Sunday, August 12, 2007

Improving science innovation

To experience most significant scientific advances, humans are dependent on the clunky unreengineered process of science innovation and deployment. Potential improvements to the innovation phase are discussed below.

In the absence of clear feedback loops aligning research investigations with implemented results, scientists can languish in isolated labs for years and the majority do not seem to care whether their findings are useful to or implemented by others. For type A scientists, the in-place incentive system is academic publishing and acknowledgment. Publishing is a codependent phenomenon with scientific publications increasingly exerting influence over the direction of research to generate more interesting reading.

Suggested Improvements

1. Open human knowledge databases
Without yet destabilizing the publishing juggernaut, some progress could be made in releasing already published and unpublishable findings into open databases of human knowledge. There are some early examples of these resources in Physics with ArXiv, the NIH's PubMed and the Earth System Grid for climate research, however there is an opportunity for a new layer of applications to make the information much more accessible to different levels of audiences.

The next three suggestions have to do with creating accountability and a better feedback loop between scientific findings and the use of that information.

2. Quantitative values attached to findings
A system of quantitative values could be applied to research so that findings and scientists could be measured and compared. Supervisors, peers and industry colleagues could rank findings based on a variety of parameters. Unpublishable and null findings would also be incorporated into the valuation program.

3. Annual performance reviews for scientists
The rigor of quarterly goal setting and review, 360 degree feedback and other performance evaluation metrics implemented decades ago in the business environment should also be de rigueur in the scientific community. Performance metrics would be a good start, incorporating what are now standard corporate principles of leadership, communication and management science to reduce subjectivity and otherwise improve scientific working environments would also be helpful.

4. Broader scientific mindset
The most successful scientists have been those who have perceived their roles as not the mere discovers and handers-off of the Truth but also as being responsible for rendering their findings implementable by others. Emphasizing full realization of pursuits and results from a more service-driven than ego-driven mindset could also produce better results more quickly.


Deepak said...

The last one in your list (broader scientific mindset) is too often overlooked. One of the most fascinating aspects of science is learning and getting ideas from outside ones own field of expertise. Too few scientists go that route.

LaBlogga said...

Hi Deepak, Thanks for the comment and it was nice meeting you recently. I think that broader mindset - both in the multidisciplinary sense that you mention and in the implementation sense, e.g.; not being satisfied with discovery and leaving implementation to others but having a greater stake in leading implementation since many findings never become useful would be helpful.

LaBlogga said...

In fact, some helpful innovation has arrived with SciVee,

The National Science Foundation, Public Library of Science and the San Diego Supercomputing Center have partnered to set up what can best be described as a "YouTube for scientists." Scientists can upload their research papers, accompanied by a video where they describe the work in the form of a short lecture, accompanied by a presentation.

Deepak said...


I've been a regular reader for a while and it was good to finally get a chance for a brief chat.

Thanks for leaving the note on my blog. SciVee is great. It sorts of validates everything we believed in when we started Bioscreencast. Science needs to move forward from the current accepted practices of communication and embrace new media and other paradigms. That the NSF is involved with SciVee is very encouraging

Another trend that this demonstrates is something that is already happening in tech with CDNs, publishing platforms and publishers all existing symbiotically. The tools are there. We just need to take full advantage.

LaBlogga said...

Hi Deepak, Thanks for the comments and for leading the way in helping create and deploy science tools.