Probabilistic inversion in priority setting of emerging zoonoses.

Area of Science:

• Decision Analysis
• Operations Research
• Risk Management

Background:

• Priority setting often relies on subjective expert judgment.
• Quantifying and validating expert preferences in complex scenarios is challenging.
• Existing methods may not fully capture the nuances of multi-criteria decision-making.

Purpose of the Study:

• To develop a quantitative methodology for priority setting using expert judgment.
• To integrate probabilistic inversion with multi-criteria analysis for robust decision support.
• To validate the proposed methodology through rigorous testing.

Main Methods:

• Experts ranked scenarios based on perceived severity.
• A linear multi-criteria analysis model was developed to represent expert preferences.
• Probabilistic inversion was employed to derive a distribution of attribute weights.
• Model validation included checking expert consistency, overall fit, and predictive accuracy on out-of-sample data.

Main Results:

• The probabilistic inversion successfully identified attribute weight distributions that reproduced expert rankings.
• The fitted model demonstrated a strong ability to replicate observed expert judgments.
• Validation confirmed the model's robustness and predictive capability when applied to new data subsets.

Conclusions:

• The combination of probabilistic inversion and expert judgment offers a powerful tool for priority setting.
• This methodology provides a data-driven approach to quantify and leverage expert knowledge.
• The validated model enhances the reliability and transparency of decision-making in complex problems.