"... In PAGE 7: ... Each parameter combination in our simulations were replicated 15 times, for 1800 time steps (of 1 s) each. We examined our model output with a suite of 11 aggregation indices that encompass individual- to population-level char- acteristics ( Table2 ). Statistics were based on time averages taken over the second half of each simulation (900 time steps), to avoid undue influence from initial conditions.... In PAGE 8: ... Population level: number of groups, number of collisions, number of stragglers, and the Clark and Evans (1954) index. See Table2 for details on how each index was computed. 303... ..."

"... In PAGE 5: ...1 Predicting Execution Times Our case analysis now proceeds with a usability evaluation with the KLM (Keystroke-Level Model) [3,4]. This model provides quantita- tive predictions of human performance based on the descriptions of the actions in Table1 . In the KLM each action is converted into a sequence of mental and motor operators (shown in Table 2), whose individual execution times have been empirically established and validated in psychological experiences [3,15].... In PAGE 6: ... This is met in our case since Sophie and Charles are highly trained in drawing line connections and using the groupware tool, as assumed in the case description. To exemplify the conversion from the detailed textual description into a KLM representation, consider the RELEASE_CELL action in Table1 and the illustration in Figure 3. Figure 3.... ..."

"... In PAGE 13: ... well understood procedures and goals . ability to communicate in small groups as well as a whole team Summary Table4 provides a summary of the different types of collaboration spaces as well as the key environmental features, benefits and constraints associated with each type of space. The research summarized in this section and in Table 4 suggests that the design of colla- borative space should derive from a careful consider- ation of the context and the nature of the work, especially the on-going communication needs, the cognitive complexity of the group work and the extent to which the work is truly mobile.... In PAGE 13: ... ability to communicate in small groups as well as a whole team Summary Table 4 provides a summary of the different types of collaboration spaces as well as the key environmental features, benefits and constraints associated with each type of space. The research summarized in this section and in Table4 suggests that the design of colla- borative space should derive from a careful consider- ation of the context and the nature of the work, especially the on-going communication needs, the cognitive complexity of the group work and the extent to which the work is truly mobile. Individual aspects of collaborative knowledge work As noted above, collaboration is defined as a system of behaviours that includes individual, focused work as well as interaction.... ..."

"... In PAGE 2: ... In this subsection, we propose a simplified collaborative group definition (SCGD) [4] that is a simplified description of the syntax, or form, of individual statements. Table1 shows a set of language specification, which is in the Backus-Naur form, for the SCGD. The syntax provides a description for efficiently collaborative session in the system.... ..."

"... In PAGE 20: ...5, and is the best trade-off between the two quantities (in the absence of information regarding the relative costs of a false alarm and a miss). Table1 shows the detection and false alarm rates for each of the four evaluators. Table 2 shows the error rates for the OR combination, MAJ combination and the BN output using the threshold just described.... ..."

"... In PAGE 19: ...5, and is the best trade-off between the two quantities (in the absence of information regarding the relative costs of a false alarm and a miss). Table1 shows the detection and false alarm rates for each of the four evaluators. Table 2 shows the error rates for the OR combination, MAJ combination and the BN output using the threshold just described.... ..."

"... In PAGE 4: ... A total of 130 undergraduate students of Brunel University participated in the second study. RESULTS Table4 shows the average rating for the individual skins. The rating ranges from 1.... In PAGE 6: ...indicated by + in Table4 . The ratings of these skins were used in aggregated category measures by taking the average value of their z-scores.... ..."

"... In PAGE 26: ...25 each one composed of 12 individually marked ows, are the input of the network, where they are remarked with di erent committed rates, looking for a weighted sharing of the available bandwidth. The committed rate of each aggregate is shown in Table7 . They also have di erent RTTs, chosen randomly from 40 to 300 ms, and the starting time goes from 0 to 1 second.... ..."

"... In PAGE 20: ...3 Contributions of Neighborhoods In order to analyze how the different neighborhood structures of VNS contribute to the whole optimization, we logged how often each one was able to improve on a current solution and their absolute gains. Table6 shows the ratios of successful improvements in contrast to how often each neighborhood structure was evaluated. These values are grouped by the different types of input instances.... ..."

"... In PAGE 4: ... Table5 : Summary: Which user group is interested in which functionality? COMPUTATIONS When executing queries, all computation is done based on the two relations likes and matches from Figure 3. Table 6 shows how the individual cases can be computed.... ..."