Discussion

---

Based on our knowledge of other collaborative applications, we do not believe that Groove is unique in its lack of privacy support. Privacy is enhanced in applications that provide access control on a document-by-document basis and that provide the capability to address chat messages to subsets of participants (e.g., a private, “whispered” chat message). In general, however, most collaborative applications are akin to Groove in that they aim to provide support to participants who wish to share everything, all the time. Such an approach is not consistent with the move toward greater computer-aided collaboration across organizations with somewhat divergent goals.

Of all the evaluation methods discussed, SCAPE is the only method to specifically focus on whether an application supports the awareness and privacy needs of a particular group of users. SCAPE does not address basic usability issues such as whether the interface was designed to be consistent or to have clearly marked exits. Inspection techniques developed for single-user applications can provide this type of insight, even for multiuser systems, and should generally be used. Thus, SCAPE should be used in conjunction with techniques such as Nielsen’s heuristic evaluation and function checklists. If an IOIS will be used in situations where awareness and privacy are important concerns, however, a SCAPE analysis would be appropriate and beneficial.

SCAPE does not involve users as evaluation subjects; this is both a strength and a weakness. Usability testing (structured testing with users) can be very expensive and time-consuming but yields rich data. Rather than forgoing evaluation completely, however, an inspection evaluation method constitutes a less expensive and less time-consuming approach. We recommend that usability testing be conducted if doing so involves small numbers of users, the users’ environment can be easily recreated or accessed, and the user population is readily available. Even if usability testing is performed and a full SCAPE evaluation or other inspection evaluation is not attempted, we recommend performing the first three steps of SCAPE to understand the roles that users play and the need to provide information to, or conceal information from, each other. Once SCAPE Awareness and Privacy Policies have been developed, they can act as a specification against which the application’s performance can be compared during a usability test. Awareness and Privacy Policies can also provide guidance in determining which tasks the users should be asked to perform to ensure that sensitive or critical information is revealed or concealed in certain situations.

If the buyer and sales representative in the case study scenario were truly looking for an IOIS to manage information sharing in an adversarial collaboration environment, they would be well advised to perform SCAPE evaluations on several systems as part of a comparative analysis. At a minimum, the most sensitive situations should be included in scenarios developed for the SCAPE evaluation. For example, SCAPE quickly highlighted the fact that Groove does not provide for documents to be read by only a subset of the application’s users; this fact alone means that Groove may not be a suitable choice for this particular group of users.

---