one of the big challenges here is persuading the Research Councils and Research England to do anything differently on the basis of the data analytics that you suggest collecting. We have always had good data about geographic distribution of grant money for example, but there has never been any willingness to do anything to the "proposal-peer review-issue grants" system that would alter the outcome in a way consistent with policy objectives like levelling up. That said, I fully support your thinking - we have focused for may years on growing the _inputs_ to the research system, with a pleasant but naive assumption that if we looked after the inputs the outputs would take care of themselves. I don't think we can or should take that for granted.
Hi Stian - as you can imagine this is all music to my ears!
Additional considerations:
- Invest in the boring stuff: build data and modeling infrastructure to deliver insights closer to real time vs. discrete reports and papers that are out of date by the time they are published.
- Co-locate data scientists and policy experts: ensure that government bodies have embedded analytical capabilities that can respond to urgent policy priorities; complement this tactical capability with R&D centres of expertise / external teams developing new methods to push the frontier. A hot take here is to minimise reliance on consultants who lack local context / advanced capabilities.
- Use mixed methods: combine data analysis, experiments and qualitative information from interviews and stakeholder engagement to create virtuous cycles of pattern recognition / hypothesis development and testing.
Stian,
one of the big challenges here is persuading the Research Councils and Research England to do anything differently on the basis of the data analytics that you suggest collecting. We have always had good data about geographic distribution of grant money for example, but there has never been any willingness to do anything to the "proposal-peer review-issue grants" system that would alter the outcome in a way consistent with policy objectives like levelling up. That said, I fully support your thinking - we have focused for may years on growing the _inputs_ to the research system, with a pleasant but naive assumption that if we looked after the inputs the outputs would take care of themselves. I don't think we can or should take that for granted.
good luck!
John
Hi Stian - as you can imagine this is all music to my ears!
Additional considerations:
- Invest in the boring stuff: build data and modeling infrastructure to deliver insights closer to real time vs. discrete reports and papers that are out of date by the time they are published.
- Co-locate data scientists and policy experts: ensure that government bodies have embedded analytical capabilities that can respond to urgent policy priorities; complement this tactical capability with R&D centres of expertise / external teams developing new methods to push the frontier. A hot take here is to minimise reliance on consultants who lack local context / advanced capabilities.
- Use mixed methods: combine data analysis, experiments and qualitative information from interviews and stakeholder engagement to create virtuous cycles of pattern recognition / hypothesis development and testing.