Nice start. This will simply be useful for DOING investigations of causes and interventions, especially new ones with limited credible evidence. It will be helpful to have an informative prior about the impactfulness about that class of interventions.
Ideally, I would use this prior in a Bayesian framework, in conjunction with the evidence presented, to generate a posterior about the impact, and the cost-effectiveness of the proposed intervention.
Without a sense of what the prior should look like, this proves extremely challenging
(Maybe Iām stating the obvious here, but...)
Nice start. This will simply be useful for DOING investigations of causes and interventions, especially new ones with limited credible evidence. It will be helpful to have an informative prior about the impactfulness about that class of interventions.
Ideally, I would use this prior in a Bayesian framework, in conjunction with the evidence presented, to generate a posterior about the impact, and the cost-effectiveness of the proposed intervention.
Without a sense of what the prior should look like, this proves extremely challenging