Analyzing Shared Workspaces Design with Human-Performance Models

Abstract

Abstract. We propose an analytic method to evaluate synchronous shared workspaces design. The method uses human-performance models, developed in the Human-Computer Interaction field, to make time predictions about collaborative actions performed in selected critical scenarios. We apply this method to two case studies: the design of a collaborative game and the redesign of a collaborative tool for software engineering requirements negotiation. The benefits and limitations of the method are discussed, as well as some implications for design.

Communicative NGOMSL: An Extension of NGOMSL to Analyze a Text-based Communication System

“…In the recent situation that cooperative work using a computerized system as a mediate becomes more popular as the related technologies are more developed, one outstanding limitation of past GOMS techniques (including NGOMSL) is that they are applicable only to analyzing individuals, and are not appropriate to analyzing a group's interaction with a cooperative system (Antunes, Borges, Pino, & Carrico, 2005; Antunes, Ferreira, & Pino, 2006;Kieras & Santoro, 2004;Min, Koo, Chung, & Kim, 1999). These GOMS techniques do not consider the allocation of subtasks and the communication among humans in a complex system (Min et al, 1999).…”

“…In this system, several persons with individual roles monitored a display and coordinated their actions via a shared radio communication channel. In other studies, a group of researchers (Antunes et al, 2005;Antunes, Borges et al, 2006;Antunes, Ferreira et al, 2006;Ferreira & Antunes, 2007) proposed a new analytic method which was inspired by GOMS to evaluate synchronous shared workspace design. To validate this method, they applied it to the design of a collaborative game and that of a collaborative tool for software engineering requirements negotiation.…”

Evaluation Methods for Groupware Systems

“…HPM describe how a person interacts with a physical interface at a low level of detail based on a cognitive architecture, e.g., the keystroke level model (KLM) approximates the interaction of a single user with an interface. HPM adapts this model to a group of users communicating through a shared workspace [2]. In this method, evaluators first decompose the physical interface into several shared workspaces.…”

“…First, methods for single-user systems are not always applicable to groupware, since the outcome depends on the various backgrounds of group members, organizational culture and group dynamics [5]. Second, evaluation may be expensive and the required resources may be unavailable [2]. Third, a groupware system's benefits may be long-term, so group observation should extend over long periods [5].…”

A Technique for Evaluating Shared Workspaces Efficiency

“…In this paper we propose a technique, based on earlier research on the benefits of using human-performance models [6], to provide additional insights about workspace collaboration, not covered by other evaluation techniques. Some advantages of this technique emerge from the following characteristics of human-performance models:…”