The Army must meet quantity and quality goals in its recruiting. Recent advances in computer technology and psychometric theory have made possible a new type of assessment technique, called computerized adaptive testing (CAT), that can provide accurate estimates of ability based on relatively few test items. The Computerized Adaptive Screening Test (CAST) was designed to estimate a prospect's Armed Forces Qualification Test (AFQT) score at the recruiting station. Recruiters use prospects' CAST scores to determine whether applicants should be sent to Military Entrance Processing Stations for further testing. These scores also forecast the various options and benefits for which the prospects will subsequently qualify. This report summarizes analyses from a nation-wide cross-validation study and recommends that changes be made to CAST to improve its utility to recruiters. Keywords: Aptitude tests; Test construction(Psychology).
The Computerized Adaptive Screening Test (CAST) is used by Army recruiters to predict prospective applicants' subsequent performance on the Armed Forces Qualification Test (AFQT). A modified version of the CAST software was used in 60 recruiting stations across the country from January through December 1985 to collect CAST item-level performance information. Screening test data were matched to applicant records from Military Entrance Processing Stations to obtain ASVAB scores and relevant demographic information. The cross-validated, corrected correlation between CAST and AFQT scores is .83. CAST's ability to predict important AFQT performance categories and Army Aptitude Area scores was also examined. Alternative subtest lengths were evaluated and item bank characteristics were described. Keywords: Recruiting, Computerized adaptive testing (CAT), Enlistment screening test (EST).
Every year, the U.S. Army must select from an applicant pool in the hundreds of thousands to meet annual enlistment targets, currently numbering in the tens of thousands of new soldiers. A critical component of the selection process for enlisted service members is the formal assessments administered to applicants to determine their performance potential. Attrition for the U.S. military is hugely expensive. Every recruit that does not make it through basic training or beyond a first enlistment costs hundreds of thousands of dollars. Academic and other professional settings suffer similar losses when the wrong individuals are accepted into the wrong schools and programs or jobs and companies. Picking the right people from the start is becoming increasingly important in today's economy and in response to the growing numbers of applicants. Beyond cognitive tests of ability, what other attributes should selectors be considering to know whether an individual has the talent and the capability to perform as well as the mental and psychological drive to succeed? Measuring Human Capabilities: An Agenda for Basic Research on the Assessment of Individual and Group Performance Potential for Military Accession examines promising emerging theoretical, technological, and statistical advances that could provide scientifically valid new approaches and measurement capabilities to assess human capability. This report considers the basic research necessary to maximize the efficiency, accuracy, and effective use of human capability measures in the military's selection and initial occupational assignment process. The research recommendations of Measuring Human Capabilities will identify ways to supplement the Army's enlisted soldier accession system with additional predictors of individual and collective performance. Although the primary audience for this report is the U.S. military, this book will be of interest to researchers of psychometrics, personnel selection and testing, team dynamics, cognitive ability, and measurement methods and technologies. Professionals interested in of the foundational science behind academic testing, job selection, and human resources management will also find this report of interest.
February issue includes Appendix entitled Directory of United States Government periodicals and subscription publications; September issue includes List of depository libraries; June and December issues include semiannual index
"This is an annotated bibliography of research by the U.S. Army Research Institute for the Behavioral and Social Sciences (ARI) on Army recruiting. Most of the research covered in this report was conducted during the period 1980 and 1999. To provide a framework for this work the research summaries are organized around a model of Army recruitment showing the important factors contributing to successful recruiting. In the model, recruiter production is conceptualized as a joint function of recruiter performance and youths' propensity to enlist. Propensity to enlist is linked to advertising effects and several other environmental factors. Recruiters' performance, in turn, is influenced by their personal characteristics, the training and mentoring they receive, and the level of technical and organizational Support provided to them. The ARI research relating to each of these major components in the recruitment model is documented in this report."--DTIC.
