Throughout the approximately 15 - 20 days of each group’s study session, participants used both the group-based and one-to-one messaging versions of Slam, each for half of the total time. Differences in the total time for each group were due to issues with scheduling all members of the groups to meet for the introductory and wrap-up sessions. The order for which messaging condition participants experienced first was counterbalanced across the groups in order to evenly distribute any novelty or other order effects. The number of weekends also was distributed evenly across the messaging conditions.
Questionnaire feedback was solicited immediately following each messaging condition. That is, at the midpoint of the study, participants were sent email asking them to complete a questionnaire regarding their social experiences with whichever version of Slam they had just completed. They then switched to the other version of the software, and completed the questionnaire on use of that version at the end of the study.
Data collected were of three types: system usage data, questionnaire responses, and round-table discussion feedback. Usage data were used to assess the quantity of social interactions as measured by the number of messages sent. All messages were also coded for message type to get a sense of how the two messaging conditions impacted the ways in which participants communicated using the software. The questionnaires were used to gather participant’s subjective sense of the impact each messaging modality had on their social relationships with other group members. Finally, the round-table discussions were used to capture qualitative feedback on participant’s experiences with both versions of the system.
3.2.1 Usage Data
Usage logs were analyzed for the number of messages sent in each messaging condition. Initial scans of the logs showed a bit of a novelty effect where participants used the system heavily the first day or two. Although the order of the two messaging
Figure 6: Participants sent more messages when communicating group-wide