A flurry of recent news articles has highlighted a new challenge faced by college students: giving the wrong impression about our personalities through our online profiles. Much of this interest can be attributed to the popularity of MySpace®, Facebook®, and other social-networking websites visited by over 10 million people each month (Nielsen/Netratings, 2007). The news articles often caution us about posting information that could disrupt others (particularly potential employers) from seeing us accurately (or at least as we see ourselves). But don’t we want to avoid giving anyone an inaccurate impression of what we are like? How can psychological research help us with this issue?
Research on "social perception" is over 60-years old and has dominated the field of social psychology in recent decades, but it has concentrated largely on the perception that one real person forms about a fictitious target (often referred to as “vignettes”). Such studies can provide useful information on how people combine information about others when they form impressions but, because the targets of the perception are not real people, it is impossible to learn about the accuracy of the impressions. To assess accuracy, the target must be a real person. Such studies fall into the subfield “interpersonal perception” and are surprisingly rare (Funder, 1999; Kenny, 1994). But recently, interpersonal perception studies have enjoyed a resurgence in laboratories across the country as well in an ambitious online research project called YouJustGetMe.com (Evans, Gosling, & Carroll, 2007).
I know what you’re wondering. People have been interacting since the beginning of time. So why is research on the accuracy of first impressions so sparse?
Until recently, psychologists avoided this topic because of a criticism in an influential article by Cronbach (1955) on the way impression accuracy was measured. At that time, psychologists measured impression accuracy by taking the average difference between a perceiver’s ratings and a target’s self-ratings on a series of traits. For example, if you rated me a 7 on extraversion and I rated myself a 4, we would say you were not accurate, because there was a 3 point difference between your rating and mine. The problem with measuring accuracy this way is that it does not take into account differences in how people use scales; maybe you prefer to use points 5, 6, and 7 on the 7-point scale, whereas I prefer points 2, 3, and 4. If so, we may both be saying I am very extraverted by our typical standards, but the absolute difference would suggest we saw my extraversion quite differently. To overcome Cronbach’s criticism, researchers now use Pearson r correlations between impressions and selfratings, which eliminates the problem by standardizing and zero-centering the ratings (Funder, 1999).
Another issue holding back research on accuracy was that psychologists in the past were uncertain whether they were gathering impressions of all personality domains or only some domains. The emergence of the Big Five personality model (McCrae & Costa, 1999), which was designed to measure a broad range of personality traits, reassured researchers that they were not leaving out anything important. The Big Five domains are broad dimensions of personality that consist of Openness to Experience, Conscientiousness, Extraversion, Agreeableness, and Neuroticism. (To remember them, think OCEAN.) Each domain consists of many related traits; for example, appreciation for art and imagination are both included under Openness.
With these improvements and the publication of Malcolm Gladwell's bestseller Blink in 2005, research in the area of first impressions is finally picking up again. Already we know a little about how acquaintanceship affects accuracy and about how different sources of input lead to accurate impressions on different personality domains.
|TABLE 1 | Source for an Accurate Impression
Note. Based on Vazire and Gosling (2004).
Research by Kenny (1994) and others (Ambady & Rosenthal, 1992; Watson, 1989) has verified that, as expected, you have a more accurate impression about someone's personality after knowing the person for a long time compared to a zero-acquaintance interaction (e.g., a brief meeting with a complete stranger). More surprising is that zero-acquaintance impressions are not completely wrong. According to some studies, strangers' impressions of your Extraversion and Conscientiousness can be almost as accurate as the impressions your long-term acquaintances have of you (Kenny, 1994). So the person you meet in an interview understands your Conscientiousness level about as well as someone you have known since kindergarten!
We learn about other personality domains from other sources. For example, your Openness can be detected by snooping around your bedroom or office (Gosling, Ko, Mannarelli, & Morris, 2002). So although first impressions are not 100% accurate, they do give us a pretty good idea about certain aspects of someone's personality.
What does your online profile say about your personality? Vazire and Gosling (2004) believe there are two different mechanisms that people use to determine your personality: your identity claims (“statements made by individuals about how they would like to be regarded," p. 124) and your behavioral residue (“physical traces of a person’s behavior left unintentionally," p. 124). Online profiles, they believe, are dominated by identity claims and contain little behavioral residue, meaning they tend to convey what the profile owner wishes to convey rather than accidentally communicate about him or her. By contrast, bedrooms, offices, and faceto- face interactions convey information via both mechanisms. This makes online profiles ideal for answering the question ”Do people form accurate impressions based on identity claims alone, or must they see behavioral evidence of personality to arrive at accurate conclusions?”
The results of Vazire and Gosling's (2004) study showed that visitors both agreed with each other about how they saw the webpage owner, and they agreed with the owner's selfratings too. The findings suggest that a stranger's impressions of your Agreeableness, Conscientiousness, and Openness are just as accurate after visiting your webpage as the impressions your long-term acquaintances have of you. And your webpage tells people as much about your Extraversion and Openness as a visit to your office or bedroom. One of the most interesting suggestions from this study is that you can form an accurate impression of someone's Agreeableness from a personal webpage. Most research has shown that zero-acquaintance strangers are typically clueless about each other's Agreeableness (Watson, 1989). That explains why so many unfortunate people only learn that someone is a sweetheart or a jerk after dating the person for a while. Vazire and Gosling's (2004) study suggests that viewing someone's online profile may be a way to avoid this uncomfortable situation.
Are the results the same for Facebook profiles? Suspecting this to be the case, Gosling teamed up with social psychologist David Evans, formerly of Microsoft® and Classmates.com® to find out. In November of 2007, they launched the YouJustGetMe application on the Facebook Platform, where over 6,300 students from over 150 universities have since formed almost 9,000 impressions of each other. They also launched the YouJustGetMe.com website so they could randomly assign people to rate each other and examine which pieces of information on your profile give others an accurate impression of your personality (e.g., through pictures, favorite films, embarrassing moments, political leanings, etc.) and which pieces do not.
On both the website and on Facebook, you first rate yourself on a 43-item Likert scale, 21 of which assess your personality on the Big Five domains while the remaining assess your preferences and attitudes. When you are done, you get to see your Big Five results and invite visitors to form impressions of you. The visitors' task is simple—they just try to rate you on the same 43 questions the way you rated yourself. YouJustGetMe immediately calculates the Pearson r "accuracy score" and displays it to both parties. It ranges from -1 (wrong) to 0 (clueless) to +1 (perfect). Thus, both parties learn what impression is given and whether the visitor is "reading" the profile owner correctly.
The Big Five results are shown with unique "bubble graphs" where the size of the bubbles indicate how strongly you demonstrate each trait. Evans and Gosling chose to modify the labels given to the Big Five domains because they wanted to use nonjudgmental terms, avoid the confusion of telling people they are “low” or “high” on a trait, and label both poles of the Big Five domains. (Historically, personality researchers have agreed there are five personality domains but allowed for some flexibility in how they are named. See Goldberg, 1990.) Thus, YouJustGetMe reports whether you are disciplined vs. casual, alternative vs. traditional, neurotic vs. unemotional, cooperative vs. competitive, and extroverted vs. introverted, which represent Conscientiousness, Openness, Neuroticism, Agreeableness, and Extroversion, respectively.
Note. Sample impression-accuracy results from YouJustGetMe.com. Reprinted with permission from Psychster, LLC.
Unpublished preliminary results of the YouJustGetMe project suggest that the methodology is comparable to past work (Gosling, Gaddis, & Vazire, 2007). The overall average accuracy correlation for randomly assigned dyads on the site is r = .29, which is comparable to the r = .27 found by Vazire and Gosling (2004). On Facebook where the dyads are not randomly assigned (and thus more likely to know each other), the overall average accuracy correlation is somewhat higher at r = .42. Also consistent with Vazire and Gosling (2004), there appears to be a gender effect where women are more accurate at guessing others (r = .33), and more easily guessed (r = .34), compared to men (respective r s = .25 and .24). These significant correlations suggest to Evans and Gosling that unlike the early days of the Internet when many people were trying on alternate identities, most people today tend to portray themselves consistent with their self-image.
Evans and Gosling also hope that the YouJustGetMe project will collect a large, diverse dataset for studying how impression accuracy differs across groups. A basic assumption of prejudice theory is that stereotypes distort people's perceptions of the members of stigmatized groups. This should lower the accuracy by which their personalities are perceived. On the Science page of the YouJustGetMe website, members can view a daily updated average of how well people read the personalities of members who selfidentify as being of African, Asian, Latin, and Caucasian descent. Indeed, the mostly Caucasian membership (so far) is less accurate at reading the personalities of non- Caucasians than other Caucasians. With more careful analysis, the data should provide new insights into the ways that stereotypes impair our ability to see others as they see themselves.
So what courses would best prepare students interested in doing research on interpersonal perception? Traditionally, the topic has been studied in two somewhat distinct traditions: personality psychology, which tends to focus on factors that contribute to accurate perceptions of others, and social psychology, which tends to focus on factors that interfere with accurate perceptions of others. So courses in either (or both) of these fields would serve as good preparation for students interested in examining these topics. Recently researchers such as David Kenny at the University of Connecticut have taken an approach that explicitly stands at the intersection of personality and social psychology. So you might be lucky enough to find a course there dedicated to person perception that takes such an integrative stance. In addition, other useful courses would be those on research design that included material on online methodology, and those on ecological or environmental or cognitive psychology that include material on Brunswikian approaches to judgment.
There are a few challenges and debates you should be aware of before you dive into research using the World Wide Web. First, there is an on-going debate about the ethics of online research. Some consider online behavior to be public behavior and therefore social scientists do not require informed consent of the participants to observe it as long as their anonymity is protected. Others argue that it must be treated like any other research using human subjects which would ensure that the procedures fulfill the principles of voluntary participation and informed consent, maintain the confidentiality of information obtained from or about human subjects, and adequately address possible risks to subjects including psychosocial stress. Second, online research is field research, which does not allow for much experimental control. This leads to high external validity but low internal validity, meaning the likelihood of the relationship happening outside the lab is high but we cannot conclude causality, only correlation. And third, a challenge that particularly affects personality research is determining accuracy criteria. Should people’s personality be defined as how they see themselves or a consensus of how their peers see them? Finally, whether you put information online about yourself or you are running an online research lab, be cautious of the information you provide about yourself and your participants because it will be available to anyone with a computer. If you decide to conduct your own research in this area be sure to be aware of these issues.
As first-impression research finally resumes, there are so many interesting questions yet to be asked. Why don't long-term friends generally become perfect in their impressions of each other? Which tells people more about your personality—your proudest moment or your most embarrassing moment? And are Asians, for example, better than other ethnic groups at judging other Asians? Online labs like YouJustGetMe are pushing social psychology into a new and exciting future that hopes to determine how we interpret others’ personalities.
Ambady, N., & Rosenthal, R. (1992). Thin slices of expressive behavior as predictors of interpersonal consequences: A meta-analysis. Psychological Bulletin, 111, 256-274.
Cronbach, L. J. (1955). Processes affecting scores on "understanding of others" and "assumed similarity." Psychological Bulletin, 52, 177-193.
Evans, D. C., Gosling, S. D., & Carroll, A. (2007). YouJustGetMe 'Learn more' page. Retrieved January 11, 2008, from http://www.youjustgetme.com/?page=learn_more
Funder, D. C. (1999). Personality judgment: A realistic approach to person perception. San Diego, CA: Academic.
Gladwell, M. (2005). Blink: The power of thinking without thinking. Boston, MA: Little, Brown.
Goldberg, L. R. (1990). An alternative “description of personality”: The Big-Five factor structure. Journal of Personality and Social Psychology, 59, 1216-1229.
Gosling, S. D., Ko, S. J., Mannarelli, T., & Morris, M. E. (2002). A room with a cue: Personality judgments based on offices and bedrooms. Journal of Personality and Social Psychology, 82, 379-398.
Gosling, S. D., Gaddis, S., & Vazire, S. (2007, March). Personality impressions based on Facebook profiles. In Proceedings of the International Conference on Weblogs and Social Media, Boulder, CO. Retrieved January 11, 2008, from http://www.icwsm.org/ papers/3--Gosling-Gaddis-Vazire.pdf
Kenny, D. A. (1994). Interpersonal perception: A social relations analysis. New York, NY: Guilford Press.
McCrae, R. R., & Costa, P. T., Jr. (1999). A five-factor theory of personality. In L. A. Pervin & O. P. John (Eds.), Handbook of personality theory and research (pp. 139-153). New York: Guilford Press.
Nielsen/Netratings. (2007). Standard metrics, month of May, 2007. Retrieved December 6, 2007, from NetRatings Inc. Web site: http://www.netratings.com
Vazire, S., & Gosling, S. D. (2004). e-Perceptions: Personality impressions based on personal websites. Journal of Personality and Social Psychology, 87,123-132.
Watson, D. (1989). Strangers' ratings of the five robust personality factors: Evidence of a surprising convergence with self-report. Journal of Personality and Social Psychology, 57,120-128.
For further reading:
- Funder, D. C. (1999). Personality judgment: A realistic approach to person perception. San Diego, CA: Academic.
- Hall, J. A., & Bernieri, F. (Eds.). (2001). Interpersonal sensitivity: Theory and measurement. Mahwah, NJ: Lawrence Erlbaum
- Kenny, D. A., Kashy, D. A., & Cook, W. L. (2006). Dyadic data analysis. New York: Guilford Press.
Interpersonal Perception Laboratories:
- University of Texas at Austin—Gosling Laboratory. Prof. Samuel D. Gosling. http://homepage.psy.utexas.edu/homepage/ faculty/gosling/
- Tufts University—Interpersonal Perception & Communication Laboratory. Prof. Nalini Ambady. http://ase.tufts.edu/psychology/ambady/
- University of California, Riverside—The Riverside Accuracy Project. Prof. David Funder. www.rap.ucr.edu/
- For additional labs visit www.psych.wustl.edu/pal/links.html
Rachel K. Green is a senior psychology major at Middle Tennessee State University and the president of the local Psi Chi chapter. She is the recipient of a URSCP Assistant Grant for research in social psychology, which she conducts with her advisor John Pennington. Ms. Green has recently been studying in France, where she missed her two cats.
David C. Evans, PhD, earned his PhD in social psychology from the University of Iowa in 1999. He is the founder of Psychster LLC, a former senior business analyst with Classmates.com® and beta program manager at Microsoft®. Prior to going online, Dr. Evans was a visiting assistant professor at Union College in Schenectady, NY.
Samuel D. Gosling, PhD, is an associate professor of psychology at the University of Texas at Austin. He earned his PhD from Berkeley and actively publishes on online social dynamics and impression accuracy. Dr. Gosling was featured in Malcolm Gladwell's bestseller Blink and is currently working on a manuscript through Basic Books.