Eugene, Oregon
May 13-15, 1998
Teresa Sullivan, Vice-President and Graduate Dean
University of Texas at Austin
Good afternoon. Today I'll address myself to positive ways in which I think the new network will affect social science research by talking about three major ideas: First, the network as the subject of research; second, the network changing the process of research; and, third, the network framing the questions that the social sciences ask. In other words, topic, data, and theory, which is just about everything that social scientists do.
So let me talk, first of all, about the network itself as a subject for social science research. Listening to the conversation this morning it was pretty clear to me how sociology, topology, and psychology are likely to make use of the resources of the new Internet as a subject of study. I believe that geography may, as well, because geography is increasingly being reinterpreted not just as the study of spatial differences but also as the study of temporal differences. And the network's existence in time and space is something many geographers will be interested in. There is a small, growing group of experimental economists who will find the network an interesting topic to research. Of course, the issues of price, consumer behavior, and incentives will be abundantly available.
Let me talk about some specific areas in which I think different disciplines might be interested in studying the new network: Learning styles, virtual communities, inequality, social network, and pathology. Learning styles first.
I was most impressed in Program Session II this morning by the fact that distributed learning is being designed for master learners, not for the relatively weak learners often in need of remedial help whom we are encountering even at the most selective institutions. College pedagogy has for a long time tried to advocate a more interactive style, breaking students away from the rather passive interaction they develop with their television sets. And so many college-level faculty fear that more network-based and distributed learning approaches will simply reinforce a passive learning style and in the process reinforce a rather superficial content. I don't think that necessarily has to be the case, but I do believe that further research is needed on how students learn better and which students learn best in distributed learning programs. People, particularly psychologists, who study the process of learning will have much here at which to look.
Secondly, virtual communities. In my own field, sociology, there has been a great deal of emphasis on groups who are dependent on face-to-face interaction as the primary social obligation. If the new network becomes as pervasive as I suspect it may be within, say, 20 years, it will raise the issue of how real communities differ from virtual communities and whether the study of virtual communities can tell us something about the existence of real communities.
You may note that social scientists are really consumers of science fiction, and I think one of the reasons is that we can't experiment with whole societies, but by reading science fiction we can think about what other societies might look like. Virtual communities may finally offer us the way to look at communities that don't necessarily exist in time and space but that give us an opportunity to look at the underlying principles of community. For example, you could analyze a chat group conversation in terms of the roles taken by the different players in the group, by the content they engaged in, or by the way in which their written communication differs from spoken communication. Many of us have noticed, for example, that there is a kind of Internet etiquette. That etiquette is in some ways analogous to and in some ways different from our everyday etiquette. We've often noticed that there are mores developing among network users. The study of that in real-time is something that will be of great interest to a number of social scientists.
Inequality of all sorts is a pervasive topic in the social sciences, and the advent of the new network and heavy technological investment offers a great opportunity to study the haves and the have-nots in different ways, not only individually but also institutionally, regionally, and nationally.
I didn't hear much said this morning about the fact that the use of computers varies a great deal by age, gender, race, and social class, but those are most certainly the issues social scientists will be looking at. Further, the issue of differential access and how to make access more universal will be a durable one.
Social network analysis has become an important aspect of modern sociology. There are many techniques developed to study social networks, and the issue will naturally come up to the extent to which a social network is isomorphic with other networks, for example, the networks that spring up around the use of the Internet. If I were a network analyst, I would, for example, have charted the different consortia that came here this morning in terms of which partners signed on with which consortia and which partners were absent. Microsoft, I noticed, was conspicuous in its absence from this meeting.
Finally, social scientists love pathology. Internet addiction, network crime, etc.--I'm sure there will be some people who will be interested in this topic, as well. For those of you who are concerned with acquisitions, look forward to a number of new journals on electronic behavior and a paper that will analyze the interplay of the network and social science.
The second topic is that the new network will change the process of social science research, which is actually a more fundamental change and in some ways more interesting. The availability of the Internet has already begun to change data collection. In the 1980s the big revolution in data collection was the widespread use of CATI (Computer Assisted Telephone Interviewing) to eliminate coding and therefore make it possible to have much faster results from national sample surveys.
CATI, combined with random digit dialing of telephone numbers, was the easy way to produce rapid, nationally representative samples. Unfortunately, marketers adopted similar techniques. As a result, CATI and its associated sampling techniques have run into an increasing difficulty of having respondents cooperate. It's possible that using the Internet and similar methods can not only assure respondents of the confidentiality and reality of the research they are undergoing, but it can also have other benefits. A study released last week indicated, for example, that adolescents questioned about their sexual behavior apparently felt less stigmatized and were more likely to tell the truth about risky behaviors when they were committed to directly enter their answers instead of responding to an interviewer.
Of course, data archiving will also be much easier with electronic data. Electronic data archiving offers the possibility for researchers all over the world to analyze a data set simultaneously, leading to more robust findings and results during a short period of time, as opposed to the current eight to ten years that it takes to have scholarly discussion of a data set and its implications. In terms of data analysis, much more robust and sophisticated work will be possible. Researchers sitting at their own desk will be able to examine multiple data sets and do applications within a single authored work using many data sets over several places at several times and offering more of a finding. What this will mean for many of you who have libraries is that you will also be data archives or will at least provide the gateway to data archives so that the social scientists on campus will be able to do that. Internet2 will be desirable not only because of the increased pipeline but because of the greater available space for large and complicated data sets to be put together and kept together in one spot. This will be of particular interest to us with the year 2000 census, on the assumption that the census actually takes place. Although it's required by the Constitution, is not necessarily required to be done in a particularly good way. The 1990 census was the first to become available in widespread form to researchers in electronic format. It's clear that if the year 2000 census is similarly available, it will receive wide use.
A third way that data analysis will be changed is by multi-investigator collaboration. The National Institute of Health recently funded a project that involved six researchers in six cities doing a study on the impact of welfare reform. This kind of study is made far easier and the analogous nature of protocols is facilitated by the fact that these researchers can remain electronically linked with one other. If a finding is made on one side, that finding can be quickly tested on the other side by relatively minor adjustments in the study. Case studies, which are now a somewhat second-class citizen among all of the types of social science data studies, will be transformed by making it possible to immediately compare sets against each another by establishing protocol standards and then by archiving them, making it possible through electronic content analysis, for example, to draw common themes from them.
Finally, peer review will alter in this new networked arena. Peer review done rapidly rather than over six months, peer review done online rather than via snail mail, is likely to make the turnaround time for findings much more rapid. That may be one of the most important changes in terms of accelerating the pace of findings within the social sciences.
I also think that the new network will frame the questions social scientists ask in the future. In that respect, it will have an impact on theory. Our current concept of privacy and even our concepts of property will be seriously challenged by the availability of Internet2. We'll begin to ask questions that we didn't ask previously because we had only one historical point from which to ask it. For example, economists have studied post-war Germany's rapid rise in productivity and recovery in fields such as the steel industry. Germany was able to leap-frog developments in the technology of steel because all their steel plants had been destroyed. Germany rapidly became more competitive internationally in steel than the United States, where many steel mills had been built around the time of World War I, or at least between World War I and World War II. Thus, the technology permitted a reversal in the relationship of the predominance of the steel industry. In this same way, countries today which we may consider developing may be able to leap-frog more advanced countries by adopting electronic methods and materials without having to go through the intermediate stages that we are now traversing. These kinds of hypothesizing and theorizing will probably be stimulated by the Internet, even though they perhaps don't necessarily have to be carried out through Internet use.
Before I finish, let me add that there will also be many areas similarly stimulated by the availability of the Internet and particularly by the availability of social science data on the Internet. Having sent this data online means that every public-interest group can reapportion every legislative district for themselves with the application of simple software programs. So legislatures looking to redistrict themselves will be faced with not just one or two plans drawn up by the major political parties, but potentially thousands of plans, each one of which benefits some groups to the disadvantage of others. The level of controversy is likely to increase rather than diminish. And so for that final field of the social sciences that I have not yet mentioned, political science, there will be no dearth of opportunities.