Tuesday, October 07, 2008

Open Space: Developing research proposals

The first open space session was based around discussing different internal and external processes to develop good research proposals. In many organizations, there are mechanisms put in place that evaluate projects before they are approved, however, there seems to be a need to look at this process from different levels and angles to see how interventions could be done at earlier stages to ensure right people are involved in the project proposal phase from the very beginning.

Some suggestions from experiences in different organizations (of varying levels of mandate and size) were:
  • Bring communications aspect early in the project development process.
  • Liaise with the final beneficiaries to see if the priorities are going to be met.
  • National priorities also need to be looked at to ensure linkages between them and the expected research outputs.
  • Have a collaborative platform to share ideas of projects as they are being developed. This will give opportunity to anyone in the organization to propose their ideas. Develop the platform with a small group as a proof-of-concept so that the results speak for themselves.
  • Scientist should discuss the project either around or just after concept note rather than towards the end of the process (like in clearance phase).
  • Another solution could be to have a repository of projects at different levels.
  • The leadership has to recognise value of sharing and discussing projects from the beginning.
  • Alliances of institutions and people is also important; assess and look at what is available and what others are doing.
  • Leverage on your quick-wins to demonstrate the immediate benefits.
  • All, those who propose projects as well as those who evaluate and approve them, have a role to ensure that projects are aligned with strategic objectives of the organization and fit in the full picture.
Notes from this session are also available at: http://www.kstoolkit.org/Report+on+session+1A

No comments: