Designing pareto-optimal selection systems for multiple minority subgroups and multiple criteria.

Area of Science:

• Organizational Psychology
• Decision Science
• Human Resource Management

Background:

• Current Pareto-optimal (PO) methods for selection systems are limited to one minority group and one criterion.
• Increasing workplace diversity and multiple selection criteria necessitate advanced approaches.
• Existing methods fail to capture the full spectrum of trade-offs in complex selection scenarios.

Purpose of the Study:

• To extend existing PO selection system design methods to handle multiple criteria and multiple minority groups.
• To develop a hybrid multiobjective PO approach for balancing quality and diversity objectives.
• To propose procedures for aiding designers in selecting optimal systems from multiple objectives.

Main Methods:

• A hybrid multiobjective PO approach was developed to compute selection systems balancing multiple quality and diversity objectives.
• Three two-dimensional subspace procedures were proposed to assist in selecting among PO systems.
• The methods were illustrated with example applications and validated through a cross-validation study.

Main Results:

• The novel multiobjective PO approaches reveal the complete set of eligible PO selection designs.
• These methods faithfully capture the Pareto trade-off front for more than two objectives.
• Validated PO selection designs demonstrated an advantage over alternative methods in new samples.

Conclusions:

• The developed multiobjective PO selection systems effectively address the complexities of diverse workforces and multiple selection criteria.
• The proposed methods provide a comprehensive framework for designing equitable and effective selection systems.
• Executable code is available to facilitate the implementation of these advanced PO approaches.