Improving distributed collaborative writing over the Internet using enhanced processes, proximity choices and a Java-based collaborative writing tool.

摘要

This research focuses on improving distributed, collaborative writing (CW) over the Internet through new CW technologies, process improvements, proximity choices, and synchronicity choices. The research methodology that is followed iteratively builds theory using design and development of CW tools, surveys, requirements analysis, laboratory and field experiments. The primary empirical work consists of two laboratory experiments and two fields experiments.; The two laboratory experiments compare a newly built CW tool, Collaboratus to Microsoft Word™. The first experiment found 3-member CW groups using Collaboratus produce higher document lengths with higher quality, but less satisfaction than similar Word™ groups. The second experiment found 3-member distributed, synchronous CW groups using Collaboratus with NetMeeting™, working over 5 weeks, produce higher document lengths and quality, and have more effective coordination, communication, and socialization than similar Word™ groups.; The two field experiments further validate that certain processes should be followed for most effective use of Collaboratus. The first experiment compares 3-member groups working in mixed work modes (part F2F part distributed, asynchronous) to similar groups conducting all work asynchronously. The only differences is that all-asynchronous groups spend less time brainstorming and more time converging on brainstormed output. The lack of process gains for the mixed-mode groups can be attributed to the difficulty of scheduling F2F meetings—an insight gained by performing this research in a field setting. The second experiment compares non-facilitated, 3-member groups using three different levels of scripted process structure to guide their distributed, asynchronous CW. Highest structure groups have the highest levels of performance (in terms of quality, productivity, satisfaction, communication, and relationships), while lowest structure groups have the lowest levels of performance. It is believed the process differences can be directly attributed to the nature of the academic writing task and the fact the groups were nearly formed, non-cohesive groups.; These findings point to the importance of properly matching technology, task, tool, proximity, synchronicity, and people choices in distributed CW to maximize the outcomes. This also provides a useful foundation for agent-based CW processes to support self-sustaining, distributed CW teams who work without professional facilitation.