I can speak mainly on cause/​intervention prioritization research—the TLDR is all of us research organizations are pretty terrible at this (e.g. by last count I think at least 4 orgs have looked into hypertension/​salt and have come to fairly similar conclusions). From various discussions, I think people are interested in increasing publication of internal research avoid this problem, but folks have also pointed out that duplication is beneficial—if a cause looks good when different organizations with different methodologies investigate it, that’s a reliable sign that it really is good; and conversely, we don’t want to permanently dismiss an idea just because a single shallow report (which may well be wrong) was negative on it.
FWIW, CEARCH keeps a longlist of causes, and we try to link to existing research where available, but it’s not remotely close to comprehensive, except for CEARCH’s and CE’s output.
Thank you Joel for pointing me in the right direction. You make a good point on not wanting to prematurely write off research findings In response to the challenges in cause/​intervention prioritization and the benefits of research duplication:
Bayesian Updating for Research Analysis: Implement a Bayesian updating system to continuously revise research certainty, accommodating new data and insights.
Visual Credibility Indicators: Use a color gradient in a table format to represent research certainty levels, making it easier to interpret findings at a glance.
Research Depth and Uncertainty Metrics: Track the extent and depth of research, focusing on areas with high uncertainty or potential for further exploration.
Estimating Further Research Potential: Assess the potential for additional insights in each research area, similar to estimating untapped resources in land surveying.
Goal: To refine research prioritization, recognize the value of duplication for verification, and avoid premature dismissal of ideas based on limited data.
Potential Pitfalls:
Over-reliance on Quantitative Metrics: Solely relying on quantitative metrics might overlook qualitative aspects of research, leading to incomplete evaluations.
Misinterpretation of Data: The color gradient system could be misinterpreted if not accompanied by clear guidelines, leading to incorrect conclusions.
System Complexity: Implementing a Bayesian updating system could be complex and require significant resources for accurate and effective functioning.
Bias in Data Input: The system’s effectiveness is contingent on the quality and neutrality of the input data, with biased inputs leading to skewed outcomes.
Ignoring Contextual Factors: The system might not fully account for contextual factors affecting research, such as socio-economic or cultural aspects.
I can speak mainly on cause/​intervention prioritization research—the TLDR is all of us research organizations are pretty terrible at this (e.g. by last count I think at least 4 orgs have looked into hypertension/​salt and have come to fairly similar conclusions). From various discussions, I think people are interested in increasing publication of internal research avoid this problem, but folks have also pointed out that duplication is beneficial—if a cause looks good when different organizations with different methodologies investigate it, that’s a reliable sign that it really is good; and conversely, we don’t want to permanently dismiss an idea just because a single shallow report (which may well be wrong) was negative on it.
FWIW, CEARCH keeps a longlist of causes, and we try to link to existing research where available, but it’s not remotely close to comprehensive, except for CEARCH’s and CE’s output.
Thank you Joel for pointing me in the right direction. You make a good point on not wanting to prematurely write off research findings
In response to the challenges in cause/​intervention prioritization and the benefits of research duplication:
Bayesian Updating for Research Analysis: Implement a Bayesian updating system to continuously revise research certainty, accommodating new data and insights.
Visual Credibility Indicators: Use a color gradient in a table format to represent research certainty levels, making it easier to interpret findings at a glance.
Research Depth and Uncertainty Metrics: Track the extent and depth of research, focusing on areas with high uncertainty or potential for further exploration.
Estimating Further Research Potential: Assess the potential for additional insights in each research area, similar to estimating untapped resources in land surveying.
Goal: To refine research prioritization, recognize the value of duplication for verification, and avoid premature dismissal of ideas based on limited data.
Potential Pitfalls:
Over-reliance on Quantitative Metrics: Solely relying on quantitative metrics might overlook qualitative aspects of research, leading to incomplete evaluations.
Misinterpretation of Data: The color gradient system could be misinterpreted if not accompanied by clear guidelines, leading to incorrect conclusions.
System Complexity: Implementing a Bayesian updating system could be complex and require significant resources for accurate and effective functioning.
Bias in Data Input: The system’s effectiveness is contingent on the quality and neutrality of the input data, with biased inputs leading to skewed outcomes.
Ignoring Contextual Factors: The system might not fully account for contextual factors affecting research, such as socio-economic or cultural aspects.