Chairs: Clayton R. Critcher, University of California, Berkeley, USA Jonah Berger, University of Pennsylvania, USA
Paper #1: The Psychology of Humblebragging Ovul Sezer, Harvard Business School, USA Francesca Gino, Harvard Business School, USA Michael I. Norton, Harvard Business School, USA
Paper #2: Posting Posed, Choosing Candid: Photo Posters Mispredict Audience Preferences
Jonah Berger, University of Pennsylvania, USA Alixandra Barasch, University of Pennsylvania, USA
Paper #3: Passing (On) Judgment: Others Judge Us Less Extremely Than We Think
Alice Moon, University of California, USA Muping Gan, University of California, USA Clayton R. Critcher, University of California, USA
Paper #4: First Impressions and Consumer Mate Preferences in Online Dating and Speed-Dating
Consumers make decisions not merely to please themselves, but to prompt pleasing impressions in others. Saving up money for a designer purse, editing (and reediting) a witty tweet, or trying a couple dozen times to capture the perfect selfie all reflect the lengths consumers will go to in order to be viewed well by others. Although much previous research has examined how the motivation to im-pression manage influences consumer behavior, comparatively little research has examined whether consumers truly know what impres-sions they have encouraged. In this session, the papers converge in addressing a series of related questions: Do consumers realize how their actions are viewed by others? Might consumers’ attempts to manage impressions ironically backfire? If consumers often err in these ways, why is this the case, and what actually does drive such social impressions?
The first two papers identify two ways that consumers fail at strategic impression management. Sezer, Gino, and Norton examine a sociocultural phenomenon that has been much discussed, but little researched: the humblebrag. Instead of directly touting one’s own achievements (e.g., “Just published in JCR!”), people will hide them within a supposedly negative pronouncement (e.g., “so embarrass-ing…my phone keeps ringing and bothering everyone—reporters always call when I’ve just published in JCR!”). But given one’s ul-terior motive is so transparent, humblebrags can prompt more nega-tive impressions than would more straightforward self-promotion.
Berger and Barasch examine a failure of impression management in a different domain—on-line profile pictures. Although users of social networking and dating websites tend to post posed photos, those who peruse such profiles draw more positive inferences from candid pic-tures. Much like humblebrags, posed photos fail because they come across as contrived, disingenuous, and insincere. Consistent with
this account, candid photos no longer prompted an impression boost when they were known to be staged.
The final two papers examine impression management that is less deliberate, but that still reflects a disconnect between the sig-nals consumers think they send and what others infer. In some cases, events simply befall consumers, who may then try to guess how others’ impressions react. Moon, Gan, and Critcher show that focal evaluative episodes (e.g., winning or losing at a game show, being accepted or rejected by a peer) loom larger in the actor’s mind than in social observers’. Although actors think they will then be judged as a genius or a dunce, a social favorite or an outcast, observers tend to see less information in these events. Olivola and colleagues comple-ment these findings by examining situations in which observers pick up on a surprising amount of signal. In particular, they show that dating profile pictures convey not merely looks, but also personal-ity. The latter is an important incrementally-valid predictor of the behavioral interest people receive on the dating market. Moreover, that so many consumer post photos conveying negative personality characteristics suggests they fail to appreciate these effects.
In total, session attendees will leave with a stronger understand-ing of what drives social impressions of consumers and consumers’
(mis)estimates of those impressions.
The Psychology of Humblebragging EXTENDED ABSTRACT
People are frequently confronted with a new type of self-praise:
“Hair’s not done, just rolled out of bed from a nap and still get hit on, so confusing!” “Two master degrees and I can’t make a tablet work!”
Such statements convey a complaint, an expression of dissatisfac-tion, or an unfavorable attitude (Alicke et al., 1992; Kowalski, 1996, 2002), yet simultaneously allow speakers to highlight their wonder-ful qualities. This kind of humblebragging—masking success in the guise of complaint—is increasingly common.
In this paper, we explore the psychology of humblebragging.
Although humblebragging is ubiquitous, our results suggest that humblebragging fails to leave a favorable impression, and can even be worse than simply bragging—because people do not find it sin-cere.
Why do people brag in the first place? An extensive body of research has documented that people are motivated to manage the impressions that others form of them (Jones & Pittman, 1982; Leary
& Kowalski, 1990; Schlenker, 1980). People desire to be viewed positively (Baumeister, 1982; Jones & Wortman; 1973; Schlenker
& Weigold, 1992) and pay attention to how they present themselves in any given social interaction (Goffman, 1959). A commonly used impression management strategy is self-promotion, which allows in-dividuals to bring their good qualities to others’ attention (Tedeschi
& Norman, 1985).
Certainly, successful impression management requires a bal-ance. Because modesty is also a highly valued quality (Eagly &
Acksen, 1971; Wosinka, Dabul, Whetstone-Dion & Cialdini, 1996), efforts to self-promote can backfire, causing the individual to be per-ceived as conceited or a braggart (Powers & Zuroff, 1988; Schlenker
& Leary, 1982). Thus, people often seek to present their qualities and accomplishments indirectly (Schleinker & Weigold, 1992). For
instance, individuals may glorify accomplishments of their associ-ates (Cialdini, Finch & DeNicholas, 1990), or may make modest at-tributions by giving credit to others (Tetlock, 1980). We suggest that humblebragging is an understudied yet ubiquitous indirect strategy that attempts to mask a brag in the guise of a complaint: to brag without seeming to be bragging.
Yet, humblebragging may not be an effective indirect way of self-promotion. Prior research has shown that the success of an im-pression management strategy depends on whether the person is able to hide his ulterior motive to be viewed positively (Giacalone
& Rosenfeld, 1986; Jones & Pittman, 1982). If the desire to make a favorable impression seems to be the main goal of the person, the person’s attempt to be viewed positively backfires, given people do not perceive the person to be genuine or sincere (Nguyen, Sears &
Hartman, 2008; Turnley & Bolino; 2001). In short, perceived sincer-ity is one of the key ingredients that determine whether the individu-als are successful in their attempt to be seen in a favorable light (Bo-lino, 1999; Eastman, 1994). Given these findings, we predicted that humblebragging would fail to leave a favorable impression, because masking bragging in complaint would seem insincere.
We started our exploration of this phenomenon by looking at social media. Prior research has shown that individuals share content online to communicate desired impressions (Dellarocas, 2003; Kol-lock, 1999; Lampel & Bhalla, 2007). This desire to signal favorable aspects of the self has given rise to a wide array of strategies for constructing a positive image online (Schau & Gilly, 2003). Thus, we started our investigation of humblebragging by analyzing a da-taset of statements that are categorized as “humblebrags” on Twit-ter (Wittels, 2012). In Study 1, two independent raTwit-ters judged 740 tweets from https://twitter.com/humblebrag on liking, perceived sin-cerity and perceived competence, and the extent to which they felt the person was humblebragging. All three metrics were negatively correlated with ratings of humblebragging, offering initial evidence for the danger of humblebragging.
In Study 2, we compared the effectiveness of humblebragging to “regular” bragging. In Study 2a, two hundred and one participants viewed either brags (e.g., “I have got so many internship offers.” or humblebrags (e.g “So many internship offers give me a headache.”), and rated how much they liked the target, how competent the target seemed, and how socially attractive they found the target to be. In addition, they rated the target for their level of sincerity. Not only are humblebraggers liked less and seen as less competent and socially unattractive than braggers, but they are also seen as less sincere than braggers. In Study 2b, two hundred and one participants evaluated either a brag (“I get hit on all the time.”) or a humblebrag (“Just rolled out of a bed and still get hit on all the time”). Again, humble-braggers are liked less and seen as less sincere than humble-braggers, and also found to be less attractive than braggers. Humblebragging leads people to infer that the target possesses the trait less than regular bragging. In short, people do not like braggers, but they acknowl-edge that at least braggers are being straightforward. The effort to mask bragging in complaint is seen as disingenuous; moreover, these decreased ratings of sincerity are what drive lower levels of liking.
In Study 3, we examine how humblebragging compares to com-plaining. Prior research suggests that complaining can be used as a self-presentation tool (Kowalski, 1996). Individuals may complain to induce similarity and liking (Brehm, 1992) or to signal specific knowledge and selectivity (Alicke et al., 1992). Because bragging, complaining and humblebragging are all used in the service of creat-ing desired impressions, we tested the relative efficacy of the three.
In Study 3, three hundred and two participants rated targets who engaged in complaining (e.g “I am so bored.”), bragging (e.g “People
mistake me for a model.”), or humblebragging (e.g “I am so bored of people mistaking me for a model.”) As in Study 2, targets who engaged in straightforward bragging were liked more than those who humblebragged; in addition, targets who engaged in straightforward complaining were also liked more than those who humblebragged.
While people do not love braggers or complainers, they at least see them as more sincere than humblebraggers, and these ratings of sin-cerity drive liking ratings.
Posting Posed, Choosing Candid: Photo Posters Mispredict Audience Preferences
EXTENDED ABSTRACT
Photos are a ubiquitous channel of social communication. They are used to make inferences about the personalities of others and to facilitate decisions in a variety of domains. Changes in technology have only increased such reliance. Today, Web and App users share 1.8 billion photos a day. Such images allow people to keep in touch with existing social ties and generate new ones.
But do people pick correctly? That is, do people choose the photos that will lead to the most favorable impressions by others?
We examine this question in the context of posed and candid photos. Posed photos capture people in a particular position to pres-ent themselves in a certain way, usually looking directly at the cam-era. Candid photos, in contrast, don’t involve posing, and capture people acting naturally or spontaneously.
We suggest there is an important disconnect between what people post and what observers prefer: while posters choose posed photos of themselves, observers may actually be more favorably disposed to candid ones (e.g., want to be friends with or date the person).
The notion that posters prefer posed photos is rather intuitive;
posed photos allow people to shape how they come across and more effectively manage impressions.
Observers, however, may respond more favorably to candid photos because they seem more genuine. Genuineness or sincerity is a valued feature of people’s personalities and interactions (Camp-bell & Kirmani, 2000; Tuk, Verlegh, Smidts, & Wigboldus, 2008).
Seeming genuine can have positive effects on how much others like or are persuaded by an individual because they are less likely to think that person has ulterior motives. Along these lines, because candid photos capture people acting naturally or spontaneously, observers should see them as more genuine, which should elicit more favor-able responses.
Twelve studies test this possibility in the lab and in the field.
Study 1a-1d examined various websites to test whether people post candid or posed photos. We coded over 2500 photos from dif-ferent sources: Facebook profiles (Study 1a), OKCupid (Study 1b), the activity site Meetup.com (Study 1c), and Facebook newsfeeds (Study 1d). Across all four contexts, people overwhelmingly posted posed photos (Ms = 87-93%).
To examine posters’ preferences more directly, Study 2a had the same person submit both posed and candid photos and choose which one they would post as their Facebook profile picture. As predicted, participants were more likely to choose the posed photo than the candid one (M = 85% vs. 15%, χ2 (1) = 22.26, p < .001). We find the same results when people chose which photo to post on an online dating site (Study 2b), which photo would lead people to want to date them (Study 2c), or which photo would lead people to want to be friends with them (Study 2d).
The first two studies demonstrate that people post posed photos, and think posed photos will generate desired responses. But are they correct?
Study 3a and 3b focused on observers. All participants were shown the same 30 photos (randomly chosen from Study 1a): half candid, half posed, but rated as equally attractive. As predicted, ob-servers were more interested in being friends with (F(1, 58) = 7.83, p
= .007) or dating (F(1, 65) = 34.01, p < .001) people who used candid photos rather than posed ones.
Study 4 measured the hypothesized underlying process while providing an even cleaner test of our theory by using a yoked design.
Each participant was yoked to one Study 2 participant and viewed ei-ther that participant’s posed or candid photo before rating how inter-ested he or she was in getting to know that person. This allowed us to examine whether candid photos generate more positive responses while perfectly controlling for the individual pictured.
As predicted, observers were more interested in getting to know someone if that person used a candid rather than posed photo (F(1, 91) = 5.42, p = .02). Further, perceived genuineness mediated the impact of photo type on desire to get to know the person better (total indirect effect = .38, 95% CI [.04, .85]). Another study found similar results even when observers were considering photos of people they already knew.
A final study provided further evidence for the underlying pro-cess through moderation. Authenticity perceptions are shaped not only by a photo’s image, but also by the underlying intent. Learn-ing that someone knew a seemLearn-ingly candid photo was beLearn-ing taken should make it seem less genuine. It might seem like the individual had ulterior motives when taking the photo and was thus not sincere (e.g., Campbell & Kirmani, 2000). As a result, intent should mod-erate the positive effect of candidness, and lead otherwise candid-looking photos to have the same effect as posed ones.
Study 5 found exactly this. Observers were more interested in getting to know someone when their photo was candid rather than posed (t(110) = 2.73, p < .01). But telling participants that the person in the photo was aware that the seemingly candid photo was being taken decreased observers’ interest in getting to know that person (t(108) = 3.51, p = .001) to the level of the obviously posed photo (t(111) = -.81, p = .42). These effects were mediated by perceived genuineness.
Taken together, the present research demonstrates an important disconnect between photo posters and observers. People tend to post posed photos of themselves (Study 1a-d) and think that posed photos will generate more favorable responses (i.e., get them more dates or make them more friends, Study 2a-d). Observers, in contrast, actu-ally respond more favorably to candid photos (Study 3-5), in part because they seem more genuine (Study 4-5).
These findings have important implications for self-presenta-tion. People often assume that a curated, polished version of the self will generate the most favorable responses – that by smooth-ing rough edges and presentsmooth-ing one’s best side will make others like them more. But as these results demonstrate, this assumption is not always correct. In some cases, observers may actually prefer an un-varnished perspective because it seems more genuine.
Passing (On) Judgment: Others Judge Us Less Extremely Than We Think
EXTENDED ABSTRACT
Consumers are interested in managing their social identities.
They buy cars to be seen as higher-status, attend high-profile social occasions to be seen as popular, or recount recent visits to cultural
events to signal sophistication. But people are also thrust into situ-ations in which the impressions they make are less controllable, but where the evaluative stakes remain (or seem) high. Company meet-ings, cocktail parties, or team trivia games are all circumstances in which people’s competence, likeability, and intelligence feel on dis-play. In light of people’s interest in and concern with how they are viewed by others, it is natural to ask whether people are accurate in such meta-perceptions—perceptions of how others view them.
Previous research on the spotlight effect (Gilovich, Medvec,
& Savitsky, 2000) has demonstrated that people often overestimate the extent to which their actions are noticed. People are sensitive to the hard-to-see coffee stain on their sleeve or the two hairs that keep moving out of place, even as others notice nothing amiss. The present work builds on this work in three ways. First, we examined if people understand how impressions of them change even when people unambiguously notice their actions. Second, we examine people’s meta-perceptions not merely in circumstances when people might present poorly, but in cases when they present well. Third, we distinguish between mechanistic accounts of when and why meta-perceptions are in error.
The first three studies each test whether actors’ meta-percep-tions would be more reactive to focal successes or failures than ob-servers’ actual perceptions were. We tested this hypothesis in three evaluative domains: intelligence (Study 1), desirability as a dating partner (Study 2), and likeability (Study 3). In each study, actors were yoked to observers, meaning that the actors’ behavior was publicly observable. They then went through a situation in which we could rig things to go well (success condition) or poorly (failure condition). We measured observers’ social perceptions and actors’
meta-perceptions at two times: before the focal event (baseline) and after it (final). All analyses were at the level of the dyad, such that we used a 2(Role: actor or observer) X 2 (Focal Event: success or fail-ure) X 2(Time: baseline or final) mixed design, with the final factor measured within-subjects. In each study, we predicted a significant 3-way interaction indicating that actors’ meta-perceptions (their as-sumptions of how the observers would view them) would be more reactive to the focal event (in a positive or negative direction when they succeeded or failed, respectively) than observers’ impressions actually were.
In Study 1, 99 actor-observer pairs were contestants or view-ers of a mock game show. Actors were videotaped, and observview-ers watched the actors’ experience on tape. In order to provide a con-text for baseline impressions, actors answered 10 dichotomous-choice trivia questions (e.g., “Which city is larger: San Antonio or Detroit?”), reasoning through their answers out loud. Actors and observers then offered baseline meta-perceptions and social percep-tions of the actors’ intelligence. Actors then went through the focal event—a single question that would allow the actors to double their prize or lose everything. Random assignment determined whether they were told they succeeded or failed. Actors and observers then offered their final meta-perceptions and social perceptions along the same intelligence dimension. The predicted 3-way interaction was significant, F(1, 97)=4.25, p=.04. Actors assumed that observers’ im-pressions would be more responsive to the focal success or failure than the observers’ impressions actually were.
Did Study 1’s asymmetry result from actors not appreciating how “sticky” observers’ prior impressions are, or did actors overes-timate the evaluative implications of the focal event? To disentangle
Did Study 1’s asymmetry result from actors not appreciating how “sticky” observers’ prior impressions are, or did actors overes-timate the evaluative implications of the focal event? To disentangle