SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. A Web Storybase
  5. Storybase Use
  6. The Story Corpus
  7. Story Themes
  8. Evolution of the Storybase
  9. Story Annotations
  10. Discussion
  11. Acknowledgments
  12. References
  13. Copyright Information

Although the use of the World Wide Web has expanded tremendously in the past few years, we still know very little about how users are working with this new medium, what they attempt to accomplish, what works, what doesn’ t. One way to answer such questions is to simply ask users for descriptions of their own activities. This paper presents an analysis of 133 stories of Web use contributed by users over a period of 40 months, since the Web Storybase began operation in December 1994. Usage of the Storybase is examined in general, and the stories are analyzed along several dimensions. The stories convey usage experiences that not only involve global information retrieval and person-to-person contact, but also the development of both good and bad interpersonal relationships, as well as extensive reflection on how the Web is changing our lives.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. A Web Storybase
  5. Storybase Use
  6. The Story Corpus
  7. Story Themes
  8. Evolution of the Storybase
  9. Story Annotations
  10. Discussion
  11. Acknowledgments
  12. References
  13. Copyright Information

Over the past few years, the Internet has become the place to be. Work groups routinely use it as a place to archive and share group activities [7, 9]. Educators increasingly rely on Web-based research as an enhancement to normal classroom activities [2, 10, 13]. Individuals faced with a wide range of tasks, from adoption of a puppy to filling out a college application form, now often begin their tasks by searching the Web for relevant background information. Businesses have spent massive amounts of money building company pages and services [19]. But perhaps what is most striking about all of this frantic activity is the rapidity with which Web use has evolved.

Three years ago, the Web was still largely the dominion of the technological elite, the researchers and scientists who had been using networks for some time, and had followed the development of HTML and associated browsers. But once several robust (and free!) HTML browsers became commonly available, the user population expanded tremendously, making the last three years a fascinating case of technology adoption. In this paper we consider this adoption process from the perspective of the users who were part of it–we analyze a corpus of stories and annotations contributed to a Web Storybase over a period of 40 months. These stories suggest that the Web has been successful in its most salient role, helping users find and contact information and other people. However, the stories also reveal a certain amount of turmoil in users’ more personal experiences–activities involving relationships with others, and reactions to how the Web is intruding upon and changing our lives.

A Web Storybase

  1. Top of page
  2. Abstract
  3. Introduction
  4. A Web Storybase
  5. Storybase Use
  6. The Story Corpus
  7. Story Themes
  8. Evolution of the Storybase
  9. Story Annotations
  10. Discussion
  11. Acknowledgments
  12. References
  13. Copyright Information

In the Fall of 1994, the World Wide Web was still a novel domain for research in human-computer interaction (HCI). The ready availability of free HTML browsers was beginning to attract a wide variety of new users, and businesses were working frantically to build Web presences that could exploit the widespread interest. Most Web interactions still involved only information retrieval. Interactive forms had just been added to the HTML markup language, and many users didn’ t even know that an opportunity for greater interactivity existed. There was much excitement, confusion, and hope about how Web technology might affect our lives.

We created the Storybase at this time with several goals in mind. First and foremost, we simply wanted to learn something about how the Web was being used. Other HCI researchers were beginning to carry out demographic surveys [14], and a few more detailed lab studies of Web navigation techniques were beginning to appear [3]. However virtually nothing was known about the tasks people were actually carrying out on the Web, what was being tried, what worked, and what didn’ t. Our prior work on scenario-based design [4, 5] suggested that the collection of user “war stories” (see also [12]) would be a useful source of data for characterizing use.

A secondary goal was to investigate interactive Web applications–to experiment with the emerging interactive forms technology, and to investigate its potential for building collaborative authoring systems. We could have created a newsgroup to collect Web use experiences, but HTML and the freely available Web browsers allowed us to provide a more attractive and functional user interface (e.g., HTML-based graphics and text formatting, hyperlinks, automatic notification of story annotations). By making the story capture mechanism an instance of Web use itself, we hoped to gather stories that were more directly situated in contributor's real world usage context.

The Web Storybase was designed for general Web users, so we kept its functionality to a minimum. Visitors to the site (see http://miso.cs.vt.edu/story) are welcomed with an introduction inviting them to share a notable experience concerning their use of the Web:

  • We feel confident that YOU too have a story to tell. Maybe it's one you've already shared with friends or colleagues; maybe it's something that just happened and is fresh in your mind. Or, maybe something in the stories below will cause a reaction, or make you remember a related experience. Browse or search the stories we have now; when you're ready to contribute your own story, follow the link below.

On this same page is a complete index of the stories that have been submitted, sorted in reverse chronological order (where order is defined by the date of submission or of last annotation if any). Figure 1 provides a partial view of this index; the stories themselves are accessed through their title links. If the author provides contact information, a name, and/or email address, this is listed; if provided an email link is created to aid in contacting the author. Other summary information is also provided–the story's date of submission, any keywords selected by the author, and its thematic category (this is provided after submission by the moderator after reviewing the submission and will be discussed later in the paper). Finally there is a summary of annotation activity, the number of annotations added thus far and the date of the most recent.

image

Figure 1. Partial listing of stories as indexed on the Storybase home page.

Download figure to PowerPoint

Visitors wishing to add a new story follow the “compose” link. This takes them to a page with several input forms, one corresponding to each of the pieces of story meta-data (e.g. the title, author, etc.), and one for the text of the story itself. Tips are available for how to use the forms, including a recommendation to write the text of the story in another text editor and then paste it in. Authors are told that they can use HTML tags to format the text or provide links to other URLs, but that conventional formatting (e.g., carriage returns, tabs, extra spaces) will be ignored. After submitting the story, an author is shown the new story page. The browser's “Back” button can be used to return to the original forms, which can be edited and resubmitted if desired. When a story is submitted an acknowledgement is sent by email if an address is provided; the Storybase administrator is also notified.

At the bottom of each story page (i.e., following all annotations), browsers are invited to “Add an Annotation”, a link that brings them to a page with forms for entering an annotation. The originating story is displayed at the top, followed by a set of forms similar to those used for submitting stories; however less summary data is collected (just author contact information). After submission, the annotations appear in chronological order directly following the story; there is no further nesting of discussion threads. Each time an annotation is submitted, an email notification is sent to the originating author (if an email address has been provided) and to the Storybase administrator.

In keeping with our goals of simplicity, the initial Storybase design provided just one mechanism for browsing the stories–the story index on the welcome page (although we did sometimes receive visitors who arrived at a specific page due to its inclusion as a search result set). However approximately two years into the project, we decided that the list was too long to be conveniently browsed (we had over 100 stories at that point). Thus we added two new mechanisms, a within-site search and six thematic filters. The search engine matches a user-provided string against all story texts and meta-data. The filters are based on a high-level content analysis of the stories: we identified six themes (Browsers and tools, Cyber-relationships, Finding it all, Getting started, Making connections, Web culture) that were sufficient to categorize the entire set, and added a category attribute to each story's meta-data. If users select one of these themes (see Figure 2) they are taken to a story index page where the listing has been filtered to include only stories with that theme.

image

Figure 2. Icons/links leading to story lists filtered by story theme.

Download figure to PowerPoint

Storybase Use

  1. Top of page
  2. Abstract
  3. Introduction
  4. A Web Storybase
  5. Storybase Use
  6. The Story Corpus
  7. Story Themes
  8. Evolution of the Storybase
  9. Story Annotations
  10. Discussion
  11. Acknowledgments
  12. References
  13. Copyright Information

The Storybase application was launched on December 14, 1994. Initially it was publicized only internally (i.e., within the Computer Science Department at Virginia Tech). In March 1995, after several months of trouble-free operation, we announced the system on two major Web launching pads (Yahoo, Mosaic What's New). This produced a flurry of site accesses (see Figure 3; note that we began collecting logging data in March 1995), and the application was quickly picked up by other recommendation pages (e.g., Delphi Innovative Web Sites, the Collaboration sites of the World Wide Web Consortium, GEnie Hotspots). However, although the site was getting quite a few visitors by this point, not many were posting material, perhaps due to a lack of critical mass in the archive [18]. Thus we made an effort to recruit users more directly, using both personal contacts and postings on various newsgroups. By the summer of 1995, we had about 25 stories and the application was well underway.

image

Figure 3. Frequency of accesses and posts to Web Storybase over 38 months during which use was monitored.

Download figure to PowerPoint

The graph in Figure 3 summarizes the site's activity over the 38 months for which we archived server logging data. Posts (combining stories and annotations) are graphed as a line on top of histograms depicting monthly page requests (for any page within the site, including the compose and annotate forms). The initial blip in accesses is presumably due to its announcement on two well-known Web launching pads; accesses then died down somewhat but began to climb again in a few months, presumably due to individuals and organizations adding links from elsewhere on the Web.

It is interesting to note the somewhat cyclical nature of site accesses–it appears that site accesses rise, fall, then rise and fall again, and so on. We can only speculate as to possible causes for such a phenomenon, as we have no systematic data about the users who visit the site. Our main source of insight comes from informal browsing of the usage logs, and from looking for patterns in the population of users contributing stories. These informal reviews suggest that the site may be at times “shared” within a community, who visit, use it, but then perhaps move on. In a few cases, such a pattern seems quite clear, because there were many requests from workstations at a single location. One interesting line of research might be to develop log analysis techniques that could identify such “pockets” of users, and develop a model for sharing news of Web applications within work-related or social networks.

As is typical of online forums, users who are interested only in browsing far out-number those who are willing to contribute. Over the 38 months of logging, the Storybase pages were accessed 65,623 times, and contributions were posted 789 times, yielding an access-to-post ratio of approximately 84:1. Note that posts also show local peaks and valleys. In this case it is more clear that this is due to temporary adoption of the site by a group of active users. Although we have not carried out a formal analysis of individual use (this is actually quite difficult to do, as users are not required to identify themselves, and even those that do may use different identities at different times), we can point to many examples of a single individual making numerous contributions over several days or weeks; some contributions also make specific reference to an earlier visit or visits.

As one would expect, the frequency of posts is related to the frequency of accesses (r = 0.59, df = 36, p < .001). This general result has a simple explanation: assuming that likely contributors are evenly distributed throughout the general Web population, then the larger the number of Web users who visit, the larger the number there are who will choose to contribute something. However, the last ten months or so of activity appear to tell a different story. While the frequency of accesses has remained relatively high–at the same general magnitude as the previous ten months–the frequency of posts has fallen dramatically, back to the level observed in the first few months of use. The ratio of accesses to posts is 70:1 during the period of Mar95-May98, but rises to 160:1 during Jun97-Apr98. For the first 28 months the access-to-post correlation is strong (r = 0.77, df = 27, p < .001), but in the last 11 months this relationship disappears (r =−0.27, df = 9, ns).

It may be premature to conclude that this is a real difference, but as of now there appears to be no evidence of a “corrective” shift in the frequency of posts. Perhaps the Storybase has become too large; visitors coming for the first time now may feel that they have nothing new to add, or may be unwilling to spend enough time browsing to find something that inspires annotation. Alternatively, Web users may be starting to “outgrow” the need for such a site–perhaps most have worked through their early experiences and are no longer interested in reflecting on and sharing the things that have happened along the way.

The Story Corpus

  1. Top of page
  2. Abstract
  3. Introduction
  4. A Web Storybase
  5. Storybase Use
  6. The Story Corpus
  7. Story Themes
  8. Evolution of the Storybase
  9. Story Annotations
  10. Discussion
  11. Acknowledgments
  12. References
  13. Copyright Information

In the 40 months of its operation, users have posted 281 stories to the archive. However, only 133 of these stories remain in the listing. Every time a story is posted, the Storybase moderator (Rosson) is notified so that the story content can be vetted for conformance with the archive's goals; this process led to the removal of 146 contributions. The majority of these (100, or 68.5%) were rejected because they were off-topic; users attempted to use this forum to “publish” material on a variety of topics unrelated to Web use (e.g., rock band preferences, short fiction, poetry); 34 of the discards included obscene or off-color remarks. A smaller number of stories (22) were removed because they duplicated other stories–users sometimes became confused about the status of their contribution and submitted it several times before realizing they needed to refresh the index screen. Finally, seven contributions were simply tests of the site's posting mechanism by users who were presumably curious to see how the interactive forms worked.

Two stories were submitted and appeared in the listing for a while but were eventually withdrawn by request of the authors. Both of these were stories describing relationship issues associated with the Web. In one case, the author began to feel awkward about the number of “doubting Thomas” remarks she was receiving in response to her upbeat–but perhaps somewhat naive–story about a cyber-relationship that was just about to culminate in an in-person meeting, soon to be followed by marriage. In the second case, a wife shared a rather bitter story of how her marriage broke up as the result of extra-marital cyber-affairs, but then became concerned that the friends of the husband so accused were using her posting of this story to incite even greater friction.

The stories vary considerably in length, ranging from several lines to over a page of single spaced text. In contrast, most of the story annotations are quite short, often a single line comment that essentially conveys a reaction such as “Me too!” or “Good for you!”. The style of most of the stories is quite informal; many contain misspellings and/or grammatical errors. Indeed one rather heated discussion thread distributed throughout the archive concerned the presence of small errors such as these, and whether the authors ought to be criticized for submitting such pieces. A few stories had clearly been written as examples of creative writing, and one interesting outcome of several such stories and their annotations was the reported creation of an email-based “writing club”.

The majority of the stories related something of a personal nature about the author. Sometimes this was simply a detail about his or her life activities or interests (e.g., that she likes to ride horses, that he recently took a trip to North Carolina, that she hopes to be a writer). Other times the stories included rather intimate details about the authors’ (or in a few cases a friend's) love life or inner turmoil about some issue. Overall, 81 of the stories (about 60%) contained personal information of this sort (see Figure 4). The remaining stories were classified as generic, because they described a usage experience or concern without mentioning personal information about the author. These general findings are quite consistent with a prior analysis of the first 27 stories, where we also found that about 60% of the stories contained personal information [15].

image

Figure 4. Stories grouped by level of intimacy.

Download figure to PowerPoint

Our conclusion now as then is simply that users seem to be quite comfortable revealing personal–even quite intimate–details about their lives in this very public forum. We suspect that users’ apparent comfort in sharing their personal life is related in part to the social distancing created by a Web forum. Even though publishing details about yourself on the Web means that a huge number of people can now know something about your life, you still remain relatively anonymous, because of the size of the audience and the absence of real world proximity. Greater social distance has been found to decrease the social inhibition of communicators [16]. Thus some of our story authors may have felt comfortable using this forum to share an experience or work through an issue that they otherwise might have been too shy or inhibited to broach [17].

Both positive and negative experiences were shared (81 vs. 52 stories respectively, 61.7% vs. 38.3%). Positive stories often described a successful venture relating to the Web, a pleasant experience, or an optimistic view of what the Web has to offer. Negative stories often complained about Web-related events, services or tools, or worried about potential or current problems with how the Web is being used. The presence of both positive and negative stories indicates that users are as willing to recount good things about this new technology domain as to complain about it; indeed the preference for positive contributions suggests that users are still generally positive about this new communication medium [15].

A majority of the stories received at least one annotation; only 27 (20.3%) evoked no response at all. A total of 571 annotations were submitted, for an average of 4.29 annotations per story. However the variance in annotations was large, ranging from 1 to 25. Interestingly, although annotations were also reviewed by the archive moderator, only two were removed due to obscene language. It may be that the considerably lower occurrence of rude or obscene remarks as annotations versus stories is also related to perceived social distance. An off-color comment added to someone else's story implicitly directs the obscenity to the originating author, whereas contributing a story containing profane language directs it to the world in general. Having a more specific target may decrease perceived social distance and thus the likelihood of socially inappropriate remarks [16].

Story Themes

  1. Top of page
  2. Abstract
  3. Introduction
  4. A Web Storybase
  5. Storybase Use
  6. The Story Corpus
  7. Story Themes
  8. Evolution of the Storybase
  9. Story Annotations
  10. Discussion
  11. Acknowledgments
  12. References
  13. Copyright Information

In order to characterize the kinds of experiences that users were reporting, we developed a content analysis that categorized each story by the main theme. Because our goal was also to use these themes for organizing access to the growing list of stories, we attempted to keep the number of categories small. Ultimately we settled on six general themes:

  • Browsers and tools–stories conveying success or problems with physical interactions with the Web, (e.g., browser features, network service problems, platform differences, HTML authoring)
  • Cyber-relationships–stories describing virtual relationships of all sorts, often including some discussion of real-life components
  • Finding it all–experiences in which the Web served as an excellent information source, sharing a specially useful or inspiring bit of information found, describing situations in which in material was retrieved just at the moment it was needed
  • Getting started–similar to Browsers and tools, but for the special category of users encountering the Web for the first time, or still identifying themselves as “newbies”
  • Making connections–similar to Finding it all, but focusing on using the Web to find or contact people, including friends, family members, professional contacts, etc.
  • Web culture–reflections on what it is like to use the Web, what new experiences it enables (e.g., publishing by hopeful authors), how it is changing people's lives

Table 1 provides information about each category, the total number of stories it accounts for and some illustrative titles. As the table indicates, the themes of Cyber-relationships and Web culture are by far the most common, accounting for over half of the 133 stories currently in the archive (27.1% and 26.3% respectively). The themes Browsers and tools, Finding it all, and Getting started, were relatively uncommon (10.5%, 10.5%, and 9.8% respectively), while Making connections was intermediate (15.8%). This general pattern suggests that what our story authors found most interesting–or at least most worthy of sharing–was experiences that conveyed their personal interactions and their views of the Web as a new “place to hang out”. Descriptions of how particular Web technology, services or information were contributing to or interfering with these activities seemed to be less worthy of sharing.

Table 1.  Story Themes with Examples
ThemeExample TitlesFrequency
Browsers and toolsShouldn't Have Bought a Mac; Hot List Hemorrhage; Color!; Trying to Negotiate the Web with One Hand Tied14
Cyber-relationshipsWhy Must it End?; A Broken Relationship; Internet Love; I Get By With a Little Help From My Cyber-Friends36
Finding it allWeb Homework; Yahoo! The Hotspot is a Cool Monastery; Dorm Sweet Dorm; Vacation Planning14
Getting startedI Can't Get Anywhere; Virgin; Helpless on the Internet; Where I am Now!13
Making connectionsFinding Long Lost Friends; Looking for Lisa; Pour Ceux Qui Parle Francais; Grandma at the Keyboard21
Web cultureBoneheads on the Web; What Happened to the Soul?; Confessions of an E-Bully; Overcoming Openness35

Although stories with a positive message were more frequent overall, the relative frequency of positive to negative varied across the six categories (χ2(5) = 17.71, p < .005). Recall that the ratio of positive to negative stories was approximately 3:2. As depicted graphically in Figure 5, only the Web culture stories approximate this ratio, although Getting started and Cyber-relationships are close. However, Finding it all and Making connections contain much higher proportions of positive stories (85.7% and 81.0% respectively), while Browsers and tools differs in the opposite direction (27.3%).

image

Figure 5. Number of stories with positive or negative tone within each general theme.

Download figure to PowerPoint

It is interesting that the proportion of positive experiences related to Web technology is so small relative to the overall proportion; ironically this conflicts with one of our starting goals, to understand what was both good and bad about the technology. However, it may be that users are likely to reflect on the tools they are using

only when the tools become “present-to-hand” through “breakdowns” [18]. The relatively high proportion of positive experiences falling in the Finding it all and Making connections categories may simply indicate that these are the services that the Web has been most successful in providing–locating information, including information about other persons you have an interest in contacting. The more intermediate ratios in Web culture and Cyber-relationships are perhaps a sign that virtual relationships and efforts at understanding this new culture are still quite a mixed bag; although many are excited and convinced all will be wonderful, many others are experiencing various forms of distress, or worrying that the technology is pushing us in the wrong direction.

Evolution of the Storybase

  1. Top of page
  2. Abstract
  3. Introduction
  4. A Web Storybase
  5. Storybase Use
  6. The Story Corpus
  7. Story Themes
  8. Evolution of the Storybase
  9. Story Annotations
  10. Discussion
  11. Acknowledgments
  12. References
  13. Copyright Information

The Storybase has been in continuous operation for three and a half years, since mid-December 1994. Coincidentally this time period corresponds closely to the emergence of the Internet as an accepted element of popular culture. For example, one of our first stories (from December 1994) remarks on the unusual experience of seeing a URL in USA Today. In June 1998, it is difficult to find a mass media offering (including television and radio) that does not include at least some mention of Web-based material.

The longevity of the story archive enables an interesting evolutionary view of Web use. Storybase is an open system, so we cannot control who visits or who decides to contribute a story. However we are able to answer questions about the kinds of stories that have been contributed and whether this content seems to have changed in character over the three years.

Figure 6 illustrates the evolution in the number of stories contributed across the six themes. Within each theme the stories have been grouped according to when they were contributed: in the first 13 months of operation (38 stories, dark bars), the middle 13 months (63 stories, medium bars), or the final 14 months (32 stories, light bars. As the figure suggests, there was tremendous variation in thematic content over these three time periods (χ2(10) = 34.03, p < .001). For example, the category of Cyber-relationships contained 0% of the stories in the first period, 38.1% in the second, and 37.5% in the third; Web culture went from 18.4%, to 27%, to 34.4%. In contrast, the relative number of stories in the other four categories diminished over the 40 months, with Finding it all showing the largest drop, from 21%, to 9.5%, to 0%.

image

Figure 6. Percent of stories in each category broken down into thirds (i.e., the first 13 months, the middle 13 months, and the final 14 months)

Download figure to PowerPoint

This pronounced shift in the content of the stories contributed has at least two plausible interpretations. Storybase is an open system and thus anyone on the Web can visit and contribute material. The evolution we observed may reflect a change in the user population visiting and contributing material, for example toward an increasing percentage of users who happen to be interested in virtual relationships and Web culture. We have no way of evaluating this interpretation, though it seems that at the least such a change would have to have been a complex one–while the Cyber-relationships stories seem to be authored primarily by Chat room afficiandos, the authors of the Web culture stories appear to be much more diverse, with the stories covering everything from personal interests in Web authoring, to the immensity and often impersonality of the Internet, to characterizations of different kinds of Web users.

Another interpretation would see this evolution as a more straightforward reflection of how users’ experiences with–and concerns about–the Internet are changing. This interpretation has good face validity: it seems quite logical that there would be a relative decrease in stories about use of Web tools, or of getting started in general, over a time period in which the Internet was transformed from a specialized network for researchers, to a relatively commonplace communication tool for the general populace. This view would argue that many users have moved beyond the initial excitement of finding some particular piece of information or locating an old college friend (see e.g., our analysis [15] of the first nine months of use), and are wondering what more there is. Some are exploring the kinds of interpersonal relationships that can be created and sustained, but are curious or concerned about how these will work, and how they compare to real life relationships. Others are taking the time to sit back and reflect on how this immense global network is changing the way we live and carry out activities.

A related analysis shows that the relative number of stories of a personal nature also increased over the 40 months of operation (see Figure 7; χ2(4) = 10.05, p < .05). During the first 13 months of operation, none of the stories contained any intimate content, although about half contained some personal information about the author's life. During the next 13 months, however, stories revealing intimate information began to appear, and continued to be contributed during the final 14 months.

image

Figure 7. Number of stories broken down by level of intimacy across three different time periods.

Download figure to PowerPoint

This finding is clearly related to the evolution of story themes: 18 of the 20 stories classified as intimate are in the Cyber-relationships category; the first of these did not appear until March 1996, the 16th month of Storybase operation. Like the changes in story themes, the changes in intimacy could be due to changes in user population, such that the more recent contributors represent a less socially inhibited sample of users. Alternatively, the effect may be due to some sort of critical mass phenomenon: after a few “free spirit” authors led the way with very personal stories, others may have felt comfortable disclosing details of their own private lives [18]. At the same time, authors wanting to share personal information may have felt more protected from possible ridicule or censure, as the database of stories grew in size.

Story Annotations

  1. Top of page
  2. Abstract
  3. Introduction
  4. A Web Storybase
  5. Storybase Use
  6. The Story Corpus
  7. Story Themes
  8. Evolution of the Storybase
  9. Story Annotations
  10. Discussion
  11. Acknowledgments
  12. References
  13. Copyright Information

Many contributions to Storybase are in the form of annotations to existing stories rather than (or in addition to) first-order stories; the archive contains more than four times as many annotations as stories. Of course, these annotations are sometimes stories themselves, but more often they convey a reaction to the content of the story, expressing approval, disapproval, encouragement, and so on. Given the wide variability in number of annotations per story (values ranged from 0 to 25, with standard deviation of 5.38), we wondered whether annotation frequency might be related to some of the variables used to characterize the stories.

Several factors might have differentially encouraged annotations. If Cyber-relationships and Web culture are especially “hot” topics (as evidenced by the relative number of stories contributed) then they may also provoke longer “conversations” in response to the experiences and concerns aired. Figure 8 shows the number of annotations contributed in each category, graphed on top of the number of annotations; the scale on the left is for story frequency, the one on the left is for annotations. As the figure suggests, there is some tendency for stories in Cyber-relationships, Getting started, Making connections, and Web culture to receive a higher proportion of annotations (averaging 4.89, 4.92, 5.00, and 4.54 respectively) than those in Browsers and tools, or Finding it all (3.21 and 1.57). However, this variation is only modestly significant (χ2(5) = 11.50, p < .05), and appears to be due largely to the small number of annotations on stories in the Finding it all category. This particular effect is not surprising: these stories are reports of “good finds” on the Web; the authors are either sharing their satisfaction or suggesting that others might also find their bit of information useful. Aside from evoking a brief word of thanks or a “me too” comment, such postings seem unlikely to be read as an invitation for discussion.

image

Figure 8. Relative frequency of stories and annotations posted within the of the six general content themes.

Download figure to PowerPoint

A story's level of intimacy might also be expected to affect response rate. When an individual reveals personal information about herself, the recipient of the information often responds in kind with similar personal information [1]. In our earlier report on the first nine months of Storybase operation, we were surprised that the more personal stories tended not to be annotated [15]. At that time we speculated that this could be due to the remoteness and relative anonymity of the disclosure, i.e. perhaps it didn’ t really “feel” personal. Now, looking back at these early stories (the first 27 contributions) we can also see that what we classified at the time as “personal” was not high in terms of self-disclosure. Such stories told of a specific personal experience, for example, an individual recounting how a friend sent him a URL related to his longterm fascination with airplanes. Though such stories reveal details of an individual's life, they are certainly not intimate. In contrast, most of the stories in the Cyber-relationships category reveal intimate details about romantic or even sexual relationships. Thus the current set of stories represents a much broader range of this personal disclosure variable.

Figure 9 graphs the total number of annotations made for stories classified as generic, personal and intimate. Despite the wider range of the intimacy factor, this analysis leads to a conclusion similar to that of the earlier report: readers are just as likely to respond to a generic story as to one of a personal or intimate nature. The ratio of annotations to stories is approximately equal across the three levels of intimacy (χ2(2) = 2.14, p > .30). Again, this is unexpected, given the history of work on intimacy and self-disclosure. It appears that simply sharing personal information is not always enough to evoke responses from others. As we suggested in the earlier analysis, it may be that the remoteness and anonymity of a public Web application works against the feeling of intimacy provoked through self-disclosure. In a public forum such as this, there must be other more salient story characteristics that evoke responsiveness, perhaps having to do with the communication goals that are implicit in a narrative.

image

Figure 9. Relative frequency of annotations to stories, broken down by the stories’ level of intimacy.

Download figure to PowerPoint

A story's tone–i.e., whether it reports a positive or a negative experience–may provide one indication of the author's goals in sharing his or her story. As mentioned earlier, negative stories typically relate some complaint about a Web tool or service, or about an inability to find satisfaction in some endeavor conducted over the Web. As a consequence, these stories may often contain an implicit request for discussion, possibly searching for an answer to a specific question, sending out a plea for help, or trying to understand why some problematic situation occurred. The data summarized Figure 10 are consistent with this reasoning: readers are indeed more likely to respond to negative than to positive stories. Even though positive stories are more common than negative stories, the ratio of average annotations per story is much greater for those of negative tone (5.8 vs. 3.3, χ2(1) = 28.23, p < .0001).

image

Figure 10. Relative frequency of annotations to stories grouped by positive or negative tone.

Download figure to PowerPoint

Although a detailed analysis of the conversational structure for each of the 133 stories is beyond the scope of this paper, a high-level examination of the “top ten” offers further insight into the effects of an author's communication goals. The first three colums of Table 2 list the title, month of submission, and number of annotations for these ten stories. The fourth column conveys an initial attempt at categorizing the communicative “goal” of the story; these goals were identified simply by studying the story and trying to infer what the author might have been trying to accomplish in sharing it.

Table 2.  Highly Annotated Stories
TitleSubmittedAnnotationsGoal
1 Frustration, confusion, aggravation, exultationDec 199625Seeking friends
2 Finding long lost friendsDec 199721Sharing success
3 I can't get anywhereApr 199721Asking for help
4 A broken relationshipJul 199619Warning
5 Boneheads on the WebMay 199519Complaint
6 Shouldn't have bought a MacDec 199418Complaint
7 A sad and happy timeNov 199618Seeking friends
8 Looking for LisaOct 199517Seeking friends
9 What happened to the soul?Mar 199716Open discussion
10 Confessions of an E-BullyFeb 199615Boasting

In three of these ten cases (stories 7, 8, and 9), the author's request for a continuing interaction was quite explicit: story 7 concludes with an invitation for others to begin a friendship (7); story 8 asks for help in locating a long lost friend (8); and story 9 opens a discussion on how the Internet is distancing people from one another (9). However, the annotators do not always respond in the ways requested. For example, in the annotations to story 8, no one was able to provide any information about “Lisa”. Instead, many annotators initiated new dialogs of their own, bootstrapping off the original author's goal to request information on their own long lost friends! In this case the discussion was not a conversation at all, but really a specific request followed by a series of “me too” similar requests.

In other cases, the discussion following a story was just what one would expect given the starting goal. In “I can’ t get anywhere” a newbie described with frustration her initial experiences on the Web. Although she didn’ t specifically request help, her implicit plea for some guidance on how to make this new medium work for her evoked many sympathetic and constructive responses. Similarly, the thoughtful piece “What happened to the soul?” produced a number of genuine and reflective replies. The stories complaining about Web content quality and tools (stories 6 and 7) tended to evoke responses of the sort “get a life, stop complaining”; a story boasting about “virtual” tricks pulled on friends (story 10) produced a number of remonstrative comments; and a story sharing the hazards of virtual relationships (story 4) elicited additional evidence for this implicit warning.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. A Web Storybase
  5. Storybase Use
  6. The Story Corpus
  7. Story Themes
  8. Evolution of the Storybase
  9. Story Annotations
  10. Discussion
  11. Acknowledgments
  12. References
  13. Copyright Information

The stories and annotations contributed to the Web Storybase over the past few years reveal an interesting mix of experiences, both good and bad, both very personal and rather generic. If the trend in story content is indicative of trends in Web use, we would conclude that users have moved beyond a purely operational use of the Web (e.g., for information retrieval) to a more complex mix of work, play, and reflection about life in general. Of course all of the findings reported here must be qualified by the fact that these story data may have been influenced by a self-selection bias: many people don’ t have the time to surf the Web in search of places to leave personal remarks, of those with the time, many will be too shy or will have no interest in sharing their experiences in public. Thus the experiences described in Storybase may or may not be representative of the Web use in general.

Another qualification concerns use of the term “Web”. Although the Storybase invitation specifically requests experiences relating to the World Wide Web, many stories involved other Internet activities such as email and chat. To some extent, this may simply reflect the integration of other network services into Web browsers. However, it may also indicate a lack of separation between the concepts of “the Web” and “the Internet”; it may be that all activities involving remote communication and information gathering are merged together in the minds of many users. As a result, the experiences conveyed in the stories analyzed here cover a broader range of network technologies than we were originally attempting to assess.

These caveats aside, the prevalence of stories concerning interpersonal relationships evokes a picture far different from the “global information highway”. The Storybase may have tapped into an unusual user population, but even so, one must conclude that a major contribution of pervasive network access has been to encourage individuals to try new forms of relating with others–as well as to explore the use of quite public communication channels for working through relationships that are problematic [17]. These new ventures entail exploration of new techniques for virtual introduction, such as the contributor who wanted to get in touch with another author, but didn’ t want to leave her own email address. Instead, she annotated this woman's story, suggesting that she contact another author who had left his email, and who in turn would know how to contact her.

The relatively large number of stories and annotations reflecting on the nature of Web culture was also interesting. We began the project with the hope that users would reflect on Web tools and their user interfaces; the more philosophical perspective conveyed by these stories suggests that it is the the Web “way of life” that is capturing user's imagination, not the tools or interaction techniques that are so often the emphasis of work in HCI. One direction for future research would be a comprehensive, longitudinal study of a sample of Web users, to better articulate the lifestyle changes and values that accompany the shift to more Web-centric–and hence remotely connected–goals and activities [6, 11].

The evolution of story themes is just what one would expect to see as this new technology (and its user population) matures and becomes more stable. While “newbies” still drop by the Web Storybase on occasion, they are becoming much more rare. Routine users of the Web are starting to take for granted the ability to find just the right piece of information (or person) at just the right time. It would be interesting to look again at self-report data such as this several years from now, to see if cyber-relationships also will become “old hat” or whether there is something inherently unsettling in developing and maintaining relationships without the benefit of physical proximity.

Finally, the recent decline of posts relative to page accesses has raised the possibility that the Storybase itself might be undergoing a usage evolution, with visit rates remaining high but fewer users choosing to contribute something. We may have been truly fortunate to create and publicize the Storybase just when the general user population was in need of a comfortable forum for sharing their experiences in these changing times. If many users are now beginning to take the Web for granted, their need to share may have died back. Indeed one of our most prolific authors made just such an assertion in one of her final annotations. On the other hand, perhaps the role of forums like this is to remain in the background of the Web, just in case someone new comes along,

Everything comes to an end eventually. I am not saying that this page has run its course. There will be new people who will stumble upon it as accidentally as I did not so long ago. They will deal with their issues in this rather secure and friendly environment, then risk going somewhere else. There are other places to write and answer. They are building momentum. Keep coming here. You’ ll see.” (Storybase annotation submitted April 1997)

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. A Web Storybase
  5. Storybase Use
  6. The Story Corpus
  7. Story Themes
  8. Evolution of the Storybase
  9. Story Annotations
  10. Discussion
  11. Acknowledgments
  12. References
  13. Copyright Information

The original Web Storybase application was built by David Messner as an undergraduate research project; it was subsequently enhanced by John Kelso, Rob Loadwick, and Jonathan Hryn. Thanks also to John Carroll who collaborated on the original design and on the analysis of the first year of its use [15].

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. A Web Storybase
  5. Storybase Use
  6. The Story Corpus
  7. Story Themes
  8. Evolution of the Storybase
  9. Story Annotations
  10. Discussion
  11. Acknowledgments
  12. References
  13. Copyright Information
  • [1]
    J.L. Archer, “Self-disclosure”, in D.Wegner & R.Vallacher (Eds), The Self in Social Psychology, (pp. 183204), Oxford University Press, London , 1980.
  • [2]
    P. Bell, E.A. Davis, and M.C. Linn, “The Knowledge Integration Environment: Theory and Design”, in Proceedings of CSCL ‘95 (pp. 1421), Indiana University, Bloomington , IN , 1995.
  • [3]
    D.R. Britton and A.A. Reyes, “Discovering Usability Improvements for Mosaic: Application of the Contextual Inquiry Technique with an Expert User”, in Proceedings of the WWW Conference ‘94, 1994.
  • [4]
    J.M. Carroll, W.A. Kellogg, and M.B. Rosson, “The Task-Artifact Cycle”, in J.M.Carroll, (Ed.), Designing Interaction: Psychology at the Human-Computer Interface (pp. 74102), Cambridge University Press, Cambridge , UK , 1991.
  • [5]
    J.M. Carroll and M.B. Rosson, “Getting Around the Task-Artifact Cycle: How to Make Claims and Design by Scenario”, ACM Transactions on Information Systems, 10(2), 1992, pp. 181212.
  • [6]
    J.M. Carroll and M.B. Rosson, “Developing the Blacksburg Electronic Village”, Communications of the ACM, 39 (12), December, 1996, pp. 6974.
  • [7]
    A. Girgensohn, A. Lee, and K. Schlueter, “Experiences in Developing Collaborative Applications Using the World Wide Web “Shell”, in Hypertext'96: Seventh Annual ACM Conference on Hypertext (pp. 246255), ACM, New York , 1996
  • [8]
    M. Heidegger, Being and Time, Harper and Row, New York , 1962.
  • [9]
    E. Isaacs, J.C. Tang, and T. Morris, “Piazza: A Desktop Environment Supporting Impromptu and Planned Interactions”, in M.Ackerman, (Ed.), CSCW'96: Proceedings of the Conference on Computer-Supported Cooperative Work (pp. 315324), ACN, New York , 1996.
  • [10]
    J. Koenemann, J.M. Carroll, C.A. Shaffer, M.B. Rosson, and M. Abrams, “Designing Collaborative Applications for Classroom Use: The LiNC Project”, in The Design of Children's Technology, A.Druin, (Ed.), Morgan-Kaufmann, San Francisco , CA , 1998.
  • [11]
    R.E. Kraut, W. Scherlis, T. Mukhopadhyay, J. Manning, and S. Kiesler, “HomeNet: A Field Trial of Residential Internet Services”, Communications of the ACM, 39(12), 1996.
  • [12]
    J. E. Orr, “Narratives at Work”, in Proceedings of 1986 Conference on Computer-Supported Cooperative Work(pp. 6272), New York , ACM, 1986.
  • [13]
    R. Pea, “The Collaborative Visualization Project”, Communications of the ACM, 36(5), 1993, pp. 6063.
  • [14]
    J.E. Pitkow and M. Recker, “Results from the First World Wide Web Survey”, in Proceedings of the WWW Conference ‘94, 1994.
  • [15]
    M.B. Rosson, J.M. Carroll, and D. Messner, “A Web Storybase”, in People and Computers XI—Proceedings of HCI'96 (pp. 369382), Springer-Verlag, London , 1996.
  • [16]
    L. Sproul and S. Kiesler, Connections: New Ways of Working in the Networked Organization., MIT Press, Cambridge , MA , 1991.
  • [17]
    S. Turkle, Life on the Screen: Identity in the Age of the Internet, Simon and Schuster, New York , 1995.
  • [18]
    S. Whittaker, “Talking to Strangers: An Evaluation of the Factors Affecting Electronic Collaboration”, in M.Ackerman, (Ed.), CSCW'96: Proceedings of the Conference on Computer-Supported Cooperative Work (pp. 409418), ACM, New York , 1996.
  • [19]
    M.S. Yeoh, “Chapter 2: Demographics and demographic tools”, in WWW: Beyond the Basics. Computer Science Department, Virginia Tech. (http://ei.cs.vt.edu/~wwwbtb/book).

Copyright Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. A Web Storybase
  5. Storybase Use
  6. The Story Corpus
  7. Story Themes
  8. Evolution of the Storybase
  9. Story Annotations
  10. Discussion
  11. Acknowledgments
  12. References
  13. Copyright Information

Copyright © 1999 Institute of Electrical and Electronics Engineers. Reprinted from the CD ROM-based “Proceedings of the Thirty-Second Annual Hawaii International Conference on Systems Sciences,” 1999 (January 5–8, 1999, Maui, Hawaii. This material is posted here with permission of the IEEE. Internal or personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution must be obtained from the IEEE by sending a blank email message to http://info.pub.permission@ieee.org. By choosing to view this document, you agree to all provisions of the copyright laws protecting it.