Web-chat and information: a
field study using CoBrow
The aim was to evaluate the uses of synchronous
chat on Web-sites in relation to their information-providing
functions. As a case study, for several days in September
1999, the CoBrow chat program was installed on a mirror
of a university prospectus Web-site. This allowed visitors
to seek advice from a representative of the university
whiIe expIoring the site. The conversational and information-seeking
behaviour of visitors was observed and a questionnaire
was administered. Fewer than 7% of the visitors to the
parent site accessed the mirror site. Most visitors
may have been looking for opportunities to chat with
other visitors rather than a representative of the university.
Those visitors who asked questions of the representative
were primarily interested in seeking clarification and
interpretation of published information in relation
to their personal circumstances. Chat could clarify
misunderstandings; but text messages were time-consuming
and lacked conversational cues. There was not enough
evidence from which to draw conclusions about attitudes
to information-seeking, the effects of cultural or gender
differences, or the effect of chat on perceptions of
the site and the university. Some technical problems
in CoBrow were identified.
In a general survey, 599 Web-sites were identified
as having offered chat facilities; only 35% still
did so. Between 45,000 and 95,000 people were estimated
to be using chat on Web-sites at any one time, as
against 164,000 to 246,000 using Internet Relay Chat,
and 4.8 million users of the Web in any one hour.
Most sites made little attempt to define uses of chat,
or to control access or behaviour in using chat. Conversations
were analysed on a sample of sites, using the same
categories as in the case study. Scheduled conversations
included more substantive items of information and
a lower proportion of social messages than casual
conversations; but the distinction between "social"
and "information" messages depends on the purposes
of the site. The categories used to analyse these
conversations helped develop a checklist of questions
that may clarify the information needs that chat on
a Web-site may serve.
(Patrick Wallace: MSc Information
Science, City University, 1999)
|
Links
|