Impressions of Intelligence and Personality Traits Due to Internet Slang in Male and Female Facebook Profiles
Meghan C. McLean, Elizabeth Boyland, Phebe Pierson, and Amy M. Falk
Social networking sites (SNS) such as Facebook allow people to form impressions of each other every day. When a Facebook profile is the sole source of how an individual is perceived, the information and the manner in which it is provided are highly influential (Hancock & Dunham, 2001; Jessmer & Anderson, 2001; Marcus et al., 2006). Internet slang, initialisms (e.g. “ttyl” meaning “talk to you later”), and letter homophones (e.g. “r” meaning “are”), used in Facebook profiles can contribute to the perception of the profile maker.
How information is conveyed through computer-mediated communication (CMC) as well as individual’s attitudes, impact impression formation. Share and Silva (2003) studied the IQ gender bias. They found that the pervasiveness of reading disabilities in boys is partially due to the stereotype that boys have higher IQs than girls. Accordingly some people assume that boys will more easily learn to read. Share and Silva’s (2003) results indicate that boys are perceived as having the ability to read due to an erroneous gender bias. In other contexts, such as in the political sphere, women who are equally as qualified as men appear less competent to voters (Paul & Smith, 2008). Biernat and Vescio (2002) explained differences in perceived intelligence between men and women through what they call “shifting standards.” If women are perceived as being less intellectual, they fall victim to a “shifted standard,” and must demonstrate more understanding and complex thought in order to be considered a man’s equal. Gill et al. (2006) commented that impressions from texts can be influenced by stereotypes (e.g. gender beliefs). If a gender bias is prevalent, a shifted standard would cause a female text to be judged as less intelligent than an equivalent male text. Research has also supported the idea that politeness and grammar in text is related to the author’s perceived gender (Jessmer & Anderson, 2001). In fabricated emails, polite and grammatical messages were perceived as written by women, whereas impolite and ungrammatical messages were perceived as written by men.
Jessmer and Anderson (2001) also investigated how the grammar and politeness of an email can affect the recipient’s impressions of the sender’s competency. Authors of polite grammatical messages were rated as more competent and likeable compared to authors of other messages. A similar study by Lewandowski and Harrington (2006) found that phonetic abbreviations caused participants to conclude that writers were less capable and exhibited less effort than e-mailers who did not employ abbreviations. There is also research regarding other kinds of computer mediated communications. Hancock and Dunham (2001) investigated whether text-based computer-mediated conversations or face-to-face interactions were broader (numerous characteristics identified) and more intense (the strength of each characteristic). They discovered that the impressions formed from CMC were more intense than and not as broad as the impressions formed from face-to-face interactions.
Marcus, Machilek, and Shutz (2006) looked at personal Web sites in order to explore the owners’ personalities. They compared their own interpretations of the Web sites and visitors’ ratings to the owners’ self-reports of personality. They found that the visitors of the Web site made many inferences about the owner’s personality after only visiting the site for a few minutes. Similarly, information provided through social networking sites causes people to form exaggerated, stereotypical, and biased impressions. SNS’s lack contextual cues that influence impressions, therefore the manner in which information is provided is highly influential (Hancock & Dunham, 2001; Jessmer & Anderson, 2001; Marcus et al., 2006).
Profiles on the SNS Facebook can influence how a person is perceived, particularly in the area of personality. Gosling, Gaddis and Vazire (2007) examined the accuracy of personality impressions formed on Facebook profiles. Results indicated that Facebook based personality impressions reflected participants’ results in almost all of the Big Five dimensions. Students were most accurate at identifying extroversion in the profiles. Gosling et al. (2007) concluded that Facebook profiles can be used to communicate personality. Gill, Oberlander and Austin (2006) produced similar results. After exposure to a 300 word text, participants consistently agreed on the author’s level of extroversion and to a lesser degree, the author’s neuroticism. Accurate impressions can be formed through social-networking sites (Back et al., 2010; Gill et al., 2006; Gosling et al., 2007).
The aforementioned studies suggest that the manner in which information is communicated on Facebook profiles may influence viewers’ impressions of the user. Commonly used Internet abbreviations, such as “LOL” and “TTYL,” may impact perceptions of the author’s intelligence. This study investigated the effects of content and Internet slang on impressions of intelligence and personality traits in Facebook profiles, with an additional focus on how gender affects people’s perceptions of the slang usage. It was predicted that students would perceive profile makers who used internet slang as less intelligent, and that the gender of the profile maker would impact students’ perceptions of intelligence in otherwise identical profile pages. It was expected that male profiles would produce higher ratings of intelligence than their female counterparts.
Method
Participants
There was a total of 89 participants all students of Connecticut College who signed up for the study on the second floor of Bill Hall and received 30 minutes of research participation credit. These included 15 men and 74 women. Of the total participants, 27 were 18 years old, 38 were 19 years old, 15 were 20 years old, eight were 21 years old, and one was 22 years old. There were 35 freshmen, 37 sophomores, nine juniors, and eight seniors. Seventy-seven of the participants were white, three were black, two were Asian, three were Hispanic, three were Asian and Hispanic, and one was white and Asian. One hundred percent of participants had a Facebook account and all but one reported using it at least once a day. All participants reported using Facebook at least once a week. The mean amount of time spent on Facebook each day was 1.93 hours, with a range from zero to ten hours. The mean number of texts sent per day among the participants was 71.
Materials
Materials included four Facebook profiles, a social desirability scale, and a survey measuring the participants’ impressions of personality traits of the provided Facebook profiles. There was one male Facebook profile without use of any internet slang (e.g. “I love to play basketball”, “I love to travel, eat, and relax on the beach”) and one male Facebook profile with use of internet slang (e.g. “i luv 2 play bball”, “I luv 2 travel eat and relax on the beach”). There was also one female Facebook profile with no internet slang usage, and one female Facebook profile with internet slang usage (See appendices A-D). The content of the male and female profiles were identical including the gender neutral name “Taylor Thomas”; the only differences were in the gender and slang usage. The “bio” and “favorite quotes” sections were filled out and the year of birth, number of friends, names of schools attended, and dates of graduations were blurred out so as to control for as many variables as possible . The profile picture will be the standard Facebook silhouette of a man or woman depending on the profile’s assigned gender.
The Crowne and Marlowe Social Desirability Scale was used to assess the participants' desire to be viewed favorably (i.e., “I never hesitate to go out of my way to help someone in trouble”). This scale has 33 true/false questions and has a Cronbach’s alpha of .74. In the present study, the Cronbach’s alpha was .73. Scores can range from 0 to 33, where low scores indicate low social desirability and high scores indicate high social desirability. Those who are low scorers act and respond in socially undesirable ways, average scorers follow social norms, and high scorers show high conformity to social rules (Sierra et al., 2010).
The participants’ impressions of the Facebook profiles were rated on a 7 point Likert scale, in which they rated the profile on attributes such as confidence, intelligence, and friendliness. This scale was created by the authors. A rating of 1 meant “not at all,” a rating of 4 meant “average,” and a rating of 7 meant “extremely.” This scale was moderately reliable with a Cronbach’s alpha of .616.
Participants completed a demographics form which asked about their racial/ethnic background, gender, sexual orientation, class year, age, and what social networking sites and instant messaging accounts they used. The last question on the demographics page was a manipulation check about the gender of the profile that was viewed.
Procedure
Participants were obtained through the Introductory Psychology courses’ participant pool at Connecticut College. The setting was a college classroom equipped with a projector. When participants arrived to the study, experimenters asked them to sign an informed consent form. Once all students were ready to begin, the experimenters explained the purpose of the study as an investigation of first impression of Facebook profiles. The experimenter then handed out the survey packet and turned on the projector, which displayed the Facebook profile. Participants were shown one of four profiles: male/slang, male/no slang, female/slang, or female/no slang. Once all of the participants finished filling out the questionnaires regarding the Facebook profile and Social Desirability, they arrived at a message at the bottom of the Facebook profile evaluation page stating “Please put down your writing utensils and wait for further instructions.” The projection was turned off once all participants had put down their pencils in order for the manipulation check to be successful. If the projector had remained on, the participants would have been able to read the profile’s gender off of the screen and write it as the answer to the manipulation check. Since the purpose of the check was to determine whether the gender of the profile had influenced the participants’ perceptions, it was necessary that the participants not be able to see the gender when completing the check. At this point, participants completed the demographics page, which included the manipulation check. As participants finished the study, they handed it in in exchange for a debriefing statement and 30 minutes of research credit.
Results
A MANOVA was conducted to analyze the hypotheses that the gender of the Facebook profile would influence the impressions formed, and that profiles containing slang would produce different perceptions of personality traits than the profiles with proper grammar. In order to test the hypothesis that Facebook profiles with slang would be perceived as less intelligent than profiles without slang, a MANOVA was conducted using slang and no slang profiles as the levels of the independent variable and all the personality traits as the dependent variables, F(1, 86) = 84.952, p < .001, η2=.497. Overall, slang use had a significant effect on all personality traits at the multivariate level, Wilks’ Lambda = .296, F(15, 72) = 11.441, p < .001.
At the univariate level, many personality traits were significantly affected by slang usage. Profiles with proper grammar had significantly higher ratings of politeness, competency, seriousness, confidence, honesty, creativity, and trustworthiness, and were rated as more goal oriented. These profiles also had significantly lower ratings of outgoingness, aggressiveness, sociability, and were rated as less judgmental (see Tables 1 and 2).
A MANOVA was conducted to analyze the hypothesis that the profile’s gender would impact the participants’ personality ratings of the profile. A manipulation check filtered the data so only the results of students who correctly reported the profile gender were used. Because of this manipulation check, 20 participants were excluded from the analysis of this hypothesis. Gender caused no significant differences across the dependent variables at the multivariate level, Wilks’ Lambda = .775, F(15, 72) = 1.009, p = .461.
Further analysis at the univariate level revealed statistically significant differences between the perceived friendliness and politeness of male and female profiles where male profiles were perceived as less friendly, F(1,66) = 5.158, p= .026, η2 =.071 , and less polite, F(1,66) = 5.077, p= .028; η2 =.093 (see Table 3). There was no significant difference between the profile gender and the other personality attributes.
Discussion
The hypothesis, which was that slang usage would significantly reduce intelligence perceptions of the Facebook profile user, was supported. The results of the experiment indicated that profiles using phonetic abbreviations influenced the perceptions of personality traits. The other hypothesis, which was that gender would significantly alter ratings of intelligence, was not supported.
Phonetic abbreviations lowered the participants’ impressions of the intelligence of the profile owner. Initialisms, such as “lol” instead of “laugh out loud,” and internet slang, are conditioned stimuli that cause lower levels of assumed competence and intelligence. Generally, people associate the use of such abbreviations with laziness. For instance, if someone is not willing to entirely write out a word, he or she may be perceived as putting in less effort (Jessmer & Anderson, 2001). Additionally correct spelling and grammar is a conditioned stimulus for a higher education level and a more extensive vocabulary. Proper spelling and grammar are indicative of education, and are assumed skills that are necessary for success in higher education. Consequently, profiles utilizing proper grammar and spelling were perceived as more goal-oriented, another of the personality traits rated by participants. Profiles that utilize phonetic abbreviations also produced decreased ratings of creativity, a personality trait associated with intelligence. Perhaps the participants view a high level of intelligence as necessary for creativity and inventiveness.
At the univariate level, the slang-using profiles produced lower perceptions of politeness and seriousness. The reasons behind the politeness results are relatively straightforward; when one wants to convey respect and when one is communicating with those of higher statuses, proper grammar is considered good manners. For instance, if a student were to communicate with a professor, he or she would most likely try to use correct spelling and eliminate grammatical errors. As a result, more grammatical messages seem more polite, and ungrammatical messages seem impolite. The results regarding seriousness are less straightforward. One explanation could be that those who write out grammatically correct messages have invested more energy in them. These individuals may be perceived as more serious because they are concerned with how each word is spelled and the clarity of their message, while those who employ abbreviations such as “lol” and “jk” do not seem as concerned. Thus, the findings regarding seriousness may be explainable due to the participants’ views of how much effort and thought they believed were invested in the Facebook profiles.
Ratings of trustworthiness were also found to be higher for the profiles that used no slang. Because proper grammar is evidently seen as requiring more effort, more intelligence, and more motivation, it does seem logical that participants would rate those who use proper grammar as more trustworthy. Proper grammar causes people to be perceived as less lazy. Individuals would most likely be hesitant to trust someone who may not understand them or may be too lazy or careless to do a task entrusted to them. This logic explains the results for honesty as well, since trustworthiness and honesty are closely linked. Like trustworthiness, profiles with no slang were also rated as more honest than those using slang.
Slang usage also caused certain personality traits to be rated higher. The profile owners who used slang were perceived as more social, outgoing, aggressive, and judgmental than the profile owners who used correct grammar. The result for sociability could be explained by assuming that people who use more modes of computer mediated communication, or who use them more frequently, are more likely to use abbreviations. Thus, abbreviation use could be considered an indicator of an active social life, and therefore those who use them are perceived as more social and outgoing. Both male and female profiles with slang were also rated as more aggressive than profiles without slang. Previous research has found that individuals assume polite emails to be written by women (Jessmer & Anderson, 2001), a finding that in turn links men to impolite and, presumably, badly written emails. This link between men and impoliteness can be extrapolated to encompass slang, as well. Perhaps participants in the present study unconsciously associate slang with masculinity. Because men are stereotypically more aggressive than women, profiles using slang could have been rated as more aggressive because slang usage is associated with masculine characteristics.
The results from the present study also indicated that profile owners who use slang are considered more judgmental. This could be due to the influence of the media on our participants (and the greater high school and college-age populations in general). The media contributes to the idea that phonetic abbreviations indicate negative characteristics. Commercials, television shows, and movies like Mean Girls (2004) typically show a “Bitchy Teenage Girl” archetype; the girls are shown texting and gossiping with their friends, and they generally seem shallow and immature. Although it is not always explicit, it may be assumed by some that these girls use abbreviations as they use computer-mediated communication and texting. Consequently, those who use abbreviations may be perceived as judgmental. In addition, there are and increasing number of news stories about cyber-bullying, whereby one student is ostracized and ridiculed by his or her peers over the internet. Again, the use of abbreviations is not explicit in these news stories, but the assumption here may also be that those who use abbreviations are more frequent users of computer mediated communication. The links between these media influences are presumably enough for participants in the present study to form the impression that the profile owner who uses slang is more judgmental.
The hypothesis that profile gender would alter ratings of intelligence was not supported. At the multivariate level, there was no significant effect of gender on the dependent variables; however, univariate results revealed that the politeness and friendliness of the profiles were influenced by gender. The research of Jessmer and Anderson (2001) showed that participants rated polite messages as a product of a female author, a conclusion that was supported by the present study. Female Facebook profiles were rated as more polite than male Facebook profiles, regardless of slang use. Other research supports these findings as well. For instance, there is a stereotype that women use more polite and mild mannered language, while men have less polite, less mild mannered, and more aggressive language (Leaper & Ayres, 2007). Gender also influenced how friendly the profile owner seemed. Perhaps due to a benevolent sexism stereotype that women are kinder than men (Eagly, 1994; Haddock & Zanna, 1994), female profiles were rated as friendlier than male profiles.
There were a few confounding variables that may have affected the results of the present study. For instance, there was a technical difficulty with the projector during the second session of the study. There was a brief delay in viewing the profile, and although students were able to see the entire profile initially, they could only see the ungrammatical text while they completed the survey; they had all the necessary information but could not see the generic Facebook silhouette picture. Additionally, the profiles had a quotation from Helen Keller. In the ungrammatical, slang using profile, the abbreviation “hahaha” followed the quotation. Some students may have been offended or found it comical, which could have changed their view of the profile owner.
Another confounding variable was a single typo in the grammatical profile. The letter “i” should have been capitalized when it was not. The typo may have caused us to find fewer significant differences than we would have otherwise; however, it seems unlikely that this mistake caused any significant difference in our results.
The present study also suffers from the “college sophomore problem.” The convenience sample of psychology students primarily consisted of female, Caucasian, sophomores. The unbalanced nature of this sample may have threatened the ecological validity of our research. Despite this lack of diversity within the sample, people of different populations could conceivably react similarly to the profiles. The sample in the present study consisted of college age students who use texting and abbreviations themselves. If future research sampled an older population that uses fewer abbreviations, the participants might be less used to the slang and therefore more likely to form judgments. However, there was a high level of mundane realism due to the screen shots of real Facebook profiles presented to the students, and there was a high level of control due to identical messages, the difference being only the presence or absence of abbreviations.
A final confounding variable is a gender bias that could have caused the female profile to be perceived as more masculine. In attempting to keep the profiles as controlled as possible, the same information was given in both the male and female profiles. Therefore, both the male and the female profile owners played basketball. Whereas perceptions of men are not altered by whether or not they play basketball, there is a bias that causes people to perceive women who play basketball as more masculine than women who do not (Csizma et al., 1988).
Future research should address the other aspects of Facebook profiles that may contribute to impression formation. For example, people’s likes, dislikes, photographs, and wall posts could be a source for judgment.
Future research could also investigate how long a first impression based on slang on Facebook lasts, and how it affects two people’s face-to-face interaction. An individual may treat another differently if he or she has pre-formed attitudes based on his or her use of phonetic abbreviations. Research should also address the use of phonetic abbreviations in daily conversation, and how vocal use of the abbreviations may influence people’s impressions of others. Spoken abbreviations may seem more salient due to their minimal use in conversation, and cause for more extreme attitudes to be formed about the individual speaking them. Also, more research investigate the importance of dialects, accents, and slang that pertains to specific population groups in the formation of impressions outside that group.
A content analysis could also be conducted regarding media portrayals of phonetic abbreviations. Commercials, advertisements, television shows, and movies could be analyzed in order to study how individuals using abbreviations are perceived. The media is likely contributing to negative stereotypes associated with abbreviation usage.
The practical and theoretical implications of the current study are vast. One hundred percent of our participants had a Facebook account, which implies a similar number within the larger population. With such a large portion of the population on Facebook, it is extremely important to understand how social networking sites are altering perceptions of people. Forms of computer-mediated communication are steadily becoming more integrated into how we talk with one another. People are tempted to use abbreviations to facilitate their messaging; however, based on the results of this study it can be inferred that they will be perceived differently after doing so. If an individual desires to seem social, outgoing, aggressive and judgmental then using abbreviations may be appropriate. However, it is imperative to avoid abbreviations and improper grammar when one wants to be perceived as intelligent, competent, creative, goal-oriented, polite, serious, honest and trustworthy.
References
Table 1
F values and significance of grammar on personality traits.
Personality Traits F(1,86) p η2
Outgoing 12.024 .001 .123
Competence 21.936 <.001 .203
Politeness 48.384 <.001 .360
Seriousness 44.558 <.001 .341
Confidence 4.300, .041 .048
Aggressiveness 11.695 .001 .120
Intelligence 84.952 <.001 .497
Honesty 6.322 .014 .068
Social 10.1127 .002 .105
Judgmental 22.767 <.001 .209
Goal-Oriented 28.935 <.001 .252
Creative 12.377 .001 .126
Trustworthy 46.211 <.001 .350
Table 2
Mean ratings of personality traits in slang and no slang profiles.
Slang No Slang
Personality Traits N M SD N M SD
Outgoing 37 5.784 1.182 51 4.892 1.197
Competence 37 3.378 1.320 51 4.431 0.781
Politeness 37 2.946 0.970 51 4.471 1.046
Seriousness 37 2.027 0.726 51 3.647 1.339
Confidence 37 5.487 1.387 51 4.882 1.321
Aggressiveness 37 4.541 1.592 51 3.373 1.574
Intelligence 37 2.487 1.044 51 4.275 0.777
Honesty 37 3.865 1.251 51 4.471 1.007
Social 37 5.865 1.110 51 4.902 1.578
Judgmental 37 4.405 1.257 51 3.177 1.144
Goal-Oriented 37 2.676 1.270 51 4.137 1.250
Creative 37 2.568 1.345 51 3.608 1.387
Trustworthy 37 2.891 1.197 51 4.373 0.848
Table 3
Means and standard deviations for friendliness and politeness of male and female profiles.
Male Female
Personality Traits N M SD N M SD
Friendliness 49 4.778 1.149 40 5.344 1.231
Politeness 49 3.528 1.006 40 4.188 1.271
How information is conveyed through computer-mediated communication (CMC) as well as individual’s attitudes, impact impression formation. Share and Silva (2003) studied the IQ gender bias. They found that the pervasiveness of reading disabilities in boys is partially due to the stereotype that boys have higher IQs than girls. Accordingly some people assume that boys will more easily learn to read. Share and Silva’s (2003) results indicate that boys are perceived as having the ability to read due to an erroneous gender bias. In other contexts, such as in the political sphere, women who are equally as qualified as men appear less competent to voters (Paul & Smith, 2008). Biernat and Vescio (2002) explained differences in perceived intelligence between men and women through what they call “shifting standards.” If women are perceived as being less intellectual, they fall victim to a “shifted standard,” and must demonstrate more understanding and complex thought in order to be considered a man’s equal. Gill et al. (2006) commented that impressions from texts can be influenced by stereotypes (e.g. gender beliefs). If a gender bias is prevalent, a shifted standard would cause a female text to be judged as less intelligent than an equivalent male text. Research has also supported the idea that politeness and grammar in text is related to the author’s perceived gender (Jessmer & Anderson, 2001). In fabricated emails, polite and grammatical messages were perceived as written by women, whereas impolite and ungrammatical messages were perceived as written by men.
Jessmer and Anderson (2001) also investigated how the grammar and politeness of an email can affect the recipient’s impressions of the sender’s competency. Authors of polite grammatical messages were rated as more competent and likeable compared to authors of other messages. A similar study by Lewandowski and Harrington (2006) found that phonetic abbreviations caused participants to conclude that writers were less capable and exhibited less effort than e-mailers who did not employ abbreviations. There is also research regarding other kinds of computer mediated communications. Hancock and Dunham (2001) investigated whether text-based computer-mediated conversations or face-to-face interactions were broader (numerous characteristics identified) and more intense (the strength of each characteristic). They discovered that the impressions formed from CMC were more intense than and not as broad as the impressions formed from face-to-face interactions.
Marcus, Machilek, and Shutz (2006) looked at personal Web sites in order to explore the owners’ personalities. They compared their own interpretations of the Web sites and visitors’ ratings to the owners’ self-reports of personality. They found that the visitors of the Web site made many inferences about the owner’s personality after only visiting the site for a few minutes. Similarly, information provided through social networking sites causes people to form exaggerated, stereotypical, and biased impressions. SNS’s lack contextual cues that influence impressions, therefore the manner in which information is provided is highly influential (Hancock & Dunham, 2001; Jessmer & Anderson, 2001; Marcus et al., 2006).
Profiles on the SNS Facebook can influence how a person is perceived, particularly in the area of personality. Gosling, Gaddis and Vazire (2007) examined the accuracy of personality impressions formed on Facebook profiles. Results indicated that Facebook based personality impressions reflected participants’ results in almost all of the Big Five dimensions. Students were most accurate at identifying extroversion in the profiles. Gosling et al. (2007) concluded that Facebook profiles can be used to communicate personality. Gill, Oberlander and Austin (2006) produced similar results. After exposure to a 300 word text, participants consistently agreed on the author’s level of extroversion and to a lesser degree, the author’s neuroticism. Accurate impressions can be formed through social-networking sites (Back et al., 2010; Gill et al., 2006; Gosling et al., 2007).
The aforementioned studies suggest that the manner in which information is communicated on Facebook profiles may influence viewers’ impressions of the user. Commonly used Internet abbreviations, such as “LOL” and “TTYL,” may impact perceptions of the author’s intelligence. This study investigated the effects of content and Internet slang on impressions of intelligence and personality traits in Facebook profiles, with an additional focus on how gender affects people’s perceptions of the slang usage. It was predicted that students would perceive profile makers who used internet slang as less intelligent, and that the gender of the profile maker would impact students’ perceptions of intelligence in otherwise identical profile pages. It was expected that male profiles would produce higher ratings of intelligence than their female counterparts.
Method
Participants
There was a total of 89 participants all students of Connecticut College who signed up for the study on the second floor of Bill Hall and received 30 minutes of research participation credit. These included 15 men and 74 women. Of the total participants, 27 were 18 years old, 38 were 19 years old, 15 were 20 years old, eight were 21 years old, and one was 22 years old. There were 35 freshmen, 37 sophomores, nine juniors, and eight seniors. Seventy-seven of the participants were white, three were black, two were Asian, three were Hispanic, three were Asian and Hispanic, and one was white and Asian. One hundred percent of participants had a Facebook account and all but one reported using it at least once a day. All participants reported using Facebook at least once a week. The mean amount of time spent on Facebook each day was 1.93 hours, with a range from zero to ten hours. The mean number of texts sent per day among the participants was 71.
Materials
Materials included four Facebook profiles, a social desirability scale, and a survey measuring the participants’ impressions of personality traits of the provided Facebook profiles. There was one male Facebook profile without use of any internet slang (e.g. “I love to play basketball”, “I love to travel, eat, and relax on the beach”) and one male Facebook profile with use of internet slang (e.g. “i luv 2 play bball”, “I luv 2 travel eat and relax on the beach”). There was also one female Facebook profile with no internet slang usage, and one female Facebook profile with internet slang usage (See appendices A-D). The content of the male and female profiles were identical including the gender neutral name “Taylor Thomas”; the only differences were in the gender and slang usage. The “bio” and “favorite quotes” sections were filled out and the year of birth, number of friends, names of schools attended, and dates of graduations were blurred out so as to control for as many variables as possible . The profile picture will be the standard Facebook silhouette of a man or woman depending on the profile’s assigned gender.
The Crowne and Marlowe Social Desirability Scale was used to assess the participants' desire to be viewed favorably (i.e., “I never hesitate to go out of my way to help someone in trouble”). This scale has 33 true/false questions and has a Cronbach’s alpha of .74. In the present study, the Cronbach’s alpha was .73. Scores can range from 0 to 33, where low scores indicate low social desirability and high scores indicate high social desirability. Those who are low scorers act and respond in socially undesirable ways, average scorers follow social norms, and high scorers show high conformity to social rules (Sierra et al., 2010).
The participants’ impressions of the Facebook profiles were rated on a 7 point Likert scale, in which they rated the profile on attributes such as confidence, intelligence, and friendliness. This scale was created by the authors. A rating of 1 meant “not at all,” a rating of 4 meant “average,” and a rating of 7 meant “extremely.” This scale was moderately reliable with a Cronbach’s alpha of .616.
Participants completed a demographics form which asked about their racial/ethnic background, gender, sexual orientation, class year, age, and what social networking sites and instant messaging accounts they used. The last question on the demographics page was a manipulation check about the gender of the profile that was viewed.
Procedure
Participants were obtained through the Introductory Psychology courses’ participant pool at Connecticut College. The setting was a college classroom equipped with a projector. When participants arrived to the study, experimenters asked them to sign an informed consent form. Once all students were ready to begin, the experimenters explained the purpose of the study as an investigation of first impression of Facebook profiles. The experimenter then handed out the survey packet and turned on the projector, which displayed the Facebook profile. Participants were shown one of four profiles: male/slang, male/no slang, female/slang, or female/no slang. Once all of the participants finished filling out the questionnaires regarding the Facebook profile and Social Desirability, they arrived at a message at the bottom of the Facebook profile evaluation page stating “Please put down your writing utensils and wait for further instructions.” The projection was turned off once all participants had put down their pencils in order for the manipulation check to be successful. If the projector had remained on, the participants would have been able to read the profile’s gender off of the screen and write it as the answer to the manipulation check. Since the purpose of the check was to determine whether the gender of the profile had influenced the participants’ perceptions, it was necessary that the participants not be able to see the gender when completing the check. At this point, participants completed the demographics page, which included the manipulation check. As participants finished the study, they handed it in in exchange for a debriefing statement and 30 minutes of research credit.
Results
A MANOVA was conducted to analyze the hypotheses that the gender of the Facebook profile would influence the impressions formed, and that profiles containing slang would produce different perceptions of personality traits than the profiles with proper grammar. In order to test the hypothesis that Facebook profiles with slang would be perceived as less intelligent than profiles without slang, a MANOVA was conducted using slang and no slang profiles as the levels of the independent variable and all the personality traits as the dependent variables, F(1, 86) = 84.952, p < .001, η2=.497. Overall, slang use had a significant effect on all personality traits at the multivariate level, Wilks’ Lambda = .296, F(15, 72) = 11.441, p < .001.
At the univariate level, many personality traits were significantly affected by slang usage. Profiles with proper grammar had significantly higher ratings of politeness, competency, seriousness, confidence, honesty, creativity, and trustworthiness, and were rated as more goal oriented. These profiles also had significantly lower ratings of outgoingness, aggressiveness, sociability, and were rated as less judgmental (see Tables 1 and 2).
A MANOVA was conducted to analyze the hypothesis that the profile’s gender would impact the participants’ personality ratings of the profile. A manipulation check filtered the data so only the results of students who correctly reported the profile gender were used. Because of this manipulation check, 20 participants were excluded from the analysis of this hypothesis. Gender caused no significant differences across the dependent variables at the multivariate level, Wilks’ Lambda = .775, F(15, 72) = 1.009, p = .461.
Further analysis at the univariate level revealed statistically significant differences between the perceived friendliness and politeness of male and female profiles where male profiles were perceived as less friendly, F(1,66) = 5.158, p= .026, η2 =.071 , and less polite, F(1,66) = 5.077, p= .028; η2 =.093 (see Table 3). There was no significant difference between the profile gender and the other personality attributes.
Discussion
The hypothesis, which was that slang usage would significantly reduce intelligence perceptions of the Facebook profile user, was supported. The results of the experiment indicated that profiles using phonetic abbreviations influenced the perceptions of personality traits. The other hypothesis, which was that gender would significantly alter ratings of intelligence, was not supported.
Phonetic abbreviations lowered the participants’ impressions of the intelligence of the profile owner. Initialisms, such as “lol” instead of “laugh out loud,” and internet slang, are conditioned stimuli that cause lower levels of assumed competence and intelligence. Generally, people associate the use of such abbreviations with laziness. For instance, if someone is not willing to entirely write out a word, he or she may be perceived as putting in less effort (Jessmer & Anderson, 2001). Additionally correct spelling and grammar is a conditioned stimulus for a higher education level and a more extensive vocabulary. Proper spelling and grammar are indicative of education, and are assumed skills that are necessary for success in higher education. Consequently, profiles utilizing proper grammar and spelling were perceived as more goal-oriented, another of the personality traits rated by participants. Profiles that utilize phonetic abbreviations also produced decreased ratings of creativity, a personality trait associated with intelligence. Perhaps the participants view a high level of intelligence as necessary for creativity and inventiveness.
At the univariate level, the slang-using profiles produced lower perceptions of politeness and seriousness. The reasons behind the politeness results are relatively straightforward; when one wants to convey respect and when one is communicating with those of higher statuses, proper grammar is considered good manners. For instance, if a student were to communicate with a professor, he or she would most likely try to use correct spelling and eliminate grammatical errors. As a result, more grammatical messages seem more polite, and ungrammatical messages seem impolite. The results regarding seriousness are less straightforward. One explanation could be that those who write out grammatically correct messages have invested more energy in them. These individuals may be perceived as more serious because they are concerned with how each word is spelled and the clarity of their message, while those who employ abbreviations such as “lol” and “jk” do not seem as concerned. Thus, the findings regarding seriousness may be explainable due to the participants’ views of how much effort and thought they believed were invested in the Facebook profiles.
Ratings of trustworthiness were also found to be higher for the profiles that used no slang. Because proper grammar is evidently seen as requiring more effort, more intelligence, and more motivation, it does seem logical that participants would rate those who use proper grammar as more trustworthy. Proper grammar causes people to be perceived as less lazy. Individuals would most likely be hesitant to trust someone who may not understand them or may be too lazy or careless to do a task entrusted to them. This logic explains the results for honesty as well, since trustworthiness and honesty are closely linked. Like trustworthiness, profiles with no slang were also rated as more honest than those using slang.
Slang usage also caused certain personality traits to be rated higher. The profile owners who used slang were perceived as more social, outgoing, aggressive, and judgmental than the profile owners who used correct grammar. The result for sociability could be explained by assuming that people who use more modes of computer mediated communication, or who use them more frequently, are more likely to use abbreviations. Thus, abbreviation use could be considered an indicator of an active social life, and therefore those who use them are perceived as more social and outgoing. Both male and female profiles with slang were also rated as more aggressive than profiles without slang. Previous research has found that individuals assume polite emails to be written by women (Jessmer & Anderson, 2001), a finding that in turn links men to impolite and, presumably, badly written emails. This link between men and impoliteness can be extrapolated to encompass slang, as well. Perhaps participants in the present study unconsciously associate slang with masculinity. Because men are stereotypically more aggressive than women, profiles using slang could have been rated as more aggressive because slang usage is associated with masculine characteristics.
The results from the present study also indicated that profile owners who use slang are considered more judgmental. This could be due to the influence of the media on our participants (and the greater high school and college-age populations in general). The media contributes to the idea that phonetic abbreviations indicate negative characteristics. Commercials, television shows, and movies like Mean Girls (2004) typically show a “Bitchy Teenage Girl” archetype; the girls are shown texting and gossiping with their friends, and they generally seem shallow and immature. Although it is not always explicit, it may be assumed by some that these girls use abbreviations as they use computer-mediated communication and texting. Consequently, those who use abbreviations may be perceived as judgmental. In addition, there are and increasing number of news stories about cyber-bullying, whereby one student is ostracized and ridiculed by his or her peers over the internet. Again, the use of abbreviations is not explicit in these news stories, but the assumption here may also be that those who use abbreviations are more frequent users of computer mediated communication. The links between these media influences are presumably enough for participants in the present study to form the impression that the profile owner who uses slang is more judgmental.
The hypothesis that profile gender would alter ratings of intelligence was not supported. At the multivariate level, there was no significant effect of gender on the dependent variables; however, univariate results revealed that the politeness and friendliness of the profiles were influenced by gender. The research of Jessmer and Anderson (2001) showed that participants rated polite messages as a product of a female author, a conclusion that was supported by the present study. Female Facebook profiles were rated as more polite than male Facebook profiles, regardless of slang use. Other research supports these findings as well. For instance, there is a stereotype that women use more polite and mild mannered language, while men have less polite, less mild mannered, and more aggressive language (Leaper & Ayres, 2007). Gender also influenced how friendly the profile owner seemed. Perhaps due to a benevolent sexism stereotype that women are kinder than men (Eagly, 1994; Haddock & Zanna, 1994), female profiles were rated as friendlier than male profiles.
There were a few confounding variables that may have affected the results of the present study. For instance, there was a technical difficulty with the projector during the second session of the study. There was a brief delay in viewing the profile, and although students were able to see the entire profile initially, they could only see the ungrammatical text while they completed the survey; they had all the necessary information but could not see the generic Facebook silhouette picture. Additionally, the profiles had a quotation from Helen Keller. In the ungrammatical, slang using profile, the abbreviation “hahaha” followed the quotation. Some students may have been offended or found it comical, which could have changed their view of the profile owner.
Another confounding variable was a single typo in the grammatical profile. The letter “i” should have been capitalized when it was not. The typo may have caused us to find fewer significant differences than we would have otherwise; however, it seems unlikely that this mistake caused any significant difference in our results.
The present study also suffers from the “college sophomore problem.” The convenience sample of psychology students primarily consisted of female, Caucasian, sophomores. The unbalanced nature of this sample may have threatened the ecological validity of our research. Despite this lack of diversity within the sample, people of different populations could conceivably react similarly to the profiles. The sample in the present study consisted of college age students who use texting and abbreviations themselves. If future research sampled an older population that uses fewer abbreviations, the participants might be less used to the slang and therefore more likely to form judgments. However, there was a high level of mundane realism due to the screen shots of real Facebook profiles presented to the students, and there was a high level of control due to identical messages, the difference being only the presence or absence of abbreviations.
A final confounding variable is a gender bias that could have caused the female profile to be perceived as more masculine. In attempting to keep the profiles as controlled as possible, the same information was given in both the male and female profiles. Therefore, both the male and the female profile owners played basketball. Whereas perceptions of men are not altered by whether or not they play basketball, there is a bias that causes people to perceive women who play basketball as more masculine than women who do not (Csizma et al., 1988).
Future research should address the other aspects of Facebook profiles that may contribute to impression formation. For example, people’s likes, dislikes, photographs, and wall posts could be a source for judgment.
Future research could also investigate how long a first impression based on slang on Facebook lasts, and how it affects two people’s face-to-face interaction. An individual may treat another differently if he or she has pre-formed attitudes based on his or her use of phonetic abbreviations. Research should also address the use of phonetic abbreviations in daily conversation, and how vocal use of the abbreviations may influence people’s impressions of others. Spoken abbreviations may seem more salient due to their minimal use in conversation, and cause for more extreme attitudes to be formed about the individual speaking them. Also, more research investigate the importance of dialects, accents, and slang that pertains to specific population groups in the formation of impressions outside that group.
A content analysis could also be conducted regarding media portrayals of phonetic abbreviations. Commercials, advertisements, television shows, and movies could be analyzed in order to study how individuals using abbreviations are perceived. The media is likely contributing to negative stereotypes associated with abbreviation usage.
The practical and theoretical implications of the current study are vast. One hundred percent of our participants had a Facebook account, which implies a similar number within the larger population. With such a large portion of the population on Facebook, it is extremely important to understand how social networking sites are altering perceptions of people. Forms of computer-mediated communication are steadily becoming more integrated into how we talk with one another. People are tempted to use abbreviations to facilitate their messaging; however, based on the results of this study it can be inferred that they will be perceived differently after doing so. If an individual desires to seem social, outgoing, aggressive and judgmental then using abbreviations may be appropriate. However, it is imperative to avoid abbreviations and improper grammar when one wants to be perceived as intelligent, competent, creative, goal-oriented, polite, serious, honest and trustworthy.
References
- Back, M. D., Stopfer, J. M., Vazire, S., Gaddis, S., Schmukle, S. C., Egloff, B., & Gosling, S. D. (2010). Facebook profiles reflect actual personality, not self-idealization. Psychological Science, 21(3), 372-374. Retrieved from http://ovidsp.ovid.com/ovidweb.cgi? T=JS&NEWS=N&PAGE=fulltext&D=psyc&AN=2010-09431-011
- Biernat, M., & Vescio, T. K. (2002). She swings, she hits, she's great, she's benched: Implications of gender-based shifting standards for judgment and behavior. Personality and Social Psychology Bulletin, 28(1), 66-77. Retrieved from http://ovidsp.ovid.com/ovid web. cg i?T=JS&NEWS=N&PAGE=fulltext&D=psyc4&AN=2001-09711-006
- Csizma, K. A., Wittig, A. F., & Schurr, K. T. (1988). Sport stereotypes and gender. Journal of Sport and Exercise Psychology, 10(1), 62-74.
- Eagly, A. H. (1994). Are people prejudiced against women? Donald Campbell Award invited address, American Psychological Association convention.
- Gill, A.J., Oberlander, J., Austin, E. (2006). Rating e-mail personality at zero acquaintance. Personality and Individual Differences, 40 (3), 497-507. doi: 10.1016/j.paid.2005.06.027
- Gosling, S., Gaddis, S., & Vazire, S. (2007). Personality impressions based on facebook profiles. Paper presented at the Proceedings of the International Conference on Weblogs and Social Media, 26-28.
- Haddock, G., & Zanna, M. P. (1994). Preferring “housewives” to “feminists.” Psychology of Women Quarterly, 18, 25-52.
- Hancock, J. T., & Dunham, P. J. (2001). Impression formation in computer-mediated communication revisited: An analysis of the breadth and intensity of impressions. Communication Research, 28(3), 325-347. Retrieved from http://ovidsp.ovid.com/ovidweb.cgi?T=JS&NEWS=N&PAGE=fulltext&D=psyc3&AN=2002-17248-004
- Jessmer, S. L., & Anderson, D. (2001). The effect of politeness and grammar on user perceptions of electronic mail. North American Journal of Psychology, 3(2), 331-346. Retrieved from http://ovidsp.ovid.com/ovidweb.cgi?T=JS&NEWS=N&PAGE=fulltext&D=psyc3&AN=2001-11562-013
- Leaper, C., & Ayres, M. M. (2007). A meta-analytic review of gender variations in adults’ language use: Talkativeness, affiliative speech and assertive speech. Personality and Social Psychology Review, 11, 328-363.
- Lewandowski, G. W. J., & Harrington, S. (2006). The influence of phonetic abbreviations on the evaluation of student performance. Current Research in Social Psychology, 11(15), 215-226. Retrieved from http://ovidsp.ovid.com/ovidweb.cgi?T=JS&NEWS=N&PAGE=fulltext &D=psyc5&AN=2007-00643-001
- Marcus, B., Machilek, F., & Schutz, A. (2006). Personality in cyberspace: Personal web sites as media for personality expressions and impressions. Journal of Personality and Social Psychology, 90(6), 1014.
- Paul, D., & Smith, J. L. (2008). Subtle sexism? examining the vote preference when women run against men for the presidency. Journal of Women, Politics & Policy, 29(4), 251-476.
- Share, D. L., & Silva, P. A. (2003). Gender bias in IQ-discrepancy and post-discrepancy definitions of readings disability. Journal of Learning Disabilities, 36(1), 4-14. Retrieved from http://ovidsp.ovid.com/ovidweb.cgi?T=JS&NEWS=N&PAGE=fulltext&D=psyc4&A N=2003-04170-003
- Sierra, J., Santos-Iglesias, P., Gutiérrez-Quintanilla, R., Bermúdez, M., & Buela-Casal, G.. (2010). Factors Associated with Rape-Supportive Attitudes: Sociodemographic Variables, Aggressive Personality, and Sexist Attitudes. The Spanish Journal of Psychology, 13(1), 202-9. Retrieved from ProQuest Psychology Journals. (Document ID: 2092540951).
Table 1
F values and significance of grammar on personality traits.
Personality Traits F(1,86) p η2
Outgoing 12.024 .001 .123
Competence 21.936 <.001 .203
Politeness 48.384 <.001 .360
Seriousness 44.558 <.001 .341
Confidence 4.300, .041 .048
Aggressiveness 11.695 .001 .120
Intelligence 84.952 <.001 .497
Honesty 6.322 .014 .068
Social 10.1127 .002 .105
Judgmental 22.767 <.001 .209
Goal-Oriented 28.935 <.001 .252
Creative 12.377 .001 .126
Trustworthy 46.211 <.001 .350
Table 2
Mean ratings of personality traits in slang and no slang profiles.
Slang No Slang
Personality Traits N M SD N M SD
Outgoing 37 5.784 1.182 51 4.892 1.197
Competence 37 3.378 1.320 51 4.431 0.781
Politeness 37 2.946 0.970 51 4.471 1.046
Seriousness 37 2.027 0.726 51 3.647 1.339
Confidence 37 5.487 1.387 51 4.882 1.321
Aggressiveness 37 4.541 1.592 51 3.373 1.574
Intelligence 37 2.487 1.044 51 4.275 0.777
Honesty 37 3.865 1.251 51 4.471 1.007
Social 37 5.865 1.110 51 4.902 1.578
Judgmental 37 4.405 1.257 51 3.177 1.144
Goal-Oriented 37 2.676 1.270 51 4.137 1.250
Creative 37 2.568 1.345 51 3.608 1.387
Trustworthy 37 2.891 1.197 51 4.373 0.848
Table 3
Means and standard deviations for friendliness and politeness of male and female profiles.
Male Female
Personality Traits N M SD N M SD
Friendliness 49 4.778 1.149 40 5.344 1.231
Politeness 49 3.528 1.006 40 4.188 1.271