Differences in Social Network Site Usage
Who uses SNSs, and are different students equally likely to use the various services available in this realm? The survey included questions about six SNSs: Bebo, Facebook, Friendster, MySpace, Orkut, and Xanga. For each, respondents were first asked to report whether they had ever heard of the site. Next, they were asked to indicate their experiences with it, using the following options: “no, have never used it,”“tried it once, but have not used it since,”“yes, have tried it in the past, but do not use it nowadays,”“yes, currently use it sometimes,” and “yes, currently use it often.”
Overall, 88% of respondents are SNS users, and 74% report using at least one SNS often. Only one student claims not to have heard of any of the six SNSs included on the survey, so non-use is not a result of not being familiar with these services. Rather, despite knowing about such sites, over 12% of the sample does not use any of them.
Table 2 shows the proportion of SNS users by specific site. Facebook is the most popular service among these students, with almost four in five using it, and over half of the overall sample doing so frequently. MySpace is used by more than half of the sample, although just over one-third uses it often. The other four sites (Xanga, Friendster, Orkut, and Bebo, in that order of popularity) are significantly less widespread in this group, with each used by less than 10% of the sample.
Table 2. Familiarity and experience with social network sites among participants (percentages)
| ||Uses it*||Has heard of it||Has never used it||Tried it once, but no more||Used to use it, no longer|
Table 3 reports the demographic breakdown of SNS users, first in the aggregate (second column) and then by site (columns 3–6). Orkut and Bebo are excluded from the table due to their extremely low levels of use in this group.
Table 3. Descriptive statistics for the sample demographics (percentages)
| ||Full sample||SNS users||Facebook users||MySpace users||Xanga users||Friendster users|
|Race and Ethnicity|
| White, non-Hispanic||42.7||43.2||44.9||44.0||20.6||3.0|
| African American, non-Hispanic||7.7||7.4||7.9||8.2||3.2||0|
| Asian American, non-Hispanic||29.6||29.9||31.6||21.3||65.1||93.9|
| Native American, non-Hispanic||1.2||1.1||1.1||1.3||1.6||0|
|Parent’s Highest Level of Education|
| Less than high school||7.4||7.4||6.0||10.0||1.5||0|
| High school||19.0||18.3||17.6||20.1||16.9||8.6|
| Some college||20.1||19.5||18.8||20.9||20.0||11.4|
| Graduate degree||19.1||19.2||20.1||14.1||27.7||22.9|
|Lives with parents||53.1||51.4||48.2||54.5||49.2||58.8|
The differences among the user populations of these services are not particularly pronounced on most variables. Some trends, nonetheless, are notable. First, the percentage of Asian/Asian American users fluctuates considerably, depending on the service. In particular, Asian/Asian American students in the sample are least represented on MySpace, whereas Xanga and Friendster are especially popular with this group. Second, students of Hispanic origin make up a considerably larger segment of MySpace users than their representation in the sample as a whole. Third, there is a relationship between parental education and use of some SNSs. In particular, students who have at least one parent with a graduate degree are more represented on Facebook, Xanga, and Friendster than they are in the aggregate sample, while students whose parents have less than a high school education are disproportionately users of MySpace.
This rather simple look at the data shows that social network site usage in the aggregate attracts a diverse set of students across services, but that certain groups are more represented on some sites than others. The important methodological take-away point here—in addition to the substantive ones about specific groups of users—is that when studying users of one SNS, researchers should exercise caution in generalizing the findings to users of another social network site.
Another way to look at the data is to consider the levels of SNS popularity by type of user attribute. Table 4 presents SNS usage statistics broken down by gender, race and ethnicity, and parental educational background. This breakdown is presented for SNS use in the aggregate and then separately for Facebook, MySpace, Xanga, and Friendster. (Due to Bebo’s and Orkut’s low rate of use in this sample, no disaggregated figures are presented for those two sites.)
Table 4. Percentage of different groups of people who use any SNS and specific social network sites+
| ||Any SNS||Facebook||MySpace||Xanga||Friendster|
|Race & ethnicity|
| White, Non-Hispanic||89||83**||57||3***||0***|
| African American, NH||84||80||58||0||0*|
| Asian American, NH||88||84**||39***||13***||10***|
| Native American, NH||83||75||58||8||0|
| Less than high school||88||64***||73***||1*||0*|
| High school||83*||73*||57||6||2|
| Some college||85||74*||57||6||2|
| Graduate degree||88||83||41***||9*||4|
Table 4 shows significant differences according to type of user. When it comes to aggregate SNS usage, women are more likely to use such services than are men, but once disaggregated by type of site, depending on the service, the differences all but disappear. That is, while female students in the sample are much more likely to use MySpace, there is little difference between young women and young men in the group when it comes to Facebook, Xanga, or Friendster use.
Regarding race and ethnicity, the most pronounced findings concern students of Hispanic and Asian origin. Hispanic students are significantly less likely to use Facebook (60% compared to 75% or more for other groups), whereas they are much more likely than others to use MySpace (73% among Hispanic students compared to 58% or less among all others). In contrast, like White students, Asian and Asian American students are much more likely to use Facebook than others, but they are significantly less likely to use MySpace. Additionally, this group of students is especially active on Xanga and Friendster compared to others.
There are also significant differences according to parents’ level of education. The most pronounced finding is that students whose parents have less than a high school degree are significantly less likely to be on Facebook and are significantly more likely to be MySpace users. In contrast, those who have at least one parent with a college education are significantly more likely to be Facebook users, while those who have at least one parent with a graduate degree are considerably less likely to spend time on MySpace. Xanga also seems to appeal more to those whose parents have higher levels of education. However, since there is a relationship between parental education and a student’s race and ethnicity, it is best to look at these associations using more advanced statistical techniques that allow other factors to be controlled while the relationship between the various background variables and SNS usage is examined. The next section does this by considering what predicts SNS use on the whole and with regard to specific sites when controlling for other factors in the model.
Explaining Any SNS Use
The findings presented in Table 5 suggest that numerous factors influence whether a student uses social network sites, while the results in Table 6 suggest that the predictors are not uniform across different services. The figures presented in both tables are “odds ratios,” meaning that any number greater than 1 constitutes a higher propensity to engage in SNS usage, whereas a number less than 1 suggests that the type of characteristic lowers the likelihood of social network site usage. First, I consider the findings for overall SNS usage, followed by an examination of specific site uses separately.
Table 5. Results of logistic regression analyses explaining SNS use (standard errors in parentheses)
| ||SNS use(any one of the four SNS)|
|Background only||Full model|
|Age||0.946 (0.102)||0.950 (0.112)|
|Gender (Male = 0, Female = 1)||1.563* (0.307)||1.660* (0.333)|
|Hispanic||0.803 (0.220)||0.919 (0.259)|
|African/African American||0.606 (0.212)||0.621 (0.226)|
|Asian/Asian American||0.966 (0.227)||1.007 (0.243)|
|Parents’ education: Less than high school||1.294 (0.553)||1.799 (0.815)|
|Parents’ education: High school||0.852 (0.242)||0.905 (0.262)|
|Parents’ education: College degree||1.496 (0.410)||1.392 (0.387)|
|Parents’ education: Graduate degree||1.194 (0.365)||1.057 (0.329)|
|Living with Parents|| ||0.640* (0.135)|
|Has Net access @ friends’/family’s|| ||2.022** (0.537)|
|Hours on Web/week (logged)|| ||1.431** (0.186)|
|Years online (logged)|| ||0.957 (0.262)|
The first column in Table 5 shows how core background characteristics are associated with any SNS use. The only variable in the core demographic model that is related to SNS use at a statistically significant level is gender. Women are more likely to use SNS than their male counterparts are. This finding is consistent with literature on women’s larger propensity to engage in person-to-person communication online as compared to men (e.g., Pew Internet and American Life Project, 2000). In this aggregate model with demographic background characteristics only, other factors, such as a student’s race and ethnicity and parental schooling level, do not show a statistically significant relationship with use of SNS.
Once more variables are added to the model, additional correlates of use emerge (see the second column on Table 5). Gender remains an important predictor of SNS use, while the other demographic variables continue to show no statistically significant relationship with aggregate SNS usage. However, in addition to gender, both context of use and experience with the medium are related to the adoption of such services. In particular, students who live at home with their parents are considerably less likely to use SNSs than those who live with roommates or on their own. Having Internet access at a friend’s or family member’s place shows the opposite relationship: Such resources increase the likelihood of use considerably. Regarding experience, how long someone has been online is not related to SNS usage; however, time spent on the Web is. This is not surprising, since SNS usage can take a lot of time. In fact, it may be SNS use, precisely, that results in these people spending more time online. Before offering an interpretation of some of the above findings, I turn to an examination of specific site usage.
Explaining Specific SNS Use
The columns in Table 6 present results of models predicting the use of Facebook, MySpace, Xanga, and Friendster, respectively. As a reminder, use of any one service is not contingent upon any other. That is, respondents could indicate use of any or all of these services. The first in each pair of models considers background characteristics only, while the second adds context and experience variables, just as in Table 5.
Although in most cases the age of students does not relate to whether they use a service, it does seem to matter when it comes to Facebook. Generally speaking, there is little variance on this characteristic, with the majority of respondents (97%) in their late teens. Nonetheless, the older participants are sufficiently different in their use of Facebook to yield a statistically significant result. Specifically, while four in five students 18 or 19 years of age use the service, only about three in five among those 20–29 do so.
The higher likelihood of women using these services holds in the case of all but one SNS: There does not seem to be any statistically significant difference between men and women when it comes to the use of Xanga.
In the earlier bivariate analyses of the data, race and ethnicity showed a relationship to people’s propensity to use some social network sites. More refined analyses allow a consideration of whether these findings are robust, when other factors are held constant. The results suggest that they are. Hispanic students are significantly less likely, in this sample, to use Facebook than are Whites (the omitted category). Conversely, they are considerably more likely to use MySpace. That is, even when controlling for students’ parental education, context of, and experience with use, there are significant differences in whether students of Hispanic origin use Facebook and MySpace compared to White students not of Hispanic origin.
The other group in the sample showing significantly different behavior compared to Whites is Asians/Asian Americans. They are significantly less likely than both White and Hispanic students to use MySpace. In contrast, they are considerably more likely to use Xanga and Friendster than White students, when other factors are controlled.
There is also a statistically significant relationship between the parental education of students and their choice of social network sites in the case of some services. Students whose parents have a college degree are significantly more likely to use Facebook than are those whose parents have some college education without a college degree (the base category in the models). Regarding predictors of MySpace use, the results suggest that students whose parents have less than a high school degree are considerably more likely to be users than those whose parents have some college education, while those who have at least one parent with a graduate degree are significantly less likely to be on MySpace compared to the omitted category. Xanga users show the opposite trend. Students whose parents have less than a high school education are much less likely to use Xanga than those with parents who have some college education. In sum, there seems to be a positive relationship between parental schooling and the use of Facebook and Xanga, and a negative relationship between parental education and the use of MySpace.
Turning to living context, the analyses suggest that students who live with their parents (just over half of this sample) are significantly less likely to use Facebook than are other students. This variable shows no statistically significant relationship with the use of any other service.
Consistent with literature on the importance of number of access points to the network (Hassani, 2006), the results suggest that for students in this sample, having access to the Internet at a friend’s or family member’s house increases the likelihood of using both Facebook and MySpace, when controlling for other factors. This finding, yet again, implies that existing inequalities are perpetuated in online activities.
Finally, as regards whether experience with the medium is related to specific SNS uses, the data suggest that Facebook, MySpace, and Xanga usage are all associated with more hours of Internet use weekly. This is not necessarily surprising, since the mere fact of using these services could lead students to spend more time online. In contrast, long-term experience with the medium—that is, how many years a student has been an Internet user—is not associated with adoption of most of these services. Only in the case of Friendster is there a significant relationship with veteran status—namely, students with fewer Internet user years behind them are considerably less likely to use that service. This finding is to be expected, since Friendster was more popular in the U.S. a few years ago, suggesting that people who have been online longer would be more likely to have adopted it in the first place.
To summarize the findings, it is clear that while an aggregate look at SNS use does not show much systematic relationship between a student’s demographic characteristics and SNS experiences, disaggregating the analyses by site tells a very different story. Students with varying backgrounds select into different services, potentially limiting the extent to which they will interact with a diverse set of users on those services. Additionally, social context of use and experiences with the medium have predictive power when it comes to explaining both specific and general levels of SNS adoption, suggesting that students who have more resources are spending more time on these sites and have more opportunities to benefit from them.