Psychological Testing in Personnel Selection, Part II: The Refinement of Methods and Standards in Employee Selection

Scroggins, Wesley A., Thomas, Steven L., Morris, Jerry A., Public Personnel Management

During the first hundred years of scientific management and personnel selection, psychological testing became a powerful and institutionalized tool that was broadly applied. Universities and trade schools, managers, engineers, psychologists, and government officials recognized the importance of using scientific methods and tools to manage human capital, the economics of business, and the national defense. Industrial and organizational psychologists emerged as the preeminent players in the development and dissemination of these tools and the evaluation of the tools' real and perceived efficiency. This article reviews the methods, measures, and standards that have emerged to evaluate these tools.

The Issue of Validity

Efforts to develop and create tests to assist managers with personnel selection have ranged from the absurd to the hopeful and into the stage of continued refinement and growing efficiency. The scientific principles underlying the selection tests have also evolved from simple and crude to complex and powerful. Psychologists have known for some time that ineffective tools and recurring fads are not uncommon within the field of psychological testing, particularly when it comes to the manner in which tests are used. (1) Consequently, psychologists have often cautioned consumers to be wary of the variable quality of what is purported to be psychological science. (2)

In the early part of the 20th century, psychology usurped the use of scientific methods to improve performance, selection, motivation, and strategic training from engineering management and had eclipsed scientific management in terms of contributing to the national defense through selection test development. The assessment tools were designed with the hope of facilitating optimal personnel development and utilization of human capital. (3) Using scientific principles and quantitative methods to conserve and develop optimal human resources by forecasting behavioral tendencies over extended periods proved to be a considerable challenge. The challenge was met with variable success, as the opportunities for error were great. Psychologists had captured the testing agenda, but they would struggle mightily to identify instruments that were accepted and widely judged to be useful.

The most important issue in HR selection testing is determining a test's validity. The actual definition of validity can vary depending on the circumstances, the specific tools used, and the application. For most selection purposes, however, a selection test is valid if the characteristic(s) it is measuring is related to the requirements and/or some important aspect of the job the test taker is being evaluated to perform. Test scores only have meaning if the test is valid, and a test is valid if there is a link between the test score and job performance. The degree to which an employment selection test has validity tells the testing entity what it can conclude or predict about someone's job performance from his or her test scores. A test's validity is established for a specific purpose, and it may not be valid for purposes other than those that it has been validated to measure.

Important issues of validity that are discussed in most HR management textbooks are described in Section 1607.5 of the U.S. Equal Employment Opportunity Commission's (EEOC) 1978 Uniform Guidelines on Employee Selection Procedures as criterion-related validity, content validity, and construct validity. (4) Criterion-related validity is the correlation or other statistical relationship between selection test score (the predictor) and job performance (the criterion). If those who score low on a test also perform poorly (and visa versa), the test is said to have high criterion-related validity. Content-related validation is a demonstration that the content of the test reflects important job-related behaviors and measures important job-related knowledge or skills. Construct-related validity is evidence that a test measures the constructs or abstract characteristics that are important to successful performance of the job. (5) For psychological tests used in selection, a test's criterion-related validity is usually the variable of interest to researchers, and it is the validity coefficient--the actual correlation coefficient between a test score and some job performance criterion--that is referred to when validity is discussed in HR literature.

Having evidence of the validity of selection tests is essential for any organization using such tools. Collecting these data is the principal way companies demonstrate that they have met the Uniform Guidelines' requirements should hiring procedures result in adverse impact (i.e., disproportionate hiring outcomes) against protected groups.

Organizations can obtain evidence of criterion-related validity using two basic strategies. One strategy is to conduct a concurrent validation study in which predictor test scores are collected from current job incumbents and the results are correlated with either current measures of job performance or file records of past performance. A second strategy is a predictive validation study in which predictor test scores are obtained from job applicants at the time of hire. After the new employees have had sufficient time to learn the job, job performance measures are obtained, and then the correlation between the predictor test and job performance scores is determined. (6) The problem with an organization conducting its own validation study is that it can be time-consuming (especially if a predictive design is used), it can be costly, and it requires large sample sizes to yield reasonable levels of reliability. Because of these problems, only fairly large organizations and those with fairly frequent hiring have the resources and sample Sizes necessary for conducting validity studies.

A third method for establishing selection test validity is to rely on validity generalization, which is often used when organizations want to use a test for selecting individuals for a specific job and choose to use a test purchased from a commercial vendor. The EEOC's Uniform Guidelines and related standards require that there be evidence of transportability (or generalizability) when purchased tests are used. Transportability refers to the necessity that …

Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Psychological Testing in Personnel Selection, Part II: The Refinement of Methods and Standards in Employee Selection. Contributors: Scroggins, Wesley A. - Author, Thomas, Steven L. - Author, Morris, Jerry A. - Author. Journal title: Public Personnel Management. Volume: 37. Issue: 2 Publication date: Summer 2008. Page number: 185+. © 2009 International Personnel Management Association. COPYRIGHT 2008 Gale Group.

This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.