October 14, 2007

Teens and Online Stranger Contact

Teens who create social networking profiles or post photos online are more likely to be contacted online by people they do not know; Girls are more likely than boys to have unwanted encounters online.

While the number of teens who have been made uncomfortable by an online experience with someone they do not know is relatively small,1 there are certain traits and activities that are more likely to attract interactions with unknown individuals, whether unwanted or otherwise. In particular, teens who have created profiles on social networking sites (SNS) and those who have posted photos of themselves online are more likely than others to be contacted online by strangers. Girls are more likely than boys to report online contact that made them scared or uncomfortable.

Stranger contact among online teens

These findings are based on a survey of 935 teens age 12-17 by the Pew Internet & American Life Project taken from October 23 to November 19, 2006.

Other key findings include:

  • Those who have posted photos of themselves and created profiles on social networking sites are more likely to have been contacted online by people they do not know. However, these teens are no more likely to find these contacts scary or uncomfortable than other online teens who are contacted by strangers.
  • Internet monitoring software that allows parental supervision seems to be more effective than online filtering software in limiting contact with strangers online.

Further analysis

The analysis that follows covers two subjects. The first section examines the factors correlated with online contact with strangers that is scary or uncomfortable. The next section covers the factors correlated with all kinds of contact with unknown individuals, whether it is troublesome or not.

Among teens who have been contacted online by someone they do not know, gender is the primary predictor of contact that is scary or uncomfortable.

Due to sample size limitations and the fact that the majority of teens’ online contacts are benign, it is difficult to disentangle all the factors that are associated with an increased prevalence of scary or uncomfortable online encounters. However, among the factors evaluated in this study, gender consistently has a strong correlation with contact that is scary or uncomfortable. Among teens who have been contacted by someone they do not know, girls (27%) are significantly more likely to report feeling scared or uncomfortable as a result of the contact compared with boys (15%).

Despite popular concerns about teens and social networking, our analysis suggests that social networking sites are not inherently more inviting to scary or uncomfortable contacts than other online activities. Among teens who have been contacted by a stranger online, 21% of profile-owning teens say they felt scared or uncomfortable as a result of this contact, compared with 28% of non-profile owners. This result is not necessarily surprising since nearly half (49%) of social networking teens use these sites to make new friends—in other words, connect with people they do not currently know. It may also be the case that profile-owning teens see some level of unwanted contact as a known downside of maintaining a social networking profile and view it as a relatively minor “cost of doing business” in this environment.

Several factors are associated with an increased likelihood of online contact, whether it is uncomfortable contact or not.2

On the general issue of online contact, statistical modeling shows that posting photos online and creating a profile on a social network site3 are the activities most strongly associated with stranger contact, whether it is scary or not.

Although the creation of profiles on social networking sites is strongly predictive of stranger contact among online teens, the specifics of a person’s social networking profile have little influence on the likelihood of being contacted by an unknown person. For instance, there is no consistent association between stranger contact and the types of information that is posted in an SNS profile such as a person’s first or last name, his school name, or his email address. There is also no statistically significant association between stranger contact and having a public SNS profile – that is, a profile that is visible to anyone. Once factors such as age, gender or posting photos online are controlled for statistically, there is little difference between public profile creators and those whose profile information is available only to those designated by the profile creator.

Teens who use social networking sites to flirt are more likely to be contacted by people they do not know once other factors are controlled for, although a similar effect is not seen in teens who use social networking sites to make new friends. The magnitude of this “flirting effect” is roughly comparable to the impact of gender. Girls are significantly more likely than boys to be contacted by someone they do not know when other factors are held constant. However, the association between contact by strangers and gender is not as strong as the association with profile ownership or posting photos online.

Interestingly, while the presence of internet monitoring software on the computer a teen uses at home is associated with a somewhat lower likelihood of stranger contact once other factors are controlled for, there is no similar effect for internet filters that block certain websites. Some explanations for this discrepancy may relate to the different features offered with monitoring versus filtering software, or the fact that parents who use monitoring software may be relatively more likely to take an active role in observing their child’s internet usage habits than parents who rely on filtering software.

Analytical Approach:  Regression analysis details

This report makes use of bivariate analyses which measure stranger contact according to individual demographics, technologies and behaviors. Additionally, we conducted several multivariate regression analyses to test whether the associations between high levels of stranger contact and different demographic and behavioral characteristics persist once other factors are controlled for. This more in-depth analysis allows us to pinpoint the factors that are most relevant to the issue of stranger contact among teens.

Logistic regression was used in conducting the analysis, with the dependent variable being contact by online strangers. Stranger contact was modeled as a function of the following variables:

  • SNS profile ownership
  • SNS profile is visible to anyone, not just my friends
  • Posting photos online where others can see them
  • Home computer has internet filtering or monitoring software
  • Use social networking sites to make new friends or flirt
  • Contents of SNS profile (e.g., first name, last name, email address, city or town where you live)
  • Demographic traits (e.g., gender and age)
  1. For the analysis reported here, multivariate regression analysis was used to further evaluate these and other key relationships between teenage behaviors and online stranger contact. For more details, please see the analytic approach section of this report.
  2. Among profile-owning teens in this study, the overwhelming majority (85%) have profiles on MySpace. While differences may exist between MySpace and other social networking sites, our sample of non-MySpace users is too small to analyze stranger contact as it relates to different social networking platforms.