Origins of Project A
Enabling of Project A
Specific Research Objectives
Overall Research Design
Research Instrument Development: Predictors
Job Analyses and Criterion Development
Modeling the Latent Structure of Performance
Correlations of Past Performance With Future Performance
Criterion-Related Validation
Some Broader Implications
Conclusions
References

This Chapter 1 is about personnel selection and classification research on a scale never before attempted in terms of (a) the types and variety of information collected, (b) the number of jobs that were considered simultaneously, (c) the size of the samples, and (d) the length of time that individuals were followed as they progressed through the organization.

The effort, commonly known as Project A, was sponsored by the US Army Research Institute for the Behavioral and Social Sciences (ARI). For contract management reasons the research program was conducted as two sequential projects: Project A (1982^–₇1989_35ya) and Career Force (1990^–₄1994_30ya), which worked from a single overall design (described subsequently).

Collectively, these projects attempted to evaluate the selection validity and classification efficiency of systematically sampled domains of prediction information for different selection and classification goals for the entire enlisted personnel system of the US Army, using various alternative decision rules (ie. “models”). Pursuing such ambitious objectives required the development of a comprehensive battery of new tests and inventories, the development of a wide variety of training and job performance measures for each job in the sample, four major worldwide data collections involving thousands of Army enlisted job incumbents for one to two days each, and the design and maintenance of the resulting database.

The truly difficult part was the never-ending need to develop a consensus among all of the project participants regarding literally hundreds of choices among measurement procedures, analysis methods, and data collection design strategies. Although many such decisions were made in the original design stage, many more occurred continuously as the projects moved forward, driven by the target dates for the major data collections, which absolutely could not be missed. The fact that all major parts of the projects were completed within the prescribed time frames and according to the specified research design was a source of wonder for all who participated.