User Descriptions and Interpretations of Self-Presentation through Facebook Profile Images

Michele M. Strano
Bridgewater College, Bridgewater, Virginia, USA

Abstract

This article reports the findings of an open-ended qualitative survey about the choices people make when choosing profile images on Facebook. The results demonstrate that women tend to change their profile image more often and to emphasize friendship in the images they choose to display. Although both men and women are equally likely to display images of family and romantic relationships, the interpretations of such images may vary by gender. In addition, as one of the first Facebook studies to include users over 22 years old, this study provides comparative data demonstrating that older users are less likely to change their profile images frequently and more likely to display images of themselves alone. This study corroborates previous findings about the nature of self-presentation in online networking sites and offers new insights into how such impression management might vary according to age and gender.

Key words: Facebook, profile images, self-presentation, identity, digital photography

Introduction

As the popularity of online social networking sites like Facebook grows, so do concerns about the impact of such sites on the impression management and relationship practices of today's youth. As Buffardi and Campbell (2008) describe, some worry that the online environment encourages narcissistic behavior, while others show concern that meaningful friendships are replaced with a shallow list of acquaintances (Garreau, 2008). However, recent research has begun to show that such claims of social danger may be overstated, as users of networking sites have more complex understandings of self-presentation and relationships than researchers and adult non-users may have at first imagined (e.g. Zhao, Grasmuck, & Martin, 2008 and Raacke & Bonds-Raacke, 2008). In addition, sites such as Facebook, which began as a networking site utilized on college campuses, have expanded their availability in recent years, resulting in over half of all Facebook users being outside of college and the fastest growing demographic consisting of those 25 years old and older (Statistics, n.d.). Existing research, however, has concentrated on the self-presentation practices of adolescents and college-aged students. Impression management has been studied through content analysis (e.g. Siibak, 2007; Zhao, Grasmuck, & Martin, 2008), close-ended questionnaires (Raacke & Bonds-Raacke, 2008), and small-sample interviews (Livingstone, 2008). This study aimed to expand our understanding of how identity is managed in the context of online networking sites, first, by including adult users over the age of 22 in the sample and, secondly, by utilizing an open-ended online survey that was then quantified in order to compare answers across age and gender groups. This approach grounded the categories of analysis in user-defined perceptions while allowing comparisons between subgroups within a sample that is larger (N = 427) than qualitative interviewing can typically achieve.

Literature Review

While other studies have looked at Facebook profile pages as a single text including visual and textual elements, this study focuses on a particular element common to all Facebook pages: the profile image. This image appears in the upper left-hand corner of a profile page and acts as a primary identity marker for a user's profile. See Figure 1 for an example of a profile image in the context of a user profile page.

fig
Figure 1: Profile image in context

Although additional photographs may be posted of the user under the "Photos" tab, the profile image arguably acts as the most pointed attempt of photographic self-presentation on the Facebook profile. In essence, this image "stands in" for the user's body in this virtual environment. When users search for each other on Facebook, the profile image is displayed, generally even on profiles that are otherwise privacy protected (although it is possible to set the privacy settings such that the profile image will not display during searches). When a user posts a comment or sends a message to another user, his or her profile image is displayed alongside the text they write. While many users choose photographs of themselves and others as their profile images, some choose photographs of iconic objects or pets as well as drawings or graphic designs. If a user does not choose a personalized image to display, Facebook inserts this generic "body" in the places reserved for profile images:

fig

Facebook profile images can be seen as a form of "implicit" identity construction (Zhao,Grasmuck, & Martin, 2008) in which users display personal characteristics through images. Goffman (1959) distinguishes between the expressions we "give" through explicit verbal communication (like that which might be offered on Facebook user's "Info" page) and implicit expressions "given off" through visual appearance. Zhao et. al. (2008) find through a content analysis of 63 college student Facebook accounts that users rely heavily on implicit modes of self-presentation, with the average number of photographs per account at 88.4 photos. In addition, while 33.3% of their sample blocked non-friends from viewing photos on other parts of their site, over 90% of the accounts displayed their profile images publically.

To fully appreciate the potential power of photographs to act as a form of self-presentation, we must remember that despite our claims of the evidentiary force of photographs (Chalfen, 2002), such images remain but one rendition of reality. Often we talk about photographs as if they are traces of reality that have "captured" or "preserved" our past; less frequently will we acknowledge the shooting and editing practices that have shaped a photograph into an idealized image representing social norms about desirable personal characteristics and socially-accepted notions of family, gender, romantic relationships, parenthood, etc. (Chalfen, 2002; Strano, 2001). Some argue that the shift from analogue to digital photography may refocus the way we think about photography, moving us from the concept of "taking pictures" to "making pictures" (Chalfen, 2002) since digital photography offers increased power over the editing process. Alterations that once required the expertise of a professional in the analogue age can now be achieved by anyone with a computer and a photo-editing program. As Van Dijck (2008) asserts, digital photography may offer the possibility of a stronger emphasis on the role photography plays in identity formation, with a de-emphasis on traditional notions of memory preservation. In addition, the identities constructed in digital environments may be more dynamic than the print-based display contexts of analogue photography, since the technology eases the process of substituting one photograph for another (Slater, 1995).

However, despite the ability to potentially reinvent oneself through the manipulation of digital images, previous research has shown that users of "nonymous" (Zhao, Grasmuck, & Martin, 2008) networking sites tend to present profiles that are somewhat true to their offline identities, although positive traits may be emphasized and flaws may be omitted. For example, Yurchisin, Watchravesringkan, and McCabe (2005) demonstrate that users of online dating networks tend to emphasize their "hoped for selves" on their profiles, "stretching the truth" a bit in order to seem more desirable. However, as Zhao, Grasmuck and Martin (2008) and Ellison, Steinfield and Lampe (2007) demonstrate, Facebook identities are often grounded in offline relationships, thus bounding the extent to which users can create identities that would result in offline social sanctions. Zhao Grasmuck and Martin (2008) argue that, in contrast to some anonymous online environments, Facebook is not the "dreamland for deviant behaviors" (p. 1831) that some critics fear.

Instead, Facebook seems to cultivate a culture in which group identity is emphasized. Users overwhelmingly report that keeping in touch with friends is the primary gratification sought from Facebook (Raacke and Bonds-Raacke, 2008; Ellison, Steinfield, & Lampe, 2007). In addition, viewers perceive users as being more popular, sexy, attractive and self-confident if their profile includes a larger number of friends (Reese, Zieger-Behnken, Sundar, & Kleck, 2007), indicating that even simple numeric representations of group affiliation may serve as a tool of impression management. Likewise, Zhao Grasmuck and Martin (2008) found that implicit claims of group identity are apparent in Facebook profile images with 38.1% of the accounts they coded containing a profile photograph with two or more persons.

While previous research has shown that self-presentation differs according to environment (e.g. anonymous vs. nonymous environments), there has also been some research demonstrating that personal demographics may impact the ways in which users construct online identities. Specifically, age and gender act as strong identity markers in offline contexts, with strong social norms communicating how to "act your age" or "be a man." It is not surprising that in nonymous online environments, gender- and age-specific social norms may also be at work.

Gender Differences in Online Self-Presentation

While Raacke and Bonds-Raacke (2008) did not find gender differences in the most popular uses and gratifications for social networking sites (i.e. keeping in touch with friends), they do note several differences in the way men and women use online social networking sites. One finding that is applicable to this study is that women were found to change the appearance of their website more often than men. Although this finding applied to the website as a whole, and not the profile image specifically, it does prompt us to question whether women might create more dynamic identities through their profile images than men.

Perhaps more directly applicable to this study, Siibak (2007) found that female users of the dating site Rate were more likely to display a profile photo of themselves smiling than men (65% versus 24%) and more likely to choose a photograph at a close personal range, showing only the head and shoulders (28% versus 15%). In addition, Siibak reports that women tend to display themselves in more seductive poses and wear clothing that emphasizes their sexuality. However, we cannot automatically expect that these findings will apply to Facebook, an online environment that is significantly different than a dating site in that users are attempting to present an identity that appeals to both genders, with the goal more often being friendship than romantic attachment. Indeed, in Raacke and Bonds-Raacke's study, none of the women and only 16.3% of the men reported using MySpace or Facebook for dating purposes.

Age Differences in Online Self-Presentation

Little research has been done that compares Facebook self-presentation across age. Indeed, all of the studies reviewed here have focused on adolescents or college students. However, one interesting finding related to age comes from Livingstone's (2008) interviews with 16 adolescents ranging in age from 13 to 16 in the Greater London area. She found that the younger participants constructed their identities through visually elaborate and individualized profiles, while older adolescents preferred an aesthetically plain profile appearance that highlighted social connections through "links to others' profiles and by posting photos of the peer group socializing offline" (p. 402). This finding raises questions about how older users of Facebook might manage their profile images differently than younger users, perhaps using group images more than photographs of themselves alone.

Goals of the Study

In summary, the goal of this study was to further investigate user interpretations of their self-presention on Facebook, specifically focusing on their choice of profile image. The study began with a broad set of questions investigated through a qualitative survey, including: (1) What patterns are there in the types of images posted? (2) What meanings do users ascribe to images? (3) What criteria are used in choosing images? (4) What prompts users to change their images? (5) Do practices and interpretations vary by age and/or gender? From the answers to the survey, categories were identified and the qualitative responses were quantitatively coded, thus allowing more valid comparisons between gender and age categories. This somewhat unusual mixed-method approach allowed user interpretations to be central in the analysis of profile image practices.

Methods

An online survey, consisting primarily of open-ended qualitative questions, was used to gather data about Facebook profile image posting practices and interpretations. The survey (see survey questions below) asked Facebook users to describe their current profile image and comment on why they picked that image to represent their identity. The survey was distributed to various email lists and then utilized a snowball sampling method by asking participants to forward the survey to their friends and acquaintances.

While content analysis would perhaps most effectively map the range of image types in Facebook profile photographs, the main focus of the study consists of interpretive questions that can perhaps be best addressed through in-depth interviews. Indeed, as described in the conclusion of this paper, the second phase of the project will use in-depth interviews to expand on the patterns observed using the current open-ended qualitative survey. While it may seem unconventional to study photography without actually coding photographs, this method has the advantage of relying on user definitions of visual symbols to develop descriptive categories.

Of course, the snowball sampling method used for this study limits the degree to which the results of the study can be generalized to a larger population. However, when compared to the samples used for content analyses of profile pages and in-depth interview studies, the sample compares somewhat favorably. For example, Livingstone (2008) only interviewed 16 adolescents, recruited "through a market research agency" (p. 397); 101 students all from the same university answered Raacke and Bonds-Raacke's (2008) study; and Zhao Grasmuck and Martin (2008) analyzed 63 Facebook accounts, all associated with the same university. So, while this sample is far from randomly selected, the survey does include 427 responses and was distributed via several different email lists in the United States. Thus, it represents a group that is at least as representative as that used for most studies focused on interpretations. In addition, as is described in more detail below, the sample includes a substantial number of users older than college age, thus expanding beyond the samples used for previous studies.

Survey Questions

The questions included in the survey are listed below. For a PDF version of the survey as it appeared online, please email the author. The questions in bold most directly address the research questions identified above, therefore, responses to these questions are used as the basis for the analysis presented in this article.

  1. Please describe in as much detail as you can your current Facebook profile photograph or image. [textbox]
  2. Why did you choose this photograph or image? [textbox]
  3. Please complete this sentence: "I think that people who see this image would describe me as a person who_________."[textbox]
  4. How long have you had a Facebook Account? (choices: less than 1 year, 1 year, 2 years, 3 years, 4 years)
  5. How many times have you changed your profile photograph or image? (choices: 0-5 times, 6-10 times, 11-15 times, 16-20 times, 21-30 times, 31-40 times, 41-50 times, Over 50 times (how many? [textbox])
  6. What usually prompts you to change your profile picture? [textbox]
  7. Is your current profile image significantly different than the types of images you usually put up? If yes, how would you describe the difference? [textbox]
  8. Have you ever removed a "tag" identifying you on a photograph posted by another Facebook user? (choices: No, Yes -why did you remove the tag? [textbox])
  9. Are you a student? (choices: 2-year Undergraduate, 4-year Undergraduate, Graduate or Professional School, Not a student -please specify your occupation [textbox] )
  10.   When you post a profile image, whom do you imagine viewing that photograph? (select all that apply, choices given: My friends, My teachers, My students, My parents, My children, Employers, Strangers, People I have just met, Other (please specify [textbox])
  11.   Gender (choices: Female, Male)
  12.   Age (choices: 18-21, 22-30, 31-40, 41-50, 51-60, Over 60 (please specify age [textbox])
Respondents

The online survey was at least partially answered by 427 Facebook users. Although nationality was not tracked by the survey, we can assume that most of the respondents are American based on the starting points for distributing the survey and the allusions to place of residence in the survey answers. As described in the conclusion, future phases of this project will be purposefully disseminated internationally and place of residence, nationality and ethnicity data will be collected.

Figure 2 shows the age distribution of the sample. The high proportion of 18-24 users reflects the overall age distribution of Facebook users (drigotti, 2007). One difficulty of using an online survey to study Facebook usage is that U.S. Internal Review Board guidelines require researchers to obtain permission from the parents of minors (those under 18 years old) who participate in a research study. Since such permission is difficult to obtain reliably via an online channel, all participants were required to attest to being legal adults before they could participate in the study. Therefore, a large segment of Facebook users (about 18% according to drigotti) were excluded from this survey.

fig
Figure 2: Age Distribution of Respondents

Figure 3 shows the gender distribution of the sample. The high proportion of women in the sample is in part due to the fact that about 60% of U.S. Facebook users are female (Arrington, 2007). Still, the gender split of the sample is significantly different than that of Facebook overall (p=5.98E-06). For some reason, women seem to have been more willing to answer the survey. Figure 4 further investigates this bias by comparing the gender split across age groups. The gender split evens out for the oldest age group. Whether this is because the age distribution of Facebook users is different for different age groups or because men over 31 years old were more willing to fill out the survey is not clear from the available data.

fig
Figure 3: Gender Distribution of Sample

fig

(Superscripts denote significant differences between age groups: age groups with the same letter are not significantly different.)

Figure 4: Gender Distribution by Age of Sample
Approach to Analysis

Sixty-one qualitative categories were generated from the answers to the four main open-ended survey questions. After categories were identified, all responses were coded by two coders using a binary coding scheme that marked each category as "Present" or "Absent" in a given participant response. Intercoder reliability was calculated for each category, with kappa scores exceeding .70 for 49 of the 61 categories (80%). Only two of the categories with kappas under .70 were kept in the analysis, since the discrepancy was easily explained by a simple difference in the types of responses each coder was including. For all categories, coders discussed discrepancies and reached agreement for the final coding. At that point, categories with less than 15 observations were also dropped from further analysis, since the responses in these categories did not seem to be significant trends in the data. This resulted in 33 categories that were used for the main analysis of the study. Figure 5 summarizes the number of categories coded for each question, the kappa scores for each, and the number of categories that remained in the analysis.

fig
Figure 5: Summary of Coding Category Generation

Results

Common Responses

Figure 6 summarizes the categories of responses identified for each of the four main open-ended survey questions. Each of the categories is described in this chart, but the key words that will be used as abbreviations in subsequent figures are highlighted in blue. Abbreviations for the questions are included in parentheses. Note that these categories are not mutually exclusive, so if a respondent described their image as a "Head Shot," for example, it was also coded as an "Alone" shot; however, not all Alone shots are Head Shots. Likewise, in response to the Why Photo questions, participants often gave several reasons why they chose a particular image, resulting in several categories being coded as present for a single response.

fig
Figure 6: Most Common Responses

A few categories that were dropped from analysis because they had such low counts are worth mentioning because the lack of such answers may be unexpected to some. These include: people with no photo displayed (1.6%); images of support icons, like images of ribbons after the Virginia Tech shootings (1.2%; people who said they chose a photo to represent a change in their appearance (1.7%); people who said they chose a photo because it made them look professional (2.1%) or like they had a skill (2.1%); people who thought their profile image made them look intelligent (1.2%) or normal (1.2%).

Differences by Gender and Age

Chi square analyses were run on Gender and Age in order to identify patterns in the collected responses. The results of these analyses are summarized in Figures 7 and 8.

Note that binary logistic regression was initially attempted in order to test a model that included Gender, Age and an interaction (Gender x Age) variable. However, with the low frequencies for most of the response categories, the model lost considerable statistical power with the addition of each variable. As described in the conclusion of this article, a future phase of this project will use a large-scale quantitative survey that should allow for an analysis of interactions between Gender and Age.

fig
Figure 7: Significant Differences by Gender

The Age variable was collapsed into three categories, 18-21-year-olds (those most likely to be presently in college), 22-30 year olds (young professionals that may have been introduced to Facebook while still in college), and 31+ year olds (who were most likely out of college when Facebook became popular. Figure 8 summarizes the significant results of the Chi Square analysis of each response category by age. Superscripts denote significant differences within each category: percentages with the same letter are not significantly different. Note that none of the categories in response to the Why Choose question revealed a significant difference by age.

fig
Figure 8: Differences by Age

How Dynamic is the Display Context?

Frequencies were calculated for answers to the "How many times have you changed your profile photograph or image?" question. Unfortunately, it was not anticipated that there would be a sizable number of people who had never changed their photograph, so the survey instrument only provided the category "0-5 times" as an option. However, 34 respondents stated that they had never changed their profile image in response to the "Prompts Change" question. Thus, Figure 9 summarizes the frequency of each answer to Times changed with 34 responses subtracted from the "0-5 Times" category in order to estimate the difference between never changing the profile image and having changed it at least once.

fig
Figure 9: Number of Times Profile Image has been changed

Although the survey did not ask respondents how often they changed their profile image per a specified length of time (in large part because without any prior data it was difficult to determine whether the appropriate question would have been "per week," "per month" or "per year"), an estimate was calculated by using the midpoint of each "Times Changed" category (dropping the "over 50 times" responses) and dividing that number by the number of years (.5, 1, 2, 3, or 4) that a respondent has been a Facebook user. The average number of times that respondents change their profile image, according to this rough estimate, is 6.89 times per year (SD=6.49, Min=.62, Max=45.00).

Discussion

As Slater (1995) predicts about digital photography in general, Facebook profile images constitute a dynamic display context. Our estimate that users are changing their profile images seven times a year is well-beyond what we would see in analogue-based display contexts. People rarely change photographs in albums (artifacts intended to act as static relics of past moments) and few people change framed photographs seven times a year. While the technological environment alone may encourage more dynamic display practices, the fact that users who are 31 years or older are the most likely to have never changed their profile image suggests that age may also be a factor in how the display environment is used. Corroborating this idea is the fact that older users are also less likely to say that they change their profile image because they value having a recent image or they get bored with the old image. However, older users are also the most likely to be new users of Facebook, with 60.7% of 31+ year olds having joined Facebook within the last year (as compared to 9.0% of 18-21 year olds and 13.2% of the 22-30 year olds, p=.000), raising the question of how quickly older users will adapt to the social norms of the Facebook display environment which seems to expect a certain degree of change. Perhaps the most direct evidence of this norm comes from respondents who reported that they change their photograph because others pressure them:

(In response to "What usually prompts you to change your profile picture?")

Since only nine people responded with these types of explanations, this category did not remain in the extended analysis of the study, but such answers suggest that there may be an underlying social pressure to change your profile image that other respondents simply did not think to mention. Future phases of this project will follow up on this idea and more systematically measure user expectations of how often a profile image should be changed.

Within this dynamic display context, women are more likely to report that they change their photograph to make it more recent. This corroborates Raacke and Bonds-Raacke's (2008) finding that women are more likely to change the appearance of their profile as a whole. This study does not provide adequate data to explain such a difference. It could be that in a culture that puts so much emphasis on women's looks female users are more likely to be dissatisfied with their appearance. The fact that women are more likely to say they chose a profile image because it makes them look attractive suggests that this explanation may have some merit. Siibak's (2007) finding that women are focused on displaying an idealized image of female beauty on dating sites also suggests that women may change their profile image more often in an effort to better approximate a beauty ideal. Perhaps related to the finding about attractiveness, this study also found that women are more likely to describe themselves as smiling in their profile image (consistent with Siibak's findings) and to say that their image makes them look happy and fun-loving. It seems that female Facebook users focus on constructing identities that emphasize light-heartedness and beauty.

However, an alternative explanation for why women change their profile image more often may be hidden in the results showing that they are also more likely to be pictured with a friend and more likely to depict a special occasion in their profile image. Perhaps women focus more strongly than men on the construction of group identity in online contexts. Indeed, men were also more likely to report that their profile image made them look unique, an indicator that individualism was an important image to portray. Perhaps women change their profile image more often because they are trying to represent a multitude of social relationships and activities, instead of a more static individual identity.

Still, both genders are equally likely to have a photograph of themselves with family or with a significant other, suggesting that certain types of group identity are important for all Facebook users. However, the way that users interpret the identity claims they are making with their photos may still differ by gender. A few examples may help clarify this point:

(from a female respondent)
(Describe) It is a picture of me and my boyfriend, he is kissing me on the cheek

(Why photo) I liked it, it shows who I am and my boyfriend, which is like my best friend

In this example, the user describes the photograph as showing who she "is" by showing that she is in a strong romantic relationship (he is kissing her, he is her best friend). Of course, an in-depth interview is really needed here to prompt the user to explain exactly what she means by "who I am." In a second example, the user points to a social norm for women to photographically define who they are in terms of their boyfriend by seeming to resisting such a norm:

(from a female respondent)
(Describe) It is a side-view of myself looking up into the face of my boyfriend, who is just barely visible in the picture. I am dressed up ready to go to a dance. My hair is curled and my dress in full-length. I cropped the original picture, in which the two of us are equally present, so that I am the center of the picture.

(Why photo) I chose this photo because it is current and because in it I look nice. I like the angle and my expression. I think I look pretty and happy, which is the impression I'd like to give to others. I also like that my boyfriend is present, because he is important in my life, but I don't want him to be the focus of the picture. I don't want to give the impression that I depend on him too much or that I think he defines my life.

This user seems to perceive a danger that by showing too much of her boyfriend in the profile image, she would be communicating that her identity is defined by his - an interpretation she presumably has of the relationship photos of others. In contrast, consider these quotes from male users describing and explaining their pictures of themselves and their girlfriends:

(from a male respondent)
(Describe) A photo of myself and my wife (girlfriend at the time) in the snow in front of my academic place of work

(Why photo) I was mainly looking for a good picture of the two of us, mainly to clarify my non-single status for people who were wondering and secondarily to demonstrate to her my commitment.

****

(from a male respondent)
(Describe) It is a picture of my girlfriend and me before a concert.

(Why photo) I wanted a picture of the two of us, and this was one of the few she approved.

Notice that in both of these examples the men describe their motivation for posting the photographs in terms of directly impacting their relationship with their girlfriends rather than projecting as an identity. In the first example, the reference to "non-single status" could be interpreted an identity claim, but since in response to the question about what type of person people would assume he was from his photo this user replied "happy and taken," this seems to be more of a practical way of warding off advances than really asserting identity. Of course, an in-depth interview would help to clarify these answers. However, it seems possible from this limited data, that the display of relationships in profile images may serve different functions for male and female users.

In relation to age differences, it is perhaps most interesting that users in the 22-30 years age group are the most likely to say they change their profile image to reflect events and activities in which they are involved. They are also the most likely to say that their image represents a special occasion and the least likely to be pictured alone in their photograph. As users age, we might expect this movement to more group-oriented identities based on Livingstone's (2008) finding that older adolescents tend to display more social connections, while younger uses display more flashy and individual profiles. However, what is unexpected is that the 31+ users resemble the younger users in that they are equally likely to be pictured alone and to have a head shot. In addition, older adults are the least likely to be pictured with a friend and the most likely to say they have edited their profile image for a unique visual effect. Coupled with the finding reported above that 31+ users are the most likely to have never changed their profile image, these results seem to suggest that older users are constructing more unique individual identities than people in their twenties, making them look more like younger adolescents. There are a couple of possible explanations for this trend. First, older users are more likely to be new users of Facebook, given the history of availability of the site. Indeed, 60.7% of users 31+ years of age reported being users for less than a year. Perhaps older users will start to act more like the 22-30 year olds as they become more familiar with the site. However, the other explanation is that the emphasis on group identity is a temporary developmental phase and that as the 22-30 year olds age their profile image practices will begin to look those of the older adults. It is not possible to predict which explanation is correct without a longitudinal study over the next decade.

It is interesting that there are no differences across age in the reasons users give for why they have posted a particular profile image. The only difference across gender is that women are more likely to say they put up the image because it makes them look attractive. Perhaps the apparent lack of difference in the reasoning behind image choice indicates that the "why" is intuitive and not easily articulated. Many respondents said vague things about the posted photograph being "good" without much explanation about the criteria that made it good. In-depth interviews are likely to generate more detailed responses about the reasons users choose particular profile images, because of the opportunity to probe initial explanations. However, this data suggests that perhaps we should be careful in assuming too much user agency when we investigate self-presentation in online environments. Just as in offline environments, the degree to which we employ cognitive choice and intentionality in the way we present ourselves may vary between specific contexts. For example, it might be interesting to compare the intentionality of self-presentation on dating sites, where very specific relationship goals may be at stake, to that of online social networking sites like Facebook, where broader goals of maintaining friendships are at work.

Finally, some of the data collected in this study suggests that profile images are not used solely for identity construction. For example, the idea of stimulating the user's own memory was apparent in some responses, such as this user´s description of why she chose a photograph of herself and her husband for her profile image: "It reminds me of how much fun we had that day." The use of the word "me" is interesting here, because it suggests that the user is not thinking about projecting an image to others (not "to show how much fun we had that day"), but to remind herself, thus implying that the profile image serves a memory function for her, much like looking at her own photograph album or framed photograph. This suggests that future research might benefit from re-evaluating whether impression management is the only lens through which profile images might be understood.

Conclusion and Future Study

This exploratory study has shown that Facebook profile images are fertile ground for investigating the practices associated with impression management in online networking environments. While this qualitative survey has provided useful foundation for future work, the results are limited by the lack of opportunity for the researcher to probe responses and by the inability to distinguish between the likelihood of a user to offer an interpretation and more objective measures of agreement with particular interpretations.

The method used in this study is somewhat unusual in that it combined the open-endedness of a qualitative study with a sample size and statistical analysis more typical of quantitative research. While the inferential statistics presented here help us to better understand the trends present in the qualitative responses to the survey, the appearance of a quantitative technique should not be mistakenly construed as representative results. The snowball sampling technique and the grounded development of analytical categories keep the results from representing a true quantitative test that can be generalized to a larger population.

The researcher plans to expand the results of this exploratory study with two more stages of research, one clearly qualitative and one clearly quantitative. The second phase of the study is already underway and employs in-depth interviews with Facebook users across four age groups (15-17 year olds are being included in this phase), allowing the collection of richer data about the interpretations of profile images. In Spring 2009, an online quantitative survey will be launched internationally, focusing on the norms associated with posting profile images identified through the qualitative survey and the in-depth interviews. This large-scale quantitative survey will aim to track differences in perception and practice according to age, gender and culture. A much larger and more representative sample will be used for this final phase of the project.

Acknowledgements

A special thanks to all the Facebook users who answered this survey.

The author wishes to thank her research assistants, Jill Wattai and Daniel Reed for their assistance on the project. This research was funded by a Faculty Research Grant from Bridgewater College, Virginia, USA.

Preliminary results of this study were presented at the 2008 Cultural Attitudes towards Technology and Communication conference in Nîmes, France.

References

Arrington, M. (2007, November 21). Facebook is almost 2/3 women and other stats. TechCrunch. Retrieved from http://www.techcrunch.com/2007/11/21/...

Buffardi, L.E., & Campbell, W. K. (2008). Narcissism and social networking sites. Personality and social psychology bulletin, 34, 1303-1314.

Chalfen, R. (2002). Snapshots "r" us: the evidentiary problematic of home media. Visual Studies, 17, 141-149.

drigotti. (2007, November 26). Facebook stats: Age, gender, education level, political views, and relationship status. Message posted to http://www.freezinghot.com/index.php/20071126/...

Ellison, N.B., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook "friends:" Social capital and college students' use of online social network sites. Journal of Computer-Mediated Communication, 12(4). Retrieved from http://jcmc.indiana.edu/vol12/issue4/...

Garreau, J. (2008, April). Friends indeed?:As we click with more pals online, the idea of friendship multiplies. The Washington Post. Retrieved from http://www.washingtonpost.com/wp-dyn/content/...

Goffman, E. (1959). The presentation of self in everyday life. Garden City, N.Y.: Doubleday.

Livingstone, S. (2008). Taking risky opportunities in youthful content creation: teenagers' use of social networking sites for intimacy, privacy and self-expression. New Media and Society, 10, 393-411.

Mazer, J.P., Murphy, R.E., & Simonds, C.J. (2007). I'll see you on "Facebook": The effects of computer-mediated teacher self-disclosure on student motivation, affective learning, and classroom climate. Communication Education, 56,1, 1-17.

Raacke, J., & Bonds-Raacke, J. (2008). MySpace and Facebook: Applyting the uses and gratifications theory to exploring friend-networking sites. CyberPsychology & Behavior, 11, 169-174.

Reese, C., Ziegerer-Behnken, D., Sundar, S.S., & Kleck, C. (2007). The company you keep and the image you project: Putting your best face forward in online social networks. Paper presented at the 57th Annual conference of the International Communication Association, San Francisco, CA.

Siibak, A. (2007). Reflections of RL in the virtual world. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 1(1). Retrieved from http://cyberpsychology.eu/view.php?cisloclanku=2007072301

Slater, D. (1995). Domestic photography and digital culture. In M. Lister (Ed.), The photographic image in digital culture (pp. 129-146). New York: Routledge.

Statistics of Facebook. (n.d.) Retrieved November 2, 2008 from http://www.facebook.com/press/...

Strano, M. (2001). As time goes by: Ritualized remembering through wedding photography. Unpublished doctoral dissertation, University of Pennsylvania, Philadelphia, PA.

Van Dijck, J. (2008). Digital photography: Communication, identity, memory. Visual Communication, 7, 57-76.

Yurchisin, J., Watchravesringkan, K., & McCabe, D. B. (2005). An exploration of identity re-creation in the context of internet dating. Social Behavior and Personality, 33, 735-750.

Zhao, S., Grasmuck, S., & Martin, J. (2008). Identity construction on Facebook: Digital empowerment in anchored relationships. Computers in Human Behavior, 24, 1816-1836.