Where? - Collaborative Web Search: Who, What, Where, When, and Why - page 39

Table 4.1: Design space of collaborative search applications for tabletop displays
( Morris et al. , 2010a ).

Group
Configuration
Collaborative
Style

Tightly-
coupled

Transitioning Parallel work

Group size Pairs Small teams Large groups

Location Co-located Distributed Mixed-presence

Device ecology Table Table +

additional
outputs/displays

Table +
Additional input
devices/objects

Import
data

Export
data

Search
Task
Application
domain

Web Multimedia Personal

documents

Database

Search inputs Keywords Visual language Touch/Gestures Tangible A.I.

Analysis stages Search Sensemaking Reporting

joint queries are interpreted conjunctively (keyboards facing each other) or disjunctively (keyboards
side-by-side). WebSurface ( Tuddenham et al. , 2009 ) (Figure 4.16) supports co-located Web brows-
ing by providing users with a high-resolution projected display; clicking links opens pages in a
separate browser window, so that related pages can be viewed side-by-side on the large surface
for easy comparison. WeSearch ( Morris et al. , 2010b ) is a tabletop system designed to support not
only collaborative Web searching, but also the accompanying group sensemaking task; WeSearch is
discussed in greater detail below.
While systems such as WeSearch and WebSurface employ shared displays in order to take
advantage of a face-to-face work arrangement, sometimes co-located collaborators employ shared
displays not out of preference but due to resource limitations. Although computers have become
more plentiful in recent years, shared computer use remains common in many settings. For example,
in U.S. public schools, the ratio of students to computers is 3.8 to 1 ( U.S. Dept. of Education ,
2006a ), and the number of Internet-enabled computers available in U.S. public libraries is 3 for
every 5,000 people ( U.S. Dept. of Education , 2006b ). In developing countries, these ratios can be
even more skewed. For example, in rural schools in developing countries, the student-to-computer
ratio can be as high 10 to 1 ( Pawar et al. , 2006 ). Even when resource constraints are not a factor,
the social and pedagogical benefits of face-to-face collaboration and shared viewing of information
can be a compelling reason for collaborators to share a single computer ( Stewart et al. , 1999 ). For
example, in a recent survey of 204 Microsoft employees, 87.7% reported engaging in “backseat
driver” searches where they watched over another person's shoulder and suggest query terms to try
or links to click ( Morris, M.R. , 2008 ).