• Home
• About Us
• Business Process Engineering
• Training
• Usability Testing
• Knowledge Management and Decision Making
• Organizational Design
• Applied Cognition and Perception
• Medical Systems
• Workplace Ergonomics
• Command and Control
• Contact Us

Predicting Future Manpower Needs

The Approach

PSE conducted a statistical analysis of a manpower model that forecast the number of bar pilots needed to meet anticipated work shift demands.

We focused on the statistical properties and predictive power of the model.

We also suggested alternative statistical modeling approaches to improve predictive power and accuracy.

The Solution

We found that the current model included many constants (e.g., peak multiplier, sick leave factor, seasonal factor) that ignored day-to-day variations critical to predictive accuracy. The model also did not take into account the interdependence among variables, which was important for developing an accurate prediction.

Time Series Analysis was recommended in order to account for fluctuations in workload over time. Various analytical strategies (e.g., hierarchical regression analysis) were also recommended to further prioritize important predictor variables.

Additional Information

Glaser, D. N. (2001). Expert Review of 1986 San Francisco Pilots Manpower Model. Pacific Science & Engineering Group, Inc.