Return to articles list

"All interactions magazine articles are posted with the permission of ACM."

© 1997, Association for Computing Machinery. Permission to copy without fee, all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage, the ACM copyright notice and title of the publication and its date appear and notice is given that copying is by permission of the Association for Computing Machinery. To copy otherwise, or republish, requires a fee and/or specific permission.

Reprinted with permission from interactions magazine, 4[2], 13-18, March, 1997.

Previous columns and articles in this magazine have discussed ethnographic methods and have given examples of where such field methods can fit in the development life cycle. This article builds on these past articles by highlighting some of the early findings of a study that I am doing with AT&T looking at the impact of the Internet on work. This study has led us to an expansion and redefinition of traditional ethnographic methods.

Inevitably, new technology brings change in many arenas. As businesses adopt new tools and move from one technology to another, there are always changes in jobs and work activities. With the meteoric increase in Internet access, a fundamental question is what is going to happen to these as a result:

What will be the impact of the Internet on work, both in the near-term and the long-term?

How will emerging electronic tools change the way we interact with our customers, our co-workers, and our organizations?

These and similar questions were the basis for a basic research project underway at AT&T on the impact of the Internet on work. We focused on a subset of occupations and were particularly interested in how the Internet is changing jobs of people who are on the cutting edge, who we call "techno-pioneers". This is because we think specific near-term changes by those on the cutting edge suggest more long-term trends. Therefore, by learning about the trends and shifts in the way they are doing their job(s), we have a glimpse of a potential future.

We chose to use an ethnographic approach instead of more traditional market research methods for a number of reasons. The research was largely exploratory and the issues were intentionally broad. The research focus on work activities required an understanding, at some level, of the physical work environment, specific work tools, and organizational structure and communication patterns. Understanding the context was believed to be absolutely critical to understanding the work activity.

We have done a series of visits to a variety of businesses in three cities in the US, and four cities outside the US. Each visit has lasted at least four hours and at each site, we talked to and observed techno-pioneers at work. Of particular interest are some of the challenges which we had to solve, including locating and recruiting participants, expanding the physical observations to include electronic "behavior", and dealing with the vast amounts of data which such a study produces.

Locating and observing participants. The first task we faced was finding our "techno-pioneers." This proved to be somewhat of a challenge. Because we were planning visits, we were geographically constrained. We wanted a broad sample, with a variety of industries and sizes. Traditional market research recruiting is optimized for these factors as long as they are recruiting a typical sample. However, the techno-pioneers we sought were by definition, "atypical." At first, we tried combining a variety of recruiting strategies, including traditional phone recruiting from a list, posting notices on local electronic bulletin boards, and World Wide Web "surfing" to look for potentially interesting sites. We quickly found that the first two strategies were not effective, and therefore, spent our efforts looking for interesting Web sites to identify candidates, who we contacted to invite into the study. This led to a very interesting set of sites.

One interesting consequence of this strategy was that many of the sites we located in this way were small, often very small, businesses. This confirmed what every book on doing business on the Internet mentions: that one of the advantages of the Internet for small businesses is that it is possible to have an electronic "storefront" that looks big, even if you are really a very small company. Even a one-person business can reach an international clientele, and can create a presence which allows them to compete with large, well established companies. (e.g., Resnick & Taylor, 1995) We could not tell in advance the size of a number of the businesses we arranged to visit.

The Web searching we did also made it difficult to pinpoint businesses geographically. Not all Web sites had addresses on them. We tried a number of search strategies, and ended up with a combination of constrained searches using zip codes, or area codes. In addition, we identified large regional Internet Service Providers (ISPs) and used their home pages to launch a search of pages they hosted. This led to a diverse sample of businesses.

Visits to sites. Our visits followed a traditional ethnographic approach as has been described elsewhere (e.g., see Dray & Mrazek, 1996.) We established rapport prior to the visit, and focused on expanding it upon arrival. Then, we proceeded to do a contextual interview, artifact walkthroughs, and finally, observation of work processes. We collected artifacts, took photos using both conventional and digital cameras, and audiotaped our visit. We also asked participants for pointers into their electronic sources, their "virtual watercooler" where they traded information with others in their profession.

Electronic ethnographies. This traditional ethnographic approach has also been augmented by what we are calling "electronic ethnographies" such as monitoring, or "hanging out" in Listservs identified by visit participants, watching Usenet groups, searching or "surfing" the World Wide Web, and doing electronic interviews with a subset of the most interesting people we "see" electronically. In a sense, we were observing behavior which in the past might have taken place around the water cooler, or in the lunchroom. Not surprisingly, we have found that augmenting observed behavior with electronic behavior has been extremely valuable. We have been able to confirm the trends we observed directly and to gather information from a much wider community than we would have been able to do using more conventional observational techniques.

Debriefing. After each visit, a debrief session was held in which the field observations were summarized. These sessions provided initial coding of the observation data and resulted in multi-perspective representations to be used for subsequent analysis. A debrief questionnaire was used to record specific ways the business and individual jobs had changed as a result of the Internet. A mind map, which is a graphical nonlinear representation of information [Wycoff, 1991], was developed to capture important communication patterns, interesting artifacts, and highlights about the physical environment. See Figure 1 for an example of a mind map record for one of the sites. And finally, a preliminary causal loop diagram [Miles & Huberman, 1994, Senge, Kleiner, Roberts, Ross, & Smith, 1994] was developed for each site to record our understanding of the interrelationships between key variables that we observed. Traditionally used much later in the analysis phase of qualitative studies, we found early use of such diagrams to be very helpful. See Figure 2 for an example of one of the preliminary causal loop diagrams.

Because the research team was geographically dispersed, we used the Web and email extensively to communicate with each other. The internal team used the AT&T Intranet to post web sites with full debriefing summaries, including the digital photographs, mind maps, and causal loop diagrams. These were electronically shipped as HTML documents to the consultant, who was outside the firewall.

Dealing with data. Since we are interested in the effects of the Internet on jobs and work activities, we decided to examine the data using a formal job classification scheme. The current Dictionary of Occupational Titles, one of the oldest and most detailed classification schemes, is currently being revised to reflect more up-to-date job content by a multi-agency team from government and private industry in a project called "O*NET". Their job content model includes of a set of generic job elements or characteristics of work. We have used this characterization of work activities, which focuses information inputs, mental processing, work outputs, and interacting with others, as a start for categorizing the shifts in observed work activities. [Jeanneret & Borman, 1995].

As with most large qualitative studies, the task of sifting through the field data was daunting. Due to the flood of documents on the Web, however, document analysis tools have become widely available and are no longer the exclusive tools of ethnographers. Several of these tools were evaluated and the field data was imported into a document data base, called an "infobase," and the document analysis software used to code the observed changes in work activities. Color highlighting, see in Figure 3, was used to code the text for later searching, retrieval and printing.

We have identified a number of changes in work processes, in roles, and in work content itself. Accompanying these are changes in the inputs and information needed, in the products or outputs of work, and in the skills required to do the new job. The techno-pioneers' work differed, at times dramatically, from what it had been before. Each occupation which we have looked at has had all of these types of changes to differing degrees.

One of the occupations we focused on was Newspaper Reporters. As Internet technology is infusing in to the newsroom, many subtle and not-so-subtle shifts are beginning. The "new breed" of so-called "computer-assisted reporters" combine the traditional reporting skills with new technical skills such as sophisticated database manipulation, complex Internet search strategies, and multidimensional data visualization. Some of these skills are more commonly taught in doctoral level social science courses than in traditional journalism programs.

The Internet has also begun to change the content of reporters' work. The technology of data bases and Internet searches has helped these reporters to change their workflow and allows them to ask better, more focused questions, often at an entirely different level. These reporters do not see the rise of technology tools as a substitute for what they somewhat colorfully call "shoe leather reporting." Rather than providing only answers, the data also provides richer questions. In the past, a reporter would ask questions, and then obtain often pre-processed data from an informed source which they would then report. Now, with easy and rapid access to database tools, that same reporter can come prepared to ask the data-provider questions that can be both more detailed, as well as more strategic. For instance, instead of looking at an overall percentage change in a budget, the reporter can update with live figures and do on-the-fly calculations to identify line item changes from a previous year's figures. Questions can be targeted at the line item level or at a more meta-level by providing the reporter with extremely rapid access to the tools to do ad hoc data manipulation. At a more profound level, these skills reduce computer assisted reporters reliance on informed sources to mediate between them and the raw data.

Still pioneers within their own newsrooms, many of these technology-leaders have sought out an electronic community of like-minded reporters, also breaking new ground in their respective papers. In their electronic community, there is a lively give-and-take as they ask each other for leads, often to other electronic resources, make suggestions, help each other with technical and training issues, and share what they have learned. The enthusiasm, bordering on zeal, of the active participants, and the sense of being pioneers in what is becoming a swiftly changing business is infectious. It is clear why they have become the ones who have the ground breakers in their organizations. Many of them take upon themselves the mission of introducing new tools and training others, inside and outside their home organizations in addition to their "real" job of reporter, editor, and beat writer. One interesting related observation is that these internal technology leaders may be evolving a new role or status for themselves and those they teach within their organizations that has little to do with their formal hierarchical placement.

A second occupational category that we investigated was teachers. We saw dramatic changes in the work activities of teachers brought about by the Internet. Clearly the Internet has the potential to alter the way that teachers gather information. We observed teachers searching the Web for lesson plans, project ideas and content. Not surprisingly, there can be significant problems with Internet accessibility for many teachers, as well as problems finding useful information and weeding out the rest. We observed the use of software to import a small portion of the web for off-line use. For those teachers with metered Internet access this reduced online charges. It also provided content stability by making a portion of the Web static. A history lesson plan that was developed could be reused without having to worry about broken links.

We visited one charter school in which the teacher's traditional role as information gatherer and deliverer had been completely transformed. Rather than being the expert, the teacher acted as facilitator or advisor as the students set about finding, gathering and evaluating information. In this technology infused school, students now gathered online news articles about the 1996 U.S. presidential election, gathered information for a planned a student trip to Washington D.C., and researched the similarities and differences between Lake Michigan and Lake Baikal. Students also sent email queries to people outside of the school in search of additional information and learned how to assess the quality of the information they found. The teachers new role was to advise the students on search strategies and point them to knowledgeable peers. Given the great diversity of information sources available to the students, the teaching role also had shifted from evaluating content to evaluating process and critical thinking.

The Internet changed the way teachers interact with peers, students, parents, and the community. For example, school announcements, directory and calendar information is communicated through the web. Electronic mail was used by teachers to communicate with each other, parents and students. The asynchronous nature of email allowed greatly increased communication flow. While some strictly social email was observed among the students, it was also used to share project information. In another school that supported distance learning we were told of students completing entire classes without face-to-face meetings.

As part of the debrief session for each site, the research team did a short process evaluation. One lesson was that traditional recruiting methods may not be very efficient for a techno-pioneer sampling strategy such as ours. Only after trying several creative search strategies were we able to find interesting organizations to visit. A second lesson is that pioneers can be very good pointers to other pioneers. The trade publications that pioneers read, the electronic mail lists that they subscribe to, the training programs that they have developed, and the special industry meetings that they attend are all great sources of leads for other pioneer visits. A third lesson is that a mix of shoe leather and electronic ethnography leads to a much richer understanding of a work community.

There are also several lessons about how Internet technology is changing work and the way workers interact with each other. Looking at the leading edge of occupational communities has revealed many changes. These include changing ways of getting information, processing that information and producing work output. The changing communications patterns developing as a result of the Web and email are also transforming work relationships in complex ways. For those who are connected to the Internet, there can be increased communication flow to a greatly expanded electronic community.

Dray, S. & D. Mrazek (1996) "A day in the life:" Studying context across cultures. In J. Nielsen & E. del Galdo, Eds. International User Interfaces. (New York: John Wiley & Sons.)

Miles, M. & Huberman, A. (1994) Qualitative Data Analysis (Thousand Oaks, CA: Sage Publications )

Resnick, R. & D. Taylor (1995) The Internet Business Guide (Indianapolis, IN: Sams.net Publishing.)

Senge, P., Kleiner, A, Roberts, C., Ross, R., & Smith, B. (1994) The Fifth Discipline Fieldbook. (New York: Doubleday Currency).

Jeanneret, P.& Borman, W. (1995) Generalized Work Activities. In Development of Occupational Information Network (O*NET) Content Model. Utah Department of Employment Security. (See http://www.doleta.gov/programs/onet/occreq.htm)

Wycoff, J. (1991) Mindmapping. (New York: Berkley Books)

Return to articles list