This study aims to understand the social and emotional climate that teens experience in spaces where they can interact with others online. There has been considerable concern among parents, teachers, policy makers, and advocates about the nature and intensity of online social encounters among teens. In this research, we pay particular attention to teens’ experiences on social network sites, including Twitter.10 Do teens find these relatively new online social spaces friendly or hostile or somewhere in between?

As in other aspects of their lives, teens witness a range of behaviors online. A Facebook profile can be the site of a budding romance or the staging ground for conflict. Exchanges that begin online can move offline and face-to-face conversations that are initiated in person can continue in social media spaces where they are then annotated with comments, photos, and videos. When a conflict arises, some choose to air their grievances in full view of their friends, while others feel that private communication channels are a more appropriate place to deal with relationship issues.

But these social media-based interactions are not face-to-face. They are mediated, in this case by a computer or cell phone or a tablet computer. In the past, mediated interactions might have taken place via paper letter or set of wires and a phone between the conversing partners. Now, all internet users have access to a broader digital audience. And in this new environment, social norms of behavior and etiquette are still being formed.

Norms are “rules, about which there is some degree of consensus, and which are enforced through social sanctions.”11 These rules do not have to be written down or made explicit, but they are understood well enough by the group so that they shape people’s behavior. Norms also vary from group to group and can cause problems in these new social media realms where users frequently intersect with a diverse array of groups whose norms may be different.

This navigational complexity makes these spaces especially interesting and challenging for teens because some of the “work” of being an adolescent is to incorporate community norms into their lives at the same time the teen is trying to craft a personal identity.

We focused our inquiry on social media sites because they are spaces of online interaction that a large number of teens use. We wanted to understand the types of experiences teens were having online and what kinds of other online – and offline – experiences resulted from those interactions. We examined what teens observe online and how they respond to those experiences. We asked where teens get advice about how to be safe online, and who they seek out for help when they have a specific situation to discuss. We asked them who served as the greatest influence in their thinking about what behavior was appropriate and inappropriate online and on their mobile phones.

We also probed the environment around teens’ online experiences by examining their privacy controls and practices, as well as the level of regulation of their online environment by their parents. We further probed more serious experiences that teens have in their lives, including bullying in a variety of locales and the exchange of sexually charged digital images.

Understanding bullying and what we call “social media meanness” has gotten more complicated even since we began this project just under a year ago. Researchers Alice Marwick and danah boyd have recently released research that complicates the term “bullying”12 and suggests that some of the troubling interactions that adults label as bullying may be referred to as “drama” by teens. According to Marwick and boyd, the term “drama” is used by teens to assert a greater sense of agency – or control – in their social lives. The authors argue that the word “drama” sidesteps being positioned as a “victim” or “bully” and allows teens to see themselves as active participants in the things that happen to them.

This survey’s construction of “mean and cruel” online behavior attempts to get at some of these behavioral distinctions, but it may not cover the entire landscape of teens’ social experiences online fully, even in conjunction with our bullying questions.

In this study, we are using a variety of different terms to talk about these issues – “mean and cruel behavior,” “social media meanness,”“harassment,” and “bullying.” While “mean and cruel” behavior and harassment both appear in the same question and are used interchangeably, bullying is asked as a distinctly separate question. We at Pew Internet have treated bullying and meanness as different experiences, but we acknowledge that many of our respondents may not make the same distinctions.13 In this survey, each of these offered terms is left undefined, and so each teen (or adult) respondent determined her or his own meaning when answering the question.

The focus on bullying and cyberbullying has become intense in the policy community over the past two years, much in the same way concern about online predators captured public attention in prior years. This report aims to provide impartial data and context to these important and evolving conversations about online safety by giving voice to teens’ own experiences, however they may be defined.

How the study was conducted

To answer these questions, we designed a multi-modal study. We began in December 2010 with a meeting of experts (for list of experts, please see Acknowledgements) to advise us on how to think about the concept of digital citizenship and to help us refine the focus of the project. In January and February 2011, we conducted 7 focus groups with middle and high school students in the greater Washington, DC metropolitan area. Each focus group had between 7 and 14 people, for a total of 57 participants across all groups. The groups were co-ed and the ages of participants ranged from 12 to 19. The participants crossed the socio-economic spectrum. Black youth were overrepresented. After finishing the groups and synthesizing the findings from both the experts meeting and the focus groups, we wrote and fielded a telephone survey. The survey was fielded April 19 through July 14, 2011 and was administered by landline and cell phone, in English and in Spanish, to 799 teens ages 12-17 and a parent or guardian. Black and Latino families were oversampled. The margin of error for the full sample is ±5 percentage points. The margin of error for the 623 teen social network site users is ±6 percentage points. In 11% of the interviews (90 out of 799), the interviewer noted that a parent listened to the child’s interview. Very few differences were observed in the answers given by teens when parents were listening compared with interviews alone with teens. Any statistically significant differences between the responses of teens in parent-attended interviews and teen solo interviews are noted in footnotes to the text. For more details on this, or any other methodology question, please see the Methodology section at the end of this report.