A goal-based weighting for composite indicators constructed through Ordered Weighted Averaging (OWA) operator

Area of Science:

• Multicriteria decision analysis
• Composite indicator development
• Quantitative social sciences

Background:

• Ordered Weighted Averaging (OWA) operators are widely used in composite indicator research.
• Traditional OWA weight determination relies on decision-makers, introducing potential bias and errors.
• Existing objective methods for OWA weight definition do not prioritize composite indicator quality.

Purpose of the Study:

• To introduce a novel method for defining OWA operator weights based on composite indicator quality.
• To address the limitations of subjective and existing objective weight determination methods.
• To enhance the reliability and interpretability of composite indicators.

Main Methods:

• The proposed method defines OWA weights by optimizing two key quality parameters: concept capture and informational loss.
• The methodology is implementable using Microsoft Excel Solver.
• The approach allows for flexibility in setting optimization objectives and constraints, including limiting informational loss or emphasizing specific phenomenon aspects.

Main Results:

• Weights are determined to maximize the composite indicator's ability to represent the multidimensional phenomenon.
• The method incorporates constraints to manage informational loss and tailor the indicator's focus.
• The optimization framework offers adaptability for diverse multidimensional problems.

Conclusions:

• The developed method provides an objective and quality-focused approach to OWA weight determination.
• This technique improves the scientific rigor and practical applicability of composite indicators.
• The method is flexible, user-friendly, and suitable for researchers and practitioners in various fields.