Carnegie Mellon Study Reveals Negative Potential of Heavy Internet Use on Emotional Well Being

Contact: Teresa S. Thomas, 412-268-3580 or Anne Watzman 412-268-3830

PITTSBURGH--The Internet has the potential to make us socially isolated, lonely and depressed, according to the unexpected results of a study of home computer users by scientists at Carnegie Mellon University.

The findings are gathered from HomeNet, the first study to look specifically at the impact that the Internet is having over time on the social involvement and psychological well being of average Americans.

Published this month in The American Psychologist, a publication of the American Psychological Association, the findings provide a consistent picture of the downside of using the Internet extensively as a source of information or setting for friendship and or social support.

"We were surprised to find that what is a social technology has such anti-social consequences," says Robert Kraut, a professor of social psychology and human computer interaction at Carnegie Mellon who is the lead author of the article for The American Psychologist.

Even though people in the study heavily used electronic mail and other communication services on the Internet, the research found that spending time on the Internet was associated with later declines in talking among family members, reductions in the number of friends and acquaintances they kept up with, and increases in depression and loneliness.

Because the research studied the same people over time, it could rule out the possibility that people who are initially socially isolated, lonely and depressed were drawn to the Internet. Rather, according to Kraut, using the Internet seems to cause isolation, loneliness and depression.

"Our results have clear implications for further research on personal Internet use. As we understand the reasons for the declines in social involvement, there will be implications for social policies and for the design of Internet technology," he adds.

Various scientific and marketing reports say that more than 50 million Americans are using the Internet, a number that is rapidly growing. Given widespread use and with more growth expected, Kraut says the Internet could change the lives of Americans as much as the telephone did in the early 20th century or as television did in the 1950s and 1960s.

"We want to help make these changes good ones," he says.

HomeNet studied 169 personal computer users in Pittsburgh, whose communications on the Internet were monitored during their first years online. The home computer users are families with a wide range of demographic backgrounds whose common bond was a high school age student or membership in a community development group. The families used electronic mail, the World Wide Web and computer games, among other normal home computing uses. Time spent online varied a great deal among the subjects.

Members of the research team are part of Carnegie Mellon's Human-Computer Interaction Institute and include Kraut and Sara Kiesler, a professor of social and decision sciences; Tridas Mukophadhyay, a professor at Carnegie Mellon's graduate business school; William Scherlis, a senior research scientist and director of the Information Technology Center in the School of Computer Science; Vicki Lundmark, a post-doctoral fellow, and Michael Patterson, a graduate student in Social and Decision Sciences.

"We hope our findings help make things change on the Internet. We are not talking about Internet addicts, just regular people," Kraut says. "These are not just results that occur in the extremes. And these are the same people who, when asked, describe the Internet as a positive thing."

The technology that has allowed people to keep in touch with distant family members and friends, to find information quickly and to develop friendships with people around the world apparently is also replacing vital, everyday human communication.

"Many users may be substituting weak online friendships for their stronger, real-life relationships," Kiesler says. "You don't have to deal with unpleasantness, because if you don't like somebody's behavior, you can just log off. In real life, relationships aren't always easy. Yet dealing with some of those hard parts is good for us. It helps us keep connected with people."

Greater use of the Internet was associated with statistically significant declines in the social involvement that Kiesler refers to. Decreases in social involvement were indicated by a drop-off in communication within a participant's families, the size of a person's social networks and reports by participants of increases in loneliness and depression, psychological states associated with reduced social involvement.

In all, the study uses data on 169 people in 73 families. A little over half the subjects are female users, a quarter of them belong to minorities. The subject pool also represents a fairly wide income range.

Of the different demographic groups, teenagers seem the most vulnerable to potential negative effects. What's more, teenagers used the Internet for more hours than did adults.

Mukhopadhyay offers the following advice to parents: "The basic objective is to maintain open communication and to stay vigilant. As far as the computer and Internet go, you can put the machine in a public place - in the living room or kitchen rather than the basement or the kid's room. This will automatically ensure that your teen does not use the Internet too much."

Carnegie Mellon's scientists believe the findings will spark a debate, not only for Internet users and researchers, but also for government agencies looking at growth of the Internet and for companies that write Internet software.

Scherlis notes, "We are not branding the Internet as either socially good or bad. The Internet is a complex and multi-faceted social phenomenon and it is evolving rapidly. It was created more than 20 years ago for sharing technical information among scientists. It's really only recently that the Internet has become a public resource, and the average citizen who uses the 'Net has largely inherited this set of services. Our results show that there may be real benefits from greater research and development to the broad area of user level communication and information services. Both industry and government can foster this growth through research into new services, experimentation, evaluation and standards development."

The research was sponsored by the National Science Foundation, the Markle Foundation, and a consortium of computer companies (Apple Computer, Hewlett Packard, Intel, Panasonic), software companies (Lotus Development Corporation, Interval Research), and communications companies (AT&T Research, US Postal Service, Bell Atlantic, Bellcore, US West Advanced Technologies, NTT, CNET) and others (NPD).

Contact: Teresa S. Thomas, 412-268-3580 or Anne Watzman 412-268-3830

Background: Q&A Explores Social Implications of Internet Use for Families

PITTSBURGH--There are many issues related to use of the Internet - from loneliness and depression caused by too much online time to what can be done to make the Internet a more social place. These questions and answers deal with these issues in greater depth. The comments are based on recent findings gathered from HomeNet, the first study to look specifically at the impact that the Internet has on general emotional well being, particularly the emotional well being of families. Those findings were generally negative.

Respondents include Carnegie Mellon's Robert Kraut, a professor of social psychology and human-computer interaction; Sara Kiesler, a professor of social and decision sciences; Tridas Mukophadhyay, a professor at Carnegie Mellon's graduate business school; and William Scherlis, a senior research scientist and director of the Information Technology Center in the School of Computer Science. Vicki Lundmark, a post-doctoral candidate, and Michael Patterson, a graduate student in Social and Decision Sciences, contributed to the research findings.

QUESTION. Why has something perceived as being so social had such an anti-social impact on its users?

Kiesler: For many people, the Internet is wonderfully convenient and fun. No need to dress up to "meet" new people, join a group discussion, or get in touch with your brother in Alaska. So talking on the Net is fun and rewarding. But talking on the Net takes time and attention away from your "real life." Think of the teenager avidly typing away in his room instead of playing a pick-up game of softball with his friends, and avoiding that party where he doesn't know anyone. Friends in real life are sometimes more troublesome, but they are the people who play a bigger role in one's life-- the people who know you as a person and who are there to give the most all-round social support. Of course, if there *is* nobody around for you, then the Internet might be a lifesaver. But even so, maybe it would be better if the Internet helped people find local friends rather than far-flung people who may disappear from their lives as easily as they entered them This is exactly the paradox we are trying to understand, and we have to admit we don't yet have all the answers.

Kraut: We're starting to think that the problem isn't with what people get when they go on- line, but with what they give up in their real lives to achieve it. It is likely that the social contact people get on the Internet is of lower quality than the social contact they get when they talk to members of their family, go to church groups or clubs, or have dinner parties. In some ways, it's easier to get on-line social contact than to get the real thing. You don't have to be at the same place at the same time, and can communicate when it's convenient for you. With the strangers you meet in chat rooms, you can always drop the relationships they don't work out, without having to run into them again and again. The problem may be, though, that the easy is driving out the good.

Mukophadhyay: It's true that the Internet allows us to get to know a lot of people we would not meet otherwise. People can use email, go to chat rooms, subscribe to distribution lists or newsgroups. You have many ways to meet many people. Of course, you can also surf the Web as much as you want. However, the key point is that cyber-friendship may be good if you have no opportunity to meet people. But it's not a good substitute for real-life friendship. In fact, if you like someone you meet on the Internet, chances are that you would want to meet that person face to face.

QUESTION: Did the findings surprise the research team?

Kiesler: Yes, the findings surprised me for two reasons. First, many studies of work communication, including our own, were showing positive effects of electronic communication on such things as knowledge of the organization or participation in organizational life. Second, many anecdotal studies that preceded ours suggested there would be only positive effects. Third, we expected measures such as depression and social involvement to be very stable, and not to be affected by a single technology.

Kraut: When we started this research, we weren't sure whether the Internet was going to be used more like the telephone or the TV, the other important information technologies heavily used at home. Our own research had shown that for the people we were studying, the phone-like uses were more important. They used the Internet to keep in touch with friends and family and to make new friends on-line. Teenagers rushed home from school to exchange mail with kids they saw an hour earlier. Both teens and adults also started to have chitchat and to exchange information with people they met on distribution lists or in chat rooms. These social uses were more important to them than finding impersonal entertainment and information on the World Wide Web. So we were shocked when we discovered that as people used the Internet more, they became more socially isolated and lonelier. And using the Internet primarily for communication purposes didn't seem to prevent these negative effects.

Mukophadhyay: I was not surprised that the Internet had some negative effects. It's only natural that a powerful new technology like the Internet will have some unexpected consequences. What is surprising is that it seems uniformly and negatively to affect a bunch of measures of social involvement and emotional well being.

QUESTION: How can you be showing that using the Internet leads to less social involvement when common experience and your own data show that it is used heavily for keeping up with old friends and making new ones?

Kraut: To draw conclusions about the consequences of using the Internet, one needs to have comparisons. One needs to compare people who are using the Internet heavily with people who are not using it or using it very little. People can only report on what they have experienced, and many of us have experienced using the Internet for social purposes. But people can't easily report on what they are giving up, what economists call opportunity costs. For example, I find the Internet very convenient for keeping up with colleagues from my old job. The question is, does the time and energy I devote to these email messages hinder me from forming strong friendships with people in my current work setting or community? The best way to see what is being given up, is by comparing heavy and light users.

QUESTION: Why should readers take this research more seriously than many other studies on consequences of using computers and the Internet?

Kraut: First, this is one of the few pieces of research that has looked at all at the social and psychological consequences of using the Internet. Second, unlike some national surveys, this research measures actual usage. Our data tell us that there is a lot of error in people's reports about how much they use the Internet. Third and most important, we measure Internet use, social involvement and psychological well being at multiple time points. As a result, we can rule out the possibility that being socially isolated, lonely and depressed causes people to use the Internet more, and can discover what the consequences are of using the Internet once initial social and psychological state are taken into account.

QUESTION: Society survived the introduction of the telephone and television. Do you think we'll survive the Internet?

Kiesler: We humans are very confident, very smart and very adaptive. So of course we will survive the Internet. But we can make changes in technology, too. We made changes in the telephone (automated switching, privacy protections, universal service) and in television (Mr. Rogers and Sesame Street, and rating systems). We expect changes in the Internet. We hope the basis for these changes will be research rather than hyperbole or scare tactics based on anecdotes.

Mukophadhyay: Society changes. But the Internet is changing faster and will continue to do so for some time. It'll be a while before we come to terms with the Internet.

QUESTION: There are lots of other technologies that distract people from their social relationships. Is the Internet worse than them?

Kraut: When people watch TV, play video games, or even read, they are withdrawing from social interaction. We can't directly compare whether spending an hour on the Internet is better or worse than spending it watching TV. Other researchers have shown effects of watching TV that are similar to the ones we see for using the Internet. My personal opinion is similar to those expressed by some of the parents in our study. They think that much of what their kids do on-line is a waste of time, but it's a better waste of time than watching TV, because its more personally engaging, active, social and literate.

Mukophadhyay: There are two issues here. First, it's too early to have a definitive opinion about the Internet. Second, there is a fundamental difference between the Internet and TV or any other household technologies. The Internet is so flexible and has so many possibilities. We would probably see a wide range of impact from the Internet.

QUESTION: If the Internet is having these negative effects, why are people using it so much?

Kraut: It's important to remember that we have only looked at a small set of the possible effects that the Internet is having. We haven't investigated whether using the Internet provides job-relevant skills for people, gives them useful information for work or school, allows them to organize their home lives or purchases more efficiently, changes their self-esteem, or had any number of compensating benefits. It is possible that, all things considered, the Internet is actually good for people.

On the other hand, we have many other examples of people engaging in behavior that is bad for them, either because the immediate experience is pleasurable even though the long-term consequences are bad or because the cost to gain some pleasure is so low. TV is the classic example. Adults watch over two hours of TV per day and spend less than 60 minutes visiting with friends and conversing with them, even though they much prefer socializing to watching TV. The reason sees to be that watching TV is easier than socializing, even though adults don't like watching TV that much.

Mukophadhyay: People believe that the Internet is here to stay. It is the way to go for the future. They do not want to miss the boat. Let us also not forget that the Internet is a valuable tool for getting information and even for communication.

QUESTION: We're just not talking about Internet addicts are we? Should we ALL be concerned about the findings?

Kiesler: We didn't see a sudden drop in psychological well being or social involvement when people reached a certain number of hours of usage. Our results show a "linear" effect, which means the more Internet, the more the negative results. It is important to emphasize that this research report doesn't apply to educational and learning effects of the Internet. Our data suggest that there may be a tradeoff here, with more usage leading to better computer skills or worldly knowledge, and even self-esteem enhancements, even while social involvement declines somewhat.

Mukophadhyay: We're talking about average people who use the Internet heavily and experience negative consequences. The lesson is that one should not run away from friends and families and escape into cyberspace.

QUESTION: How can we recognize susceptibility in ourselves -- in other words, when can we, as individuals, recognize that we may be entering a danger zone of Internet usage?

Kiesler: Many people do things "too much." Eating quarts of ice cream at night, smoking three packs a day and sitting at the computer 10 hours at a time. Many people know very well when they are in a danger zone, and they must decide to change their behavior themselves. Most people probably do. We are planning to analyze our sample over a somewhat longer time period to see who drops and who doesn't and how their own behavioral changes affect their well being.

QUESTION: If teens are afflicted more by loneliness and depression brought on by Internet use, do you have any useful advice for parents, schools or college counselors?

Kiesler: Send them to camp, encourage them to play sports, help them be with real life friends.

Kraut: My teenaged kids have access to the Internet, which I pay for, so obviously I think that the Internet can be valuable for them. But we've put limits on how much they can use the computer. I've also talked to my kids about this research and its implications, and tried to encourage them to monitor how they use it. I try to encourage them to get involved in real social activity -- after school clubs, temple youth group, visits to friends, volunteer work. As parents, we refused to allow them to put a computer in their bedroom, so we could monitor how much they use it.

Scherlis: It would be an overreaction to stop Internet use altogether, since the educational and learning benefits would also stop. Our point is that use of the Internet has many kinds of social effects (just like the telephone and TV). We should consider all these effects in making judgments about its use for ourselves and as parents. We should also recognize that this technology is evolving rapidly, and its social impacts will also change. Finally, there are many different ways the Internet can be used (again, like the telephone and TV). Parents and teachers should focus their attention on the nature of the Internet use as well as the amount of the use.

Mukophadhyay: The advice is simple. Maintain open communication and stay vigilant. As far as the computer and Internet go, the best thing to do is to put the machine in a public place - in the living room or kitchen rather than the basement or the kid's room. This will automatically ensure that your teen does not use the Internet too much.

QUESTION: The Internet was first created for sharing scientific information and not for social activity. Would it be better if it went back to being that?

Kiesler: The Internet (it wasn't called that) was first created so engineers and computer scientists could share computers. Even in those early days, it was used extensively for communication. You aren't going to change that, and you wouldn't want to. Society depends on communication, and grows stronger with more communication. Perhaps it would be better if people talked more with local family and friends rather than chatting away on the Internet with strangers. But we don't know for sure what led to our effects, so this is just speculation on our part.

Kraut: Absolutely not. I have confidence that we can shape service on the Internet so that it will have beneficial influences. I've seen many cases where people are able to reconnect to a long-lost friend, share problems and get advice, pleasurably pass the time and provide social support on-line. Our goal should be to make this style of interaction increase and to make the likelihood of wasted interaction decrease.

QUESTION: If not, what needs to happen to make email a socially useful tool for communication -- one that brings people with strong supportive ties together in positive ways?

Kiesler: One thing preventing people from using the Internet to support local strong ties is that everyone close to them that they know doesn't have email or is wary of the technical complications. Email has to be more accessible to everyone locally and it should be as easy for grandpa as for his teenage grandson.

Kraut: I think there are ways to use the Internet that will improve how it influences lives. For example, in our study, some of the kids who graduated from a local high school created distribution lists to keep in touch with their classmates after they moved on the college. They are able to share news, and to arrange activities when they came back to town for vacations. This seems to me to be a healthier use than one that encourages discussions with strangers. Even better would be services that support communication among already existing social groups. For example, schools could have after- school clubs and homework sessions on-line. If kids find electronic communication fascinating, then channel it so that they communicate with people with whom they already have ties.

Mukophadhyay: Unfortunately, most people still do not have access to the Internet. We found how families were excited when a close friend or a distant relation went online. As the Internet becomes more widespread and it becomes easier to use, it will take its place beside the telephone, cellular telephone and beeper as a powerful medium for interpersonal communication.

QUESTION: As consumers, what should we tell government regulators, Internet providers and software companies about what we want?

Kiesler: The Internet is too often fodder for political debates and grandstanding. Policy-making should be based on careful empirical research. We don't hesitate to do evaluations of educational interventions and medical technologies. We should do much more empirical evaluation of the massive social intervention called the Internet.

Scherlis: Issues related to content have triggered much of the policy debate on the Internet. But, like the telephone, the Internet itself, along with the user-level services it supports (such as email and the Web), is neutral with respect to the content it conveys. So we should avoid simplistic appraisals and focus on understanding the issues through careful research.

QUESTION: What are the implications of this study on policy and technology development related to the Internet?

Scherlis: For more than 20 years, the Internet served primarily to support the sharing of technical information among professionals working in offices. Most of the Internet services we use today -- email, the Web and newsgroups, for example, were designed to support this technical exchange. Only recently has the Internet become a public resource, and the average citizen who uses the 'net has largely inherited this set of services. These services constitute the user experience of the Internet.

There has been a lot of attention paid to the huge improvements in the capacity and connectivity of the network. But (except for multimedia support) the actual user-level services they support have not evolved very much since their introduction. (Packet-switched email was introduced in 1972, and the email systems we use today are similar to those of a quarter-century ago. The Web was invented in 1989 to help physicists share data and scientific results, and has evolved primarily in technical respects.) Our results suggest that this set of services is not well adapted, for example, to increasing the social involvement of network users.

Historically, government research sponsorship and policy initiatives have been the major force behind the development of the Internet. Without the sustained government involvement over the past two decades, there would be no Internet today. But this attention has generally focused on network infrastructure.

Our results suggest that there may be real benefits from greater R&D attention to the broad area of user-level communication and information services. Both industry and government can help foster this growth, through research into new services, experimentation, evaluation and standards development.

QUESTION: Does this report end the debate about the social value of the Internet?

Kiesler The debate will continue as more data come in.

Kraut: No, for several reasons. First, we don't know how these results generalize and how stable these results are. Will the results be the same at different eras, with more socially isolated people, using different services? We doubt it. And second, we don't really understand the mechanisms that produce the negative effects, even in the sample we've studied. One must understand both the limits of the results and the mechanisms behind them before designing solutions or making policy recommendations.

Scherlis: Beware of monolithic characterizations of the Internet as socially "good" or "bad." The Internet is a complex and multi-faceted social phenomenon, and it is evolving rapidly. It is as diverse both in form and in content as the existing and established popular media. Any debate and subsequent policy setting must recognize this complexity.

Mukophadhyay: This report really jump-starts the debate. We will next examine how stable or generalizable these results are. We'll also try to understand the mechanisms that cause these results.

QUESTION: As your study progresses, what will you look at next?

Kiesler: Do negative effects continue or do they dissipate as people get less enamored of the Internet? Are the effects we found the same for different kinds of people and different samples? Are the effects better or worse than the effects of television? Is there a tradeoff of positive educational effects or skill for negative social effects?