137 vs. Mnarrow = 132; t(190) < .8, p > .3).


no different from the control conditions. Narrowcasting, however, should encourage people to

think more about others and reduce self-presentation.

Finally, we use a more open-ended method to test our theory. We ask participants to talk

freely about a restaurant experience and use textual analysis software (Linguistic Inquiry and

Word Count or LIWC; Pennebaker, Francis, and Booth, 2001) to measure the valence of what

they share (i.e., number of positive (e.g., love, nice, sweet) versus negative (e.g., hate, nasty,

annoyed) words used). Language content has been used to evaluate important word-of-mouth

outcomes (Moore 2012; De Angelis et al. 2012), and LIWC has been used across psychology

and marketing (Berger and Milkman 2012; Robertson and Doig 2010) to measure psychological

constructs from passages of text (Tausczik and Pennebaker 2010). Less negative emotional

expression makes the self look better (e.g., increases liking and makes one seem more popular,

Bell 1978; Locke and Horowitz 1990). Consequently, we predict that compared to

narrowcasting, broadcasting should lead people to share more positively-valenced content.

Methods

One-hundred seventy students at a Northeastern university completed a brief interaction

task in the laboratory. They were asked to have a short conversation with other student(s) about

a recent restaurant experience. Participants were randomly assigned to one of four conditions.

The first two conditions mirrored Study 1. In the narrowcasting (broadcasting) condition,

participants were assigned to speak to a single partner (small group of 4-5 individuals). Each

pair or group was introduced and asked to have a conversation with each other. To facilitate

interaction and data analysis, participants first wrote down what they wanted to share, which

allowed us to analyze the data at the individual level and avoid any concerns about the

interdependence of group conversation driving the results (Bales 1951). Then, participants read

aloud what they had written to their partner or group.

There were also two control conditions. The first was a “no audience” control, in which

people wrote about a recent restaurant experience but were not instructed to share it with others.

This allowed us to control for content type (restaurant experience) but analyze sharer focus and

self-presentation when no audience of any size is involved. To ensure that writing alone did not

shift sharer focus, we also included a second “no task” control condition, where participants did

not write or share any content and just went straight to the dependent measures.


After completing the speaking or writing task, participants were then asked “How much

are you thinking about yourself right now?” and “How much are you thinking about others right

now?” (order counterbalanced, 1 = “not at all” to 7 = “a lot,” differenced to form a measure of

sharer focus).

Finally, we collected ancillary measures to test potential alternative explanations. One

could argue that effects of audience size could be driven by increased anxiety, arousal, or

embarrassment when talking to greater numbers of people. To test this possibility, we asked

participants to report how much they were feeling anxious (i.e., anxious, apprehensive, worried,

and nervous, α = .94; Brooks and Schweitzer 2011), aroused (i.e., passive versus active, mellow

versus fired up, and low energy versus high energy, α = .84; Berger 2011) and embarrassed

(“How embarrassed do you feel?”; Kelly and Jones 1997), all on 7-point scales.

Results

Self-Presentation. For all conditions where participants wrote something, we used LIWC

to measure the valence of each participant’s passage (adapted from Berger and Milkman 2012;

difference between the z-scored percentages of positive and negative words).

As predicted, audience size influenced self-presentation (F(2,109) = 6.76, p = .002),

Figure 1. Relative to narrowcasting (M = -.78), broadcasting increased the positivity of what

participants shared (M = .42, t(76) = -3.39, p = .001). As expected, there was no difference

between the broadcasting and “no audience” control conditions (M = .31, t < .4, p > .7).4

Sharer Focus. As predicted, condition also influenced sharer focus (F(3,166) = 6.25, p <

.001), Figure 2. Compared to broadcasting (M = .44), narrowcasting shifted focus from the self

towards others (M = -1.50, t(76) = -3.31, p = .001). There was no difference between

broadcasting and either control condition (no audience: M = .11, t < .6, p > .5; no task: M = .78, t

< .7, p > .5).

Mediation Analysis. The bootstrap mediation method (MEDIATE macro; Hayes,

Preacher, and Myers 2011) illustrates that differences in sharer focus drove the impact of

condition on sharing self-presentational content (total indirect effect = .24, standard error = .16,

4 There were no differences between the conditions in number of words written (Mbroad = 153, Mnarrow = 165, Mcontrol

= 141; F(2,109) < 1.4, p > .2).


95% CI does not include zero [.01, .60]), Figure 3. Relative to narrowcasting, broadcasting led

people to think more about themselves (a1 = 1.94), which increased self-presentation (b1 = .09).

Alternative Explanations. Ancillary analyses cast doubt on several alternative

explanations. There was no effect of condition on anxiety (F < .5, p > .6), arousal (F < 1.3, p >

.3), or embarrassment (F < .2, p > .9).

Discussion

Results of Study 2 further support our conceptualization. First, audience size influenced

what people shared. Compared to narrowcasting, broadcasting increased self-presentation,

boosting the positivity of what participants passed on.

Second, the results provide preliminary support for our hypothesized mechanism.

Audience size impacts sharer focus, which in turn impacts what people share. Compared to

broadcasting, narrowcasting shifts focus towards others, and as a result, decreases self-

presentation.

A more objective measure underscores these differences in sharer focus. Personal

pronoun usage (e.g., I or you) provides information about the speaker’s focus, attention, and

priorities (Tausczik and Pennebaker 2010). First-person singular pronouns (e.g., I, my, and mine)

are used when people are focused on themselves (e.g., when sitting in front of a mirror, Davis

and Brock 1975), while second-person pronouns (e.g., you and your) are used when people are

focused on others (e.g., when high self-monitors engage in peer interactions, Ickes, Reidhead,

and Patterson 1986). Consistent with our theorizing, broadcasting led people to use more self-

focused pronouns (F(2,109) = 9.32, p < .001; Mbroad = 8.79, Mnarrow = 7.59, p = .03), while

narrowcasting increased the use of other-focused pronouns (F(2,109) = 4.30, p = .016; Mnarrow =

1.69, Mbroad = .81, p = .005).5

Third, Study 2 illustrates that broadcasting does little to move people away from their

default tendency for self-focus and self-presentation. Broadcasting had the same effects as the

control conditions on these measures.

Fourth, the study casts doubt on alternative explanations. Even in a situation where

audience closeness was controlled across conditions (i.e., both broadcasters and narrowcasters

5There were no differences between broadcasting and the “no audience” control condition for either self-focused

pronouns (M = 9.23, t(75) < .8, p > .4) or other-focused pronouns (M = .34, t(75) < 1.6, p > .1).


interacted with strangers), audience size still impacted what people shared. Further, anxiety,

arousal, and embarrassment did not different between conditions, casting doubt that they are

driving the effects.

Finally, by analyzing actual conversations, not constraining participants to write about

any particular event, and using an impartial measure of linguistic styles to code for valence and

personal pronouns, our results underscore the generalizability of these effects.

Study 3: Manipulating Sharer Focus

Study 3 looks for further evidence of the role of sharer focus by manipulating it directly

and testing whether it moderates the effect of audience size on sharing self-presentational and

useful content.

As discussed in the introduction, narrowcasting should encourage people to think about

their audience, which should discourage sharing of self-presentational content and encourage

sharing of more useful content.

We use an additional manipulation of sharer focus to test its role in these effects. In

addition to manipulating audience size, we ask half our participants to list the name(s) of the

people receiving their message. Prior work finds that reading specific information about a

comparison target (or even just seeing the person’s name) reduces egocentric tendencies and

increases how much people think about that other person (Weinstein 1983; Alicke et al. 1995).

Consequently, if audience size impacts what people share through shifting sharer focus, as we

suggest, then encouraging individuals to concentrate on their audience by listing the other(s) with

whom they are sharing should make broadcasting look more like narrowcasting.6

Study 3a investigates how audience size impacts whether people share useful content,

while Study 3b investigates how audience size impacts whether people share self-presentational

content.

6A pretest (N = 168) confirmed that listing the names of audience members interacted with audience size to affect

sharer focus (F(1,164) = 4.41, p = .04). Among broadcasters, listing recipients’ names increased how much people

reported thinking about their audience (Mlist = 6.18 vs. Mno list = 5.21, F(1,164) = 11.98, p < .01). Among

narrowcasters, however, whom we expected to already be focused on others, there was no additional effect of listing

names on the degree to which participants reported that they were thinking about the audience (Mlist = 6.00 vs. Mno

list = 5.85 F < .5, p > .5).


Study 3a

Study 3a examines how audience size affects whether people share useful content. We

predict that compared to broadcasting, narrowcasting should boost sharing of useful information.

We also further test alternative explanations. While Study 2 controlled for audience

closeness, here we measure it directly to rule it out again as an alternative driver of the influence

of audience size. In addition, we measure the public nature of the communication to see whether

it could be driving the effects.

Methods

One hundred- and-forty-two respondents were randomly assigned to condition in a 2

(Audience Size: narrow vs. broad) x 2 (Sharer Focus: Control vs. Other-focus) between-subjects

design.

Similar to Study 1, participants were randomly assigned to think about communicating

online with one friend (narrowcasting) or a group of friends (broadcasting).

We separately manipulated sharer focus by asking some participants to list the specific

other(s) with whom they were sharing. Half the participants (Other-focus condition) were asked

to write down the names of the specific person or people with whom they were communicating.

The other half (Control condition) were not.

All participants were then given a list of eight useful things they might share with others

(e.g., “Information about how to buy tickets to a really popular upcoming concert” or “A coupon

for a discount”) and asked to rate how likely they would be to share each with the person(s) with

whom they were communicating.7 Responses to these items were highly correlated (α = 0.78)

and averaged to form a likelihood to share useful content measure.

Finally, we collected ancillary measures to test potential alternative explanations.

Participants were asked “How much do you know about the recipient(s) of these items?”, “How

much do you care about the recipient(s) of these items?”, and “How good of a friend is the

7Pretest participants (N = 47) who rated each item based on its usefulness (1 = “Not at all useful to the recipient” to

7 = “Extremely useful to the recipient”) confirmed that the items were significantly more useful than the scale

midpoint (M = 4.91, t(46) = 7. 88, p < .001).


recipient(s) of these items?” Response to this set of items were averaged to form an audience

closeness measure (α = .85). To measure the public nature of communication they were also

asked “How public do you feel sharing these items would be?”

Results

In addition to a main effect of Sharer Focus (F(1,138) = 4.54, p = .04), a 2 x 2 ANOVA

revealed the predicted Audience Size x Sharer Focus interaction (F(1,138) = 5.65, p = .02), see

Figure 4. Among control participants, narrowcasting boosted the sharing of useful content

(Mnarrow = 4.39 vs. Mbroad = 3.65, F(1,138) = 7.54, p < .01). When participants focused more on

others by listing the names of the people with whom they were sharing, however, this difference

disappeared (Mnarrow = 4.34 vs. Mbroad = 4.50, F < .4, p > .5).

Looked at another way, while there was no effect of Sharer Focus in the Narrowcasting

condition (F < .1, p > .8), there was in the Broadcasting condition, such that listing the name of

the recipient(s) increased people’s willingness to share useful information (F(1,138) = 10.30, p <

.01). As predicted, boosting other-focus led broadcasting to look like narrowcasting.

Alternative Explanations. Ancillary analyses cast doubt on the notion that the effects of

audience size were driven by audience closeness (i.e., tie strength) or public nature of

communication. There was no Audience Size x Sharer Focus interaction on either audience

closeness (F < 2.7, p > .1) or how public participants felt the sharing of the items would be (F <

.1, p > .8). Further, neither construct mediated the sharing of useful content.8

Study 3b

Study 3b uses the same manipulations as Study 3a to test how audience size impacts

sharing self-presentational content.

We also examine whether audience size affects all types of self-presentation equally.

Prior work has noted that self-presentation can be either protective or acquisitive (Arkin 1981).

Protective self-presentation involves distancing oneself from things that make you look bad, such

as negative personal outcomes or experiences (Richins 1984; Sedikides 1993; Sedikides and

8The same measures were also collected for Study 3b and showed the same results. There was no significant

interaction between Audience Size and Sharer Focus for either audience closeness or the public nature of the

communication, and the mediation analysis ruled out both variables as drivers of the likelihood to share interaction.


Strube 1995). People wear fewer school colors after their team loses (Cialdini et al. 1976), for

example, and may avoid sharing negative things that happened to them because it makes them

seem like a “Debbie Downer.” Acquisitive self-presentation involves connecting oneself to

positive personal outcomes (e.g., bragging about how everyone likes your cooking, Brown,

Collins, and Schmidt 1988).

Importantly, while both types of self-presentation can occur, prior work suggests that

people are more motivated to avoid making bad impressions (Baumeister et al. 2001). For

example, people are more likely to underestimate their bad traits than overestimate their good

ones (Klein 1992; Hoorens 1996) and direct comparisons of protective and acquisitive self-

presentation find that protective is more likely to occur (Tice 1991). In the communication

domain, people tend to avoid posting negative information about themselves on social networks

(Gonzales and Hancock 2011).

Consequently, while our theory makes no prediction about whether audience size should

have stronger effects on one type of self-presentation than another, given that prior work finds

protective self-presentation to be more prevalent, we may find the same here.

Methods

One hundred- and-fifty-eight respondents were randomly assigned to condition in a 2

(Audience Size: narrow vs. broad) x 2 (Sharer Focus: Control vs. Other-focus) x 2 (Self-

Presentation Type: Acquisitive vs. Protective) between-subjects design.

First, we used the Audience Size and Other-focus manipulations from Study 3a.

Next, we manipulated acquisitive versus protective self-presentation. Participants were

asked to rate how likely they would be to share each of 15 self-presentation items. We kept the

topic of the items (e.g., grades) similar across conditions, but in the acquisitive (protective)

condition the items were framed as making the self look good (bad). For example, “The fact that

you got a good (bad) grade on your recent test,” and “How you recently got a great new pair of

shoes on sale (spent too much money on an average pair of shoes).”9 Responses to these items

were highly correlated in each condition and combined to create measures of likelihood to share

9 Pretest participants (N = 63) rated each item on how sharing it would make the sharer look (1 = “extremely bad to

the recipient” to 7 = “extremely good to the recipient”). One-sample t-tests confirmed that acquisitive items made

the sharer look good (i.e., better than scale midpoint, M = 5.19, t(62) = 12.64, p < .001), while protective items made

the sharer look bad (i.e., worse than scale midpoint, M = 3.09, t(62) = -13.42, p < .001).


acquisitive self-presentation items (things that make the self look good; α = 0.90) or protective

self-presentation items (things that make the self look bad; α = 0.79).

Results

A 2 x 2 x 2 ANOVA on willingness to share revealed main effects of Audience Size

(F(1,150) = 12.37, p = 0.001), Sharer Focus (F(1,150) = 33.12, p < 0.001), and Self-presentation

type (F(1,150) = 23.10, p < .001), and a two-way Sharer focus x Self-presentation type

interaction (F(1,150) = 4.05, p < 0.05). More importantly, consistent with prior theory, these

effects were qualified by a three-way interaction (F(1,150) = 3.58, p = 0.06). To understand the

nature of this three-way interaction, we examine the two types of self-presentation items

separately.

Among participants who considered protective self-presentation, in addition to main

effects of Audience Size (F(1,78) = 13.20, p < 0.001) and Sharer Focus (F(1,78) = 31.62, p <

.001), the results revealed the predicted Audience Size x Sharer Focus interaction (F(1,78) =

6.90, p = 0.01), see Figure 5A. In the Control condition, the effect of audience size mirrored our

prior studies. Compared to narrowcasting, broadcasting decreased participants’ willingness to

share things that would make them look bad (Mbroad = 2.37 vs. Mnarrow = 3.99, F(1,78) = 19.59, p

< .001). When participants listed the names of the people with whom they were sharing,

however, thereby increasing other-focus, this difference disappeared (Mbroad = 4.51 vs. Mnarrow =

4.77, F < .6, p > .4).

In contrast, among participants who considered acquisitive self-presentation, there was

only a main effect of Sharer Focus (F(1,72) = 6.69, p = 0.01), but no Audience Size x Sharer

Focus interaction (F < .1, p > .9), see Figure 5B.

Discussion

Results of Study 3 extend the findings of the prior studies, provide further evidence for

the mechanism behind the observed effects, and cast doubt on a number of alternative

explanations.

First, audience size influenced what participants were willing to share. Compared to

broadcasting, narrowcasting increased people’s willingness to share useful content.


Broadcasting, however, increased protective self-presentation relative to narrowcasting,

decreasing people’s willingness to share content that made them look bad.

Second, consistent with our underlying conceptualization, these effects were moderated

by sharer focus. When people listed the names of the recipient(s) with whom they would share,

the impact of audience size dissipated and broadcasting looked more like narrowcasting. The fact

that a manipulation which increased other-focus made broadcasting look like narrowcasting

supports the notion that the effects of audience size are driven by shifts in sharer focus.

Third, our findings are inconsistent with a number of alternative accounts. Audience

closeness (how much people knew and cared about their audience) and the public nature of the

conversation did not show the same interactive patterns and did not mediate the effects.

Consequently, it is difficult for these alternatives to explain the effects observed here.

Fourth, consistent with prior work, audience size had a stronger impact on protective

(than acquisitive) self-presentation. We discuss this further in the General Discussion.

Study 4: Public Sharing

Study 4 further tests the underlying mechanism. We manipulate audience size, measure

sharer focus, and examine whether it mediates the effects of audience size on whether people

share self-presentational and useful content.

To avoid any concerns about the particular nature of the items we chose for prior studies,

this study uses more general measures that directly tap self-presentation or usefulness.

We also further test whether the public nature of the communication can explain the

results. We hold the public nature of the sharing constant by looking at broadcasting and

narrowcasting within a purely public domain (i.e., Facebook).

Methods

One hundred and sixty-one respondents were randomly assigned to condition in a 2

(Audience size: narrow vs. broad) x 4 (Content Type) design, with audience size manipulated

between subjects and content manipulated within subjects.

First, we manipulated audience size. Participants were asked to imagine that they were

thinking of sharing content (e.g., a link or story) with all of their friends in a Facebook status


update (broadcasting condition) or with one of their friends in a Facebook wall post

(narrowcasting condition). These two conditions are equally public (both can be viewed by all

Facebook friends), but wall posts are more targeted (i.e., directed towards one person).10

Next, we manipulated the type of content being shared. Participants were asked how

likely (1 = “not at all” to 7 = “extremely”) they would be to share each of four types of content in

the situation described (order counterbalanced): (1) “Something that makes you look good,” (2)

“Something that makes you look bad,” (3) “Something that is useful to others,” and (4)

“Something that is not useful to others.” Self-presentation was assessed by the difference

between the first two items, while usefulness was assessed by the difference between the latter

two.

Finally, we measured sharer focus to test the proposed mechanism behind these effects.

Participants were asked “How much did you think about yourself?” and “How much did you

think about the [person/people] with whom you were sharing?” when deciding whether to share

the content (order counterbalanced, 1 = “did not think about it at all” to 7 = “thought about it a

lot,” differenced to form a measure of sharer focus, as in Study 2).

Results

Self-Presentation. Not surprisingly, a repeated-measures ANOVA on likelihood to share

revealed a main effect of Self-presentation: people were more willing to share content that made

them look good than content that made them look bad (F(1,141) = 163.09, p < .001). More

importantly, however, this was qualified by a significant Audience Size x Self-Presentation

interaction (F(1,141) = 23.97, p < .001). People always preferred to share content that made

them look good, rather than bad, but compared to narrowcasting (Mlook good = 5.37 vs. Mlook bad =

3.67, F(1,141) = 32.60, p < 0.001), broadcasting increased this tendency (Mlook good = 5.23 vs.

Mlook bad = 1.87, F(1,141) = 148.77, p < 0.001).

Looked at another way, and consistent with Study 3, relative to narrowcasting,

broadcasting decreased participants’ willingness to share content that makes them look bad

(F(1,141) = 49.25, p < .001). Audience size, however, did not significantly impact participants’

willingness to share content that makes them look good (F(1,141) < .3, p > .6).

10

A pretest (N = 40) confirmed that while status updates and wall posts are perceived to be similarly “public” (MStatus

Update = 5.18, MWall Post = 5.30, t(39) < .8, p > .4), wall posts are seen as more directed at particular audience members

than status updates (MStatus Update = 4.35, MWall Post = 5.58, t(39) = 3.81, p < .001).


Useful. Again, a repeated-measures ANOVA on likelihood to share revealed a main

effect of usefulness: people were more willing to share content that was useful to the recipient

than not useful (F(1,141) = 168.47, p < 0.001). More importantly, this was qualified by a

significant Audience Size x Usefulness interaction (F(1,141) = 6.94, p < .01). People always

preferred to share content that was useful (rather than not useful) to others, but compared to

broadcasting (Museful= 4.73 vs. Mnot useful= 2.59, F(1,141) = 51.03, p < 0.001) narrowcasting

increased this tendency (Museful= 5.48 vs. Mnot useful= 2.24, F(1,141) = 128.16, p < 0.001).

Looked at another way, narrowcasting increased participants willingness to share useful

content (F(1,141) = 5.99, p = .02). However, audience size did not have an effect on participants’

willingness to share content that was not useful (F(1,141) = 1.70, p = 0.19).

Mediation Analysis. To test the mediating role of sharer focus, we again used the

bootstrap mediation method (MEDIATE macro; Hayes, Preacher, and Myers 2011).

As predicted, sharer focus drove the impact of audience size on what people shared.

First, the effect of audience size on sharing self-presentational content (difference between

willingness to share things that make the self look good versus bad) was mediated by sharer

focus (total indirect effect = .30, standard error = .16, 95% CI [.65, .04]). Relative to

narrowcasting, broadcasting led participants to think more about the self (a = 1.26), which

increased their likelihood to share self-presentational content (b = .38, see Figure 6A).

Second, the effect of audience size on sharing useful content (difference between

willingness to share useful and not useful content) was also mediated by sharer focus (total

indirect effect = .40, standard error = .20, 95% CI [.07, .86]). Narrowcasting led participants to

think less about the self and more about the audience (a = -1.26), which increased their

likelihood to share content that is useful to others (b = -.37, see Figure 6B).

Discussion

Results of Study 4 underscore the findings of the first four studies and provide further

evidence for the underlying mechanism behind these effects.

First, audience size influenced what people shared. People were always more willing to

share content that made them look good rather than bad, but this tendency was stronger when

they were broadcasting (communicating in a Facebook status update) than narrowcasting

(communicating in a Facebook wall post). In addition, people were always more willing to share


content that was useful (as opposed to not useful) to the recipient, but this tendency was stronger

when people were narrowcasting than when they were broadcasting.

Second, these effects were driven by the degree to which individuals were thinking about

themselves versus others. Relative to narrowcasting, broadcasting led people to think more about

the self which, in turn, boosted the sharing of self-presentational content. On the other hand,

narrowcasting led people to think more about others which, in turn, boosted the sharing of useful

content.

In addition, by holding the public nature of the sharing constant (a Facebook wall post

versus a Facebook status update, both of which all Facebook friends can see), we again cast

doubt on the notion that differences in the public nature of communication are driving the

effects.

General Discussion

Whether face-to-face or online, all conversations involve communicating with an

audience. This audience may consist of just one person or it may consist of multiple people, but

does mere audience size impact what people pass on, and if so, how?

Five experiments demonstrated how audience size affects what people share.

Broadcasting does little to discourage people’s natural tendency to share self-presentational

content. Relative to narrowcasting, broadcasting led people to share things that made them look

better (Studies 1, 3b, and 4), use more positive language (Study 2), and reframe negative events

to make the self look less bad (Study 1). Narrowcasting, on the other hand, decreased the

tendency to share self-presentational content and led people to share more things that are useful

to their conversation partner (Studies 3a and 4).

The studies also demonstrate how sharer focus drives these effects. Narrowcasting

encourages people to think less about the self and more about others, which reduces the sharing

of self-presentational content and boosts the sharing of useful content. Our studies provide

support for this mechanism by establishing direct mediational evidence (Studies 2 and 4),

demonstrating differences in self- and other-focused pronoun usage (Study 2), and showing that

inducing other-focus leads to the same effects as narrowcasting (Studies 3a and 3b).


Finally, showing these effects across a wide range of manipulations and outcomes

illustrates their generalizability. The effects hold for willingness to share (Studies 3a, 3b, and 4),

written messages (Study 1), and actual communication with others (Study 2). In addition, they

persist regardless of whether writing is relatively constrained (Study 1) or uncontrolled (Study

2), and whether messages are human-coded (Study 1) or analyzed using more objective text

analysis (Study 2).

Theoretical Contributions and Implications

Our research makes several contributions. First, it sheds light on the drivers of word-of-

mouth and interpersonal communication. Recent research has begun to delve more deeply into

why people share (Berger and Milkman 2012; Berger and Schwartz 2011; Cheema and Kaikati

2010; Moldovan et al. 2011; Wojnicki and Godes 2011; Packard and Gershoff 2012; see Berger

2013 for a review), but people are just starting to address how the audience itself shapes what

people talk about. While some work has focused on audience type (e.g., strong vs. weak ties,

Frenzen and Nakamoto 1993), how number of followers influences posting frequency in social

media (Toubia and Stephen 2013), or how number of people in a focus group affects self-

disclosure (Corfman 1995), no prior work has investigated how mere audience size shapes the

type of content people share. We fill this gap, showing that whether people communicate with a

large or small audience can impact what they talk about by changing their focus.

Second, we advance prior psychological work on social influence and group behavior.

Social impact theory (Latane 1981) suggests that the force of social influence increases with a

greater number of sources (or others). But while Latane’s model bridged many diverse findings,

it does not demonstrate any unifying mechanism. In contrast, we demonstrate a specific

cognitive mechanism through which audience size impacts word of mouth. One-on-one

communication leads people to focus more on others and encourages them to share less self-

presentational and more useful content.

Similarly, Mullen’s theory on the other-total ratio in group behavior (1983) suggests that

as the number of others in a group increases, people become more concerned with standards of

appropriate behavior. While this work suggests that larger groups may increase self-attention,

our process is distinct. We demonstrate that people naturally focus on themselves and that


broadcasting does not shift people away from this default; rather, the action comes from

narrowcasting as a single individual can boost attention on others.

Third, this work provides insight into when the communication sender versus receiver

plays a relatively larger role in what people share. We examined the role of audience size, but

other factors should also play a role. For example, arousal encourages self-focus (Wegner and

Giuliano 1980), so it might increase the weight people put on their own interests and opinions

when deciding what to share. The goal of the communication may also matter. If people are

trying to sell a product, they may be particularly attentive to others’ needs, regardless of

audience size, while people may be predominantly self-focused when communicating in an

interview or on a date. Finally, there may be cultural differences: people from collectivist

societies may be less naturally self-focused than those from individualist societies.

Fourth, our work helps explain why social media posts (e.g., Facebook or Twitter) tend to

be self-focused. People often “brand” themselves through social media updates and personal

websites (Schau and Gilly 2003), and researchers have discussed the relationship between social

media and self-promotion (Buffardi and Campbell 2008). The computer-mediated (Walther

2007) and public (Ratner and Kahn 2002) nature of such communication should increase self-

presentational concerns, but our work identifies a third factor. Posting on social media often

involves a large audience: sharing comments with hundreds or thousands of “friends” or

“followers.” Our findings suggest that such broadcasting does little to shift people away from

their default tendency to focus on the self and thus may result in the disproportionate degree of

self-presentation in this medium.

This has important implications for consumer welfare. Because social media posts make

others’ lives look so fantastic, using social media can decrease wellbeing and make one’s own

life seem worse in comparison (Chou and Edge 2012). Our results suggest that broadcasting may

be unrepresentative of everyday life, and greater awareness of this bias may reduce the negative

inferences viewers make about the relative quality of their own lives.

More broadly, our results have important implications for interpersonal communication.

At an individual level, these findings can help consumers manage how they are perceived by

others. For instance, if a politician or public figure is concerned about sounding too negative, he

may want to schedule larger public events and fewer one-on-one appearances. In a business

context, it might be strategic for individuals to hit “reply” instead of “reply all” when responding


to group emails if they want to be less self-focused and more in tune with their audience. Or, if

replying to the full group is necessary, listing each person’s name (instead of a general group

salutation) might be an effective tactic for inducing other-focus and encouraging one to be more

useful.

Our results also have implications for encouraging word-of-mouth. Companies that sell

useful products (e.g., healthcare) may generate greater word-of-mouth by providing web forms

that allow for narrow, personalized messages. Conversely, companies that sell products related to

self-presentation (e.g., designer clothing) may increase sharing by facilitating easy broadcasting

(e.g., one-click posting on social media).

Directions for Future Research

As with any preliminary investigation, there is always more to explore. Consistent with

prior literature (Baumeister et al. 2001; Tice 1991), we find stronger effects on protective than

acquisitive self-presentation. While the word of mouth literature has not considered this

distinction previously, it deserves further study. Lay intuition might suggest that acquisitive

presentation occurs more frequently, but this may be due to biases in attention and memory. It is

much harder to notice the absence of negative things than the presence of bragging. Further, the

few word of mouth findings that can be categorized as acquisitive presentation in retrospect (De

Angelis et al. 2012; Wojnicki and Godes 2011) all involve explicit needs to bolster self-

enhancement. Thus while such situations may lead people to say positive things about a chosen

experience (e.g., everyone liked the restaurant I picked) to look smart, it is less clear whether

these effects apply more broadly.

One reason broadcasting may not have as strong an influence on acquisitive self-

presentation is concern about implausibility (Tice et al. 1995). Prior research shows that people

are less likely to try and claim a desirable image when there is more potential to be caught

“faking” it (Packard and Gershoff 2012). While people can tailor their message to avoid being

caught exaggerating their positive traits when talking to one person, this may become more

difficult when talking to a large group, as different people may have access to different

information. Similarly, while people can tailor their message to one person to avoid being seen

as arrogant, concerns about coming off this way (Schlenker and Leary 1982) might also dampen


bragging to larger audiences. More generally, future research might consider under what

conditions broadcasting may increase acquisitive self-presentation.

Future research might also examine if people select audiences of different sizes

depending on what they want to talk about. This paper demonstrates that manipulating audience

size has a causal impact on what consumers share. But in situations where the audience size is

not fixed, consumers may also select how many people to talk to based on the content of their

communication.

It would also be interesting to consider other ways in which audience size impacts

sharing beyond self- and other-focus. For instance, larger audiences are more likely to involve

heterogeneous groups, and may thus elicit more cautious opinions. Similarly, people who

communicate in “multiple-audience” contexts are more likely to acknowledge multiple

viewpoints and present more than one side of an argument (Schlosser 2005).

Moreover, consistent with work showing the extraordinary power of a single individual

relative to any number of multiple people (Latane 1981; Jenni and Loewenstein 1997), we define

audience size in two distinct categories (broadcasting versus narrowcasting). However, we

recognize that the influence of audience size may be more of a continuum. Consequently, the

self- versus other-focus trade-off may be more fluid, with people gradually thinking more about

others as audience size decreases. Similarly, to the extent that one can induce other-focus when

communicating with a group, broadcasting may look more like narrowcasting. For example, in

addition to the listing manipulation from Study 3, describing a large audience as consisting of

similar individuals, or as an entitative unit (Smith, Faro, and Burson 2013), could make people

share more useful content and less self-presentational content.

In conclusion, the current research demonstrates one way in which audience size affects

interpersonal communication. By integrating work on the drivers of word-of-mouth with work

on self-other trade-offs, this deepens our understanding of what people share and why.


REFERENCES

Alicke Mark D. and Olesya Govorun (2005), “The better-than-average effect,” in The self in

social judgment, eds. Mark D. Alicke, David A. Dunning, and Joachim I. Krueger, New

York: Psychology Press, p. 85–106.

Alicke, Mark D., M. L. Klotz, David L. Breitenbecher, Tricia Yurak, Debbie S. Vredenburg

(1995), “Personal contact, individuation, and the better-than-average effect” Journal of

Personality and Social Psychology, 68(5), 804-825.

Apperly, Ian A., Kevin J. Riggs, Andrew Simpson, Claudia Chiavarino, and Dana Samson

(2006), “Is belief reasoning automatic?” Psychological Science, 17, 841–844.

Arkin, Robert (1981), “Self-presentation styles,” in Impression management and social

psychological research, ed. J.T. Tedeschi, New York: Academic Press.

Bales, Robert F. (1951), Interaction Process Analysis: a method for the study of small groups,

Cambridge: Addison-Wesley.

Baumeister, Roy F., Ellen Bratslavsky, Catrin Finkenauer, and Kathleen D. Vohs (2001), "Bad is

stronger than good," Review of general psychology, 5, 323-370.

Bell, Paul A. "Affective state, attraction, and affiliation: Misery loves happy company, too."

Personality and Social Psychology Bulletin 4, no. 4 (1978): 616-619.

Berger, Jonah (2011), “Arousal Increases Social Transmission of Information,” Psychological

Science, 22 (7), 891–93.

Berger, Jonah (2013), Contagious: Why Things Catch On, New York: Simon & Schuster.

Berger, Jonah and Katherine Milkman (2012), “What Makes Online Content Viral?” Journal of

Marketing, 49 (2), 192-205.

Berger, Jonah and Eric Schwartz (2011), “What Drives Immediate and Ongoing Word-of-

Mouth?” Journal of Marketing Research, 48 (5), 869-880.

Brooks, Alison Wood, and Maurice E. Schweitzer (2011), "Can Nervous Nelly negotiate? How

anxiety causes negotiators to make low first offers, exit early, and earn less profit,"

Organizational Behavior and Human Decision Processes, 115 (1), 43-54.

Brown, Jacqueline Johnson and Peter H. Reingen (1987), “Social Ties and Word-of-Mouth

Referral Behavior,” Journal of Consumer Research, 14 (3), 350–62.

Brown, Jonathon D., Rebecca L. Collins, and Greg W. Schmidt (1988), “Self-Esteem and Direct

Versus Indirect Forms of Self-Enhancement,” Journal of Personality and Social

Psychology, 55 (3), 445–53.


http://psycnet.apa.org/index.cfm?fa=search.searchResults&latSearchType=a&term=Alicke,%20Mark%20D.

http://psycnet.apa.org/index.cfm?fa=search.searchResults&latSearchType=a&term=Klotz,%20M.%20L.

http://psycnet.apa.org/index.cfm?fa=search.searchResults&latSearchType=a&term=Breitenbecher,%20David%20L.

https://marketing.wharton.upenn.edu/files/?whdmsaction=public:main.file&fileID=3464

https://marketing.wharton.upenn.edu/files/?whdmsaction=public:main.file&fileID=3464

Buffardi, Laura E. and W. Keith Campbell (2008), “Narcissism and Social Networking Web

Sites,” Personality and Social Psychology Bulletin, 34 (10), 1303-1314.

Chambers, John R. and Paul D. Windschitl (2004), “Biases in social comparative judgments: The

role of nonmotivated factors in above-average and comparative-optimism effects,”

Psychological Bulletin, 130, 813–838.

Cheema, Amar and Andrew M. Kaikati (2010), “The Effect of Need for Uniqueness on Word-of-

Mouth,” Journal of Marketing Research, 47 (3), 553–563.

Chevalier, Judith A. and Dina Mayzlin (2006), “The Effect of Word of Mouth on Sales: Online

Book Reviews,” Journal of Marketing Research, 43, 345-354.

Chintagunta, Pradeep K., Shyam Gopinath and Sriram Venkataraman (2010), “The Effects of

Online User Reviews on Movie Box-Office Performance: Accounting for Sequential

Rollout and Aggregation Across Local Markets,” Marketing Science, 29, 944-957.

Chiu, Yi, Jin-Chern Chiou, Weileun Fang , Yung-Jiun Lin and Mingching Wu (2007), “Design,

Fabrication, and Control of Components in MEMS-Based Optical Pickups,” IEEE

Transactions in Magnetics, 43 (2), 780-785.

Chou, Hui-Tzu Grace and Nicholas Edge (2012), “’They Are Happier and Having Better Lives

than I am’: The Impact of Using Facebook on Perceptions of Others’ Lives,”

Cyberpsychology, Behavior, and Social Networking, 15 (2), 117-121.

Cialdini, Robert B., Richard J. Borden, Avril Thorne, Marchus Randall Walker, Stephen

Freeman, and Lloyd Reynolds Sloan (1976), “Basking in reflected glory: Three (football)

field studies,” Journal of Personality and Social Psychology, 34, 366-375.

Clark, Herbert H. and Edward F. Schaefer (1989), “Contributing to discourse,” Cognitive

Science, 13, 259–294.

Corfman, Kim P. (1995),"The Importance of Member Homogeneity to Focus Group Quality", in

Advances in Consumer Research, Vol. 22, eds. Frank R. Kardes and Mita Sujan, Provo,

UT: Association for Consumer Research, p. 354-359.

Davis, Deborah, and Timothy C. Brock (1975), "Use of first person pronouns as a function of

increased objective self-awareness and performance feedback," Journal of Experimental

Social Psychology, 11 (4), 381-388.

De Angelis, Matteo, Andrea Bonezzi, Alessandro Peluso, Derek Rucker and Michele Costabile

(2012), “On Braggarts and Gossips: A Self-Enhancement Account of Word-of-Mouth

Generation and Transmission,” Journal of Marketing Research, 49 (4), 551-563.

Dichter, E., (1966), “How Word of Mouth Advertising Works,” Harvard Business Review, 44,

147-166.


Dunbar, R. I. M., Anna Marriott, and N. D. C. Duncan (1997), "Human conversational

behavior," Human Nature, 8 (3), 231-246.

Dunning, David, Leaf Van Boven, and George Loewenstein (2001), “Egocentric empathy gaps

in social interaction and exchange,” in Advances in Group Processes, Vol. 18, eds.

Edward Lawler, Michael W. Macey, Shane R. Thye, and Henry A. Walker, Oxford,

England: Elsevier Science, p. 65–97.

Emler, Nick (1990), “A social psychology of reputation,” European Review of Social

Psychology, 1, 171–193.

Engel, James F., Roger D. Blackwell, and Robert J. Kegerreis (1969) "How Information Is Used

to Adopt an Innovation," Journal of Advertising Research , 9 (4), 3-8.

Epley, Nicholas, Boaz Keysar, Leaf Van Boven, and Thomas Gilovich (2004), “Perspective

taking as egocentric anchoring and adjustment,” Journal of Personality and Social

Psychology, 87, 327–339.

Fiske, Susan T. (2001), “Social and Societal Pragmatism: Commentary on Augustinos, Gaskell,

and Lorenzi-Cioldi,” in Representations of the social: Bridging research traditions, eds.

Kay Deaux and Gina Philogene, New York: Blackwell, p. 249-253.

Frenzen, Jonathan and Kent Nakamoto (1993), “Structure, Cooperation, and the Flow of Market

Information,” Journal of Consumer Research, 20, 360-375.

Fussell, Susan R. and Robert M. Krauss (1989), “The effects of intended audience on message

production and comprehension: Reference in a common ground framework,” Journal of

Experimental Social Psychology, 25, 203–219.

Godes, David and Dina Mayzlin (2004), “Using Online Conversations to Study Word-of- Mouth

Communication,” Marketing Science, 23 (4), 545–60.

Goffman, Erving (1959), The Presentation of Self in Everyday Life, New York: Anchor.

Gonzales, Amy L., and Jeffrey T. Hancock, (2011), "Mirror, mirror on my Facebook wall:

Effects of exposure to Facebook on self-esteem," Cyberpsychology, Behavior, and Social

Networking, 14 (1), 79-83.

Granovetter, Mark S. (1973), “The Strength of Weak Ties,” American Journal of Sociology, 78

(6), 1360-80.

Grice, H. Paul (1975), “Logic and conversation,” in Syntax and semantics, Vol. 3, eds. Peter

Cole and Jerry L. Morgan, New York: Academic Press, 41-58.


Hamilton, David L. and Steven J. Sherman, (1996), "Perceiving persons and groups,"

Psychological Review, 103 (2), 336-355.

Hastie, Reid, and Bernadette Park, (2005), "The relationship between memory and judgment

depends on whether the judgment task is memory-based or on-line," Psychological

Review, 93, 258-268.

Hayes, Andrew F., Kris J. Preacher and Teresa A. Myers (2011), “Mediation and the Estimation

of Indirect Effects in Political Communication Research,” in The Sourcebook for

Political Communication Research: Methods, Measures, and Analytical Techniques, ed.

E.P Bucy and R. Lance Holbert, New York: Routledge, p. 434–465.

Hennig-Thurau, Thorsten, Kevin P. Gwinner, Gianfranco Walsh and Dwayne D. Gremler,

(2004), “Electronic Word‐of‐Mouth Via Consumer‐Opinion Platforms: What Motivates

Consumers to Articulate Themselves on the Internet?” Journal of Interactive Marketing,

18 (1) 38-52.

Higgins, E. Tory (1999), ““Saying is believing” effects: When sharing reality about something

biases knowledge and evaluations,” in Shared cognition in organizations: The

management of knowledge, ed. Leigh L. Thompson, John M. Levine, and David M.

Messick, Mahwah, NJ: Erlbaum, p. 33–49.

Hoorens, Vera (1996), "Self-favoring biases for positive and negative characteristics:

Independent phenomena?" Journal of Social and Clinical Psychology, 15, 53-67.

Ickes, William, Susan Reidhead, and Miles Patterson (1986), “Machiavellianism and self-

monitoring: As different as ‘me’ and ‘you,’” Social Cognition, 4, 58-74.

Iyengar, Raghuram, Christophe Van den Bulte, and Thomas W. Valente (2011), “Opinion

Leadership and Social Contagion in New Product Diffusion,” Marketing Science, 30 (2),

195-212.

Jenni, Karen and George Loewenstein (1997), “Explaining the Identifiable Victim Effect,”

Journal of Risk and Uncertainty, 14 (3), 235-257.

Kelly, Kristine M., and Warren H. Jones (1997), "Assessment of dispositional embarrassability,"

Anxiety, stress, and coping, 10 (4), 307-333.

Klein, William M. (1992), “Constructing social reality: Greater bias for negative than for

positive behaviors?” paper presented at the annual meeting of the Eastern Psychological

Association, Boston, MA.

Kruger, Justin (1999), “Lake Wobegon be gone! The ‘below-average effect’ and the egocentric

nature of comparative ability judgments,” Journal of Personality and Social Psychology,

77, 221–232.


https://marketing.wharton.upenn.edu/profile/199/

https://marketing.wharton.upenn.edu/profile/201/

Landis, M. H., and Harold E. Burtt (1924), "A Study of Conversations," Journal of Comparative

Psychology, 4 (1), 81-89.

Latane, Bibb (1981), "The psychology of social impact,” American Psychologist, 36 (4), 343-

356.

LeRouge, Davy and Luk Warlop (2006),"Why It Is So Hard to Predict Our Partner's Product

Preferences: The Effect of Target Familiarity on Prediction Accuracy," Journal of

Consumer Research, 33 (3), 393-402.

Locke, Kenneth D., and Leonard M. Horowitz (1990), "Satisfaction in interpersonal interactions

as a function of similarity in level of dysphoria," Journal of Personality and Social

Psychology, 58 (5), 823-831.

McGuire, William J. (1985), "Attitudes and attitude change,” in The Handbook of Social

Psychology, ed. G. Lindzey and E. Aronson, New York: Random House, p. 233-346.

Moldovan, Sarit, Jacob Goldenberg, and Amitava Chattopadhyay (2011), “The Different Roles

of Product Originality and Usefulness in Generating Word of Mouth,” International

Journal of Research in Marketing, 28 (2), 109-119.

Moore, Sarah (2012), “Some things are better left unsaid: how word of mouth influences the

storyteller,” Journal of Consumer Research, 38 (6), 1140-1154.

Mullen, Brian (1983), "Operationalizing the effect of the group on the individual: A self-

attention perspective," Journal of Experimental Social Psychology, 19 (4), 295-322.

Naaman, Mor, Jeffrey Boase and Chih-Hui Lai (2010), “Is It Really About Me?: Message

Content in Social Awareness Streams,” in Proceedings of the 2010 ACM Conference on

Computer Supported Cooperative Work, Savannah, GA, 189–192.

Ogilvie, Daniel M. (1987), "The undesired self: A neglected variable in personality research,"

Journal of Personality and Social Psychology, 52, 379-385.

Packard, Grant and Andrew Gershoff (2012), “Caught Red Branded: The Social Risk of Self-

Presentational Props,” University of Michigan working paper.

Packard, Grant and David B. Wooten (2012), “Compensatory Communication: Consumer

Knowledge Discrepancies and Knowledge Signaling in Word-of-Mouth,” University of

Michigan working paper.

Paulhus, Delroy L., Peter Graf, and Mark Van Selst (1989), “Attentional load increases the

positivity of self-presentation," Social cognition, 7 (4), 389-400.

Paulhus, Delroy L. and Karen Levitt (1987), “Desirable Responding Triggered by Affect:

Automatic Egotism?” Journal of Personality and Social Psychology, 52 (2), 245-259.


http://ideas.repec.org/a/ucp/jconrs/v33y2006i3p393-402.html

http://ideas.repec.org/a/ucp/jconrs/v33y2006i3p393-402.html

http://ideas.repec.org/s/ucp/jconrs.html

http://ideas.repec.org/s/ucp/jconrs.html

Pennebaker, James W., Martha E. Francis, and Roger J. Booth (2001), "Linguistic inquiry and

word count: LIWC 2001," Mahway: Lawrence Erlbaum Associates.

Piaget, Jean (1926), The language and thought of the child. London: Routledge & Regan Paul.

Pylyshyn, Zenon W. (1978), “When is attribution of beliefs justified?” Behavioral and Brain

Sciences, 1, 592-593.

Ratner, Rebecca K. and Barbara E. Kahn (2002), “The impact of private versus public

consumption on variety-seeking behavior,” Journal of Consumer Research, 29, 246-257.

Richins, Marsha L. (1984), “Word of Mouth Communications as Negative Information,” in

Advances in Consumer Research, Vol. 11, ed. Thomas C. Kinnear, Provo, UT :

Association for Consumer Research, 697–702.

Robertson, Kirsten, and Antony Doig (2010), "An Empirical Investigation of Variations in Real‐Estate Marketing Language over a Market Cycle," Housing, Theory and Society, 27 (2),

178-189.

Ross, Michael and Fiore Sicoly (1979), “Egocentric biases in availability and attribution,”

Journal of Personality and Social Psychology, 37, 322-336.

Ross, Lee, and Andrew Ward (1996) "Naive realism in everyday life: Implications for social

conflict and misunderstanding," in Values and knowledge, eds. Edward S. Reed, Elliot

Turiel, and Terrance Brown, Mahway, New Jersey: Lawrence Erlbaum Associates, Inc.,

p. 103-135.

Schau, Hope Jensen, and Mary C. Gilly (2003), "We are what we post? Self‐presentation in

personal web space," Journal of Consumer Research, 30 (3), 385-404.

Schlenker, Barry R., and Mark R. Leary (1982), "Audiences' reactions to self-enhancing, self-

denigrating, and accurate self-presentations," Journal of experimental social psychology,

18, 89-104.

Schlosser, Ann E. (2005), “Posting versus Lurking: Communicating in a Multiple Audience

Context,” Journal of Consumer Research, 32, 260-265.

Sedikides, Constantine (1993), “Assessment, Enhancement, and Verification Determinants of the

Self-Evaluation Process,” Journal of Personality and Social Psychology, 65 (2), 317–38.

Sedikides, Constantine and Michael J. Strube (1995), “The Multiply Motivated Self,”

Personality and Social Psychology Bulletin, 21 (12), 1330–35.


Smith, Robert W., David Faro, and Katherine A. Burson (2013), "More for the many: The

influence of entitativity on charitable giving," Journal of Consumer Research, 39, 961-

976.

Sundaram, D.S., Kaushik Mitra and Cynthia Webster (1998), “Word-of-Mouth Communications:

A Motivational Analysis,” Advances in Consumer Research, 25, 527-531.

Tamir, Diana I. and Jason P. Mitchell (2012), “Disclosing Information About the Self Is

Intrinsically Rewarding,” Proceedings of the National Academy of Sciences, 109 (21),

8038-8043.

Tausczik, Yla R. and James W. Pennebaker (2010). “The Psychological Meaning of Words:

LIWC and Computerized Text Analysis Methods,” Journal of Language and Social

Psychology, 29 (1), 24-54.

Taylor, Shelley E. and Suzanne C. Thompson (1982), "Stalking the Elusive 'Vividness' Effect,"

Psychological Review, 89 (2), 155-181.

Tice, Dianne M. (1991), "Esteem protection or enhancement? Self-handicapping motives and

attributions differ by trait self-esteem," Journal of Personality and Social Psychology, 60,

711-725.

Tice, Dianne M., Jennifer L. Butler, Mark B. Muraven, and Arlene M. Stillwell (1995), “When

modesty prevails: Differential favorability of self-presentation to friends and strangers,”

Journal of Personality and Social Psychology, 69 (6), 1120-1138.

Toubia, Olivier and Andrew T. Stephen (2013), "Intrinsic vs. Image-Related Utility in Social

Media: Why Do People Contribute Content to Twitter?" Marketing Science, forthcoming.

Trusov, Michael, Randolph E. Bucklin and Koen Pauwels (2009), “Effects of Word-of- Mouth

Versus Traditional Marketing: Findings from an Internet Social Networking Site,”

Journal of Marketing, 73 (5), 90-102.

Van Boven, Leaf, George Loewenstein, and David Dunning (2005), “The illusion of courage in

social predictions: Underestimating the impact of fear of embarrassment on other

people,” Organizational Behavior and Human Decision Processes, 96, 130–141.

Van den Bulte, Christophe and Stefan Wuyts (2009), “Leveraging Customer Networks,” in The

Network Challenge: Strategy, Profit and Risk in an Interlinked World, eds. Jerry (Yoram)

Wind and Paul R. Kleindorfer, Upper Saddle River, New Jersey: Wharton School

Publishing, p. 243- 258.

Walther, Joseph B. (2007), "Selective self-presentation in computer-mediated communication:

Hyperpersonal dimensions of technology, language, and cognition," Computers in

Human Behavior, 23 (5), 2538-2557.


Wegner, Daniel M. and Tony Giuliano (1980), “Arousal-induced Attention to Self,” Journal of

Personality and Social Psychology, 38, 719–726.

Weinstein, Neil D. (1983), “Reducing unrealistic optimism about illness susceptibility,” Health

Psychology, 2, 11-20.

Wellman, Henry M. (1988), “First steps in the child’s theorizing about the mind,” in Developing

theories of mind, eds. Janet W. Astington, Paul L. Harris, and David R. Olson,

Cambridge: Cambridge University Press, p. 64–92.

Wojnicki, Andrea C. and David Godes (2011), “Word-of-Mouth as Self-Enhancement,” HBS

Marketing Research Paper No. 06-01.

Zhang, Yan, and Nicholas Epley (2012), "Exaggerated, Mispredicted, and Misplaced: When “It's

the Thought That Counts” in Gift Exchanges," Journal of Experimental Psychology, 141

(4), 667-681.


FIGURE 1: RELATIVE TO NARROWCASTING, BROADCASTING DECREASES USE OF

NEGATIVE EMOTION WORDS (STUDY 2)

0

0.5

1

1.5

2

Narrowcasting Broadcasting Control(no audience)

Ne

gati

ve e

mo

tio

n w

ord

s


FIGURE 2: NARROWCASTING SHIFTS FOCUS AWAY FROM THE SELF AND TOWARDS

OTHERS (STUDY 2)


FIGURE 3: SHARER FOCUS MEDIATES THE IMPACT OF AUDIENCE SIZE ON VALENCE

OF EXPRESSION (STUDY 2)

Note: ns p > .10; * p < .05; ** p < .01; *** p < .001. Mediation run using the Bootstrap method

with 1,000 samples (Hayes, Preacher and Myers 2011). The total indirect effect is significant,

with a 95% confidence interval of [.01, .60] and a standard error of 0.16.

β=.09*

β=1.94**

Audience size

Sharer focus

Positive vs.

negative emotion

words β=1.19** (β=-.96*)


FIGURE 4: LISTING AUDIENCE MEMBER(S) MODERATES THE IMPACT OF AUDIENCE

SIZE ON SHARING USEFUL ITEMS (STUDY 3a)

3.5

3.7

3.9

4.1

4.3

4.5

4.7

Narrowcasting Broadcasting

Will

ingn

ess

to s

har

e u

sefu

l ite

ms

Control

Other-focus(Listing)


FIGURE 5A: LISTING AUDIENCE MEMBER(S) MODERATES THE IMPACT OF AUDIENCE

SIZE ON SHARING ITEMS THAT MAKE THE SELF LOOK BAD (STUDY 3b)

FIGURE 5B: LISTING AUDIENCE MEMBER(S) HAS NO EFFECT ON THE IMPACT OF

AUDIENCE SIZE ON SHARING ITEMS THAT MAKE THE SELF LOOK GOOD (STUDY 3b)

1

1.5

2

2.5

3

3.5

4

4.5

5

5.5

6


Will

ingn

ess

to s

har

e it

ems

that

m

ake

the

self

loo

k b

ad

Control


1

1.5

2

2.5

3

3.5

4

4.5

5

5.5

6


Will

ingn

ess

to s

har

e it

ems

that

m

ake

the

self

loo

k go

od

Control



FIGURE 6A: SHARER FOCUS MEDIATES THE IMPACT OF AUDIENCE SIZE ON SHARING

SELF-PRESENTATIONAL CONTENT (STUDY 4)




β=.38***

β=1.26***

Audience size

Sharer focus

Self-presentational

content β=1.94*** (β=1.64***)


FIGURE 6B: SHARER FOCUS MEDIATES THE IMPACT OF AUDIENCE SIZE ON SHARING

USEFUL CONTENT (STUDY 4)




β=-.37***

β=-1.26***

Audience size

Sharer focus

Useful content

β=1.09** (β=.70 ns)


APPENDIX

EXAMPLES OF PARAGRAPHS WRITTEN BY PARTICIPANTS IN STUDY 1

Broadcasting

(Reframe Score = 5, Number of positive events minus number of negative events = 2)

Hey guys! I had a great weekend! I went with a couple of friends to see Iron Man 3. It was

PHENOMENAL. I really really enjoyed it! I thought it was way better than the second movie. It

was very different from the previous two movies in that it had a different focus. Regardless, it

was great. I also got popcorn, but it was really salty. / However, I did oversleep on Saturday and

missed my favorite morning show. I guess I was really tired- more tired than I thought. Oh well,

I can just watch a rerun on the internet! I sent the boyfriend this youtube video of a screaming

goat that I thought was hilarious. It was a mashup of Taylor Swift's "I knew you were trouble"

video and a goat that screamed liked a human. He thought it was funny too. Oh and I baked a

cheesecake! The boyfriend and I made a cheesecake to bring over to a friend's house and it

turned out better than we expected. The middle was a little soft, but it was still pretty yummy.

Talk to you guys later!

Narrowcasting

(Reframe Score = 2, Number of positive events minus number of negative events = -1)

I hope this email finds you well. I guess I'll start by telling you guys about the interesting day I

had yesterday. So, the night before I went out to the Irish Pub in my neighborhood and had a ton

of beer with some of my fraternity brothers. But they were out of jagermeister. So it wasn't a

shock that I woke up the next morning at 10 oclock; an hour after Mike and Mike in the

morning! I don't have to tell you guys how much I love that show. So my day had a rocky start.

After a brief meeting with my mentor, which I was late for by the way, I met up Charlize to go

see a movie. The movie was great and Charlene was even better. After the movies I took her to

the Cheesecake Factory for some dessert but they were closed and we had to settle for a Hot n

Crusty around the corner. Womp Womp! Anyways, I dropped Charlize off at home and went

home to work on the video. / / On my way home this guy complimented me on my favorite

shirt. You know? The one you guys always laugh at me for wearing. He said the shirt was

swanky and that I had a unique sense of style. So take that! But back to the video, when I got

home I made the video in like 2 hours of pure creative bliss.


broadcasting and narrowcasting: how audience size …

Documents