An agreed level of collaboration
The researcher and those being researched reach an agreement about the level of collaboration in any one project.
In a lot of research, decisions are taken by the external researcher. In mainstream research approaches, research is owned by an external (often academic) researcher who makes decisions about the research. This often means it is their ideas and priorities that drive the research, rather than those of the people or organisations being researched. Research is often most useful and empowering if it is partly or fully owned by the people who are its subject. Following this model (see for example Participatory Action Research, there is dialogue at all stages, and ideas/skills/benefits flow both ways.
But, while two-way collaboration in research is an ideal, it may not always be realistic or desired by Transition members. This was clear from the findings of WP1. Non-collaborative research may also be valuable, as Transition members may not have the time or resources to take part in research. The important principle to be adhered to in all cases, is that such research should also be discussed and the level of collaboration planned together from the start. But in all circumstances, the level of collaboration should be agreed from the outset, and continuing dialogue and checking back in allows for changes in input.
This principle does not mean that the pattern of the research (or the level of collaboration) is fixed. Research is, and should be, an emergent and sometimes unpredictable process, and plans may change during the course of the research due to new questions arising, unforeseen events, etc (see Space for the unexpected).
The level of collaboration in research between external researchers and Transition should be agreed at the outset. Everyone involved should have, at least, the opportunity for input into decisions about Transition research, its design and outcomes. Finding appropriate ways to share power and ownership, if that is required, underpins this. There is no one size fits all model for doing this, as people, contexts and circumstances vary. To make it work, knowledges/ideas/skills need to be valued equally.