Debating the societal effects of the internet

Our Shrinking Social UniverseReply to Etzioni By Norman H. Nie and Lutz Erbring

ometimes ideological or meth- ence groups” or Paul Lazarsfeld’s “in- findings suggesting that the internet may odological prejudice seems to troduction... of the concept of two- reduce the time people spend interacting S get in the way of sound judg- step communication and opinion lead- with friends and family or increase the ment and even common sense. Per- ers.” Actually, though, he seems to time people spend working. haps that is what happened to our object less to our interpretations of the colleague Amitai Etzioni who—in the data than to the fact that we “provided o what is the message our dis- sunset of a distinguished career and a summary of [our] study to the media, tinguished colleague seems un- presumably against his own better judg- which got front page attention.” In S able or unwilling to understand? ment—may have allowed himself to get any event, we feel perfectly at ease with It is, above all, our finding that the carried away by adrenaline in his com- our findings and are at a loss how to more people use the internet, the less mentary on preliminary findings from time they report spending with “real our study on social consequences of the human beings.” Now, while there internet as reported by the media. may be reasons to dislike that finding because of its implications for the fu- Actually, as behooves a scholar, he did ture quality of social life (indeed we are not stop with second-hand media re- only witnessing the beginnings of the ports but did at least take the trouble of internet’s impact), one cannot make it examining some of the more detailed go away by (a) wishing it away, (b) evidence available on our web site— arguing it away, (c) refusing to ac- though, apparently, to no avail. knowledge the facts, or (d) attacking those who report it. Our colleague seems to have been so eager to rush to judgment that he mis- It is, of course, ironic that literally read our findings as a message of within days of his venomous attack, technophobic doomsday prophets or ©2000 www.arttoday.com another national study of computer raving cultural critics. Yet even if that and internet use by National Public were true, which it is not—we are en- take advantage of the “sympathizing” Radio, the Kaiser Family Foundation, thusiastic internet users convinced of offered by Professor Etzioni (except and Harvard University reported es- the benefits that this new technology perhaps to savor it until some future sentially the same (even stronger) re- holds for society—it would hardly be day when our results happen to coin- sults: 58% of their respondents report considered a sign of scholarly conduct or cide with his predilections). computers have led them to spend less analytic competence to beat the messen- time with their families and friends, ger when one doesn’t like the message. Apparently, Professor Etzioni divides and 46% say computers have given the world of research findings (and them less free time. Indeed there are Of course, we are delighted to be put in sources) into two simple categories, reasons why textbooks on scientific the company of “survey masters” whom based on his personal values and preju- methodology emphasize the impor- he seems to chastise (or grudgingly ad- dices: (1) findings he likes (which he tance of replication! Thus, we could mire) for “coming up with post hoc dismisses as “self-evident and dull” rest our substantive case here. interpretations of their data,” such as when they come from a source he Robert Merton’s “‘discovery’ of refer- dislikes); and (2) findings he dislikes However, for the benefit of readers (which he tries to either wish away, who may not be familiar with the de- argue away, deny outright, or if all else tails of our study as caricatured by Norman H. Nie is director of the Stanford fails, discredit when they come from a Professor Etzioni, a few additional re- Institute for the Quantitative Study of source he dislikes). In the former cat- marks may be in order. Society and research professor of political egory are our findings suggesting that science, Stanford University. Lutz Erbring the internet may reduce the time people Specifically, textbooks on methodol- is professor of mass communication stud- spend shopping in stores or commuting ogy also introduce students to the fun- ies, Free University of Berlin. in traffic; in the latter category are our damental idea of relationships between

44 Public Perspective, May/June 2000 variables. That idea would have been Why else would he again restrict his they spend more time with friends and helpful to our distinguished colleague argument to “groups” of those who family—while missing the essential in understanding our findings, and spend 5 to 10 hours (10% reporting point of asking whether that number, would have saved him the misguided less time socializing), or those who as in the case of spending less time, is effort of attacking our findings by ar- spend 10 or more hours (15% report- consistently related to hours of internet guing about the size or characteristics ing less socializing), while ignoring use (it is NOT!). And he goes on to of “groups.” While we did, indeed, those who spend less than one hour make an elaborate argument as to why simplify our findings for journalistic (with 4% reporting less socializing) or the internet actually leaves people with presentation and public consumption 1 to 5 hours (with 8% reporting less more time for social relationships— in our press release by focusing on the socializing), thus evidently failing to when in fact they tell us they spend “group” of regular internet users (those recognize the systematic nature of the less. So who should we assume has got relationship it right: the great master theorist, or on which our the people speaking for themselves? conclusions For our part, we have no doubt whom “Our colleague seems to have are predi- to believe. been so eager to rush to judg- cated? Or why would he However, we have no intention of ex- ment that he misread our compare these tending our response into a seminar on figures with basic social science methodology. We findings as a message from univariate are simply amazed at the remarkable marginal per- arrogance of this line of reasoning: he technophobic doomsday centages knows better than our respondents. prophets...” which are Perhaps someone like Professor Etzioni completely ir- has no need to bother with the tedium, relevant here or the logic, of research methodology. spending 5 hours a week or more on and can throw no light whatsoever on He stands above it by virtue of access to the ’net), we also presented our survey how amount of time on the internet a superior source of insight and knowl- results for a more professional or so- affects behavioral outcomes? edge—and he does not hesitate to let phisticated audience in terms of rela- us in on what that superior source tionships between variables: between And, indeed, why would he call for a might be: “Everyday experience.” It is hours of use (ranging from less than “control group”—a concept that makes “[e]veryday experience” which one to more than 10 hours per week) no sense in the language of relation- “...shows that people use the internet... and percent reporting decreases (or ships between variables (to say nothing to reinforce existing relations among increases) in time spent with family about his quaint suggestion of asking a family, friends, and coworkers; to forge and friends (or working at home and at “control group” of non-internet users new relationships...; and to join or the office, respectively). about how the use of the internet has form communities.” The master affected their lives)? Still, after com- knows; our respondents have no idea We were assuming, of course, that plaining about our failure to include what they are talking about. someone of the caliber of Professor the non-users as a control group, he Etzioni would know how to read our goes on to complain that we did in- Meanwhile, as the master morphs into graphs and tables correctly, and would clude them (we did not!) even though a virtual communitarian in cyberspace, realize that when a relationship is con- they would be contaminated by our the rest of us toil on in the lowly sistent across the entire range of hours internet-based data collection meth- quarries of empirical research. Sur- of internet use, the results do not de- odology. Professor Etzioni seems ei- prisingly, we love our work. pend on the percentages reported for ther more confused than we thought the “group” of regular users (5 hours or possible, or simply prepared to dis- more). He does, in fact, correctly pense with logic just so he can maul us report some of our more detailed re- coming and going, or both. sults (by hours of internet use), but apparently fails to understand their His personal convictions seem so meaning in terms of a consistent, sys- strong, and his methodological insights tematic relationship between hours of so weak, that he even tries to hang on internet use and behavioral outcomes. to the handful of respondents, who say

Public Perspective, May/June 2000 45