By John K. Allen, Ed.D.
"Selecting employees is like trading horses. If you've never been fooled, you've only done it once." A veteran CEO
All organizations have to be concerned with employee performance. The benefits of selecting employees who can learn quickly and perform to expectations are obvious. Selection is without doubt one of the most powerful tools available to achieve these productivity increases. In spite of claims to the contrary there is no foolproof way to select employees. There is no magical test, interview technique, background check, or experience requirement to insure that every pick will be a winner. There is only a choice of scientific methods and procedures designed to reduce the risk of a bad hire. Psychological testing and assessment are currently the most effective methods available to implement selection.
Employment testing increased dramatically following World War II when testing procedures developed for the military were adapted for business and industrial use. At the same time, the Sputnik- inspired “Race for Space” stimulated the development of many educational tests. Most of us have taken the scholastic aptitude, achievement, and interest tests that were the result of this national research effort.
Psychological tests are systematic procedures for comparing individual behavior, often for the purpose of prediction. Most personnel tests can be classified as psychological tests. There are four basic types of such tests that are commonly used in employment selection: ability tests, aptitude tests, achievement tests, and personality tests.
Ability Tests. Ability tests measure overall skill in learning and solving problems. Ability is related to what was formerly termed “intelligence.” Most people think of intelligence as a genetic or inherited potential to solve problems but employment testing is more concerned with demonstrated performance
Aptitude Tests. Aptitude tests are like ability tests but focused on more specific traits. Mechanical aptitude and verbal reasoning are examples of aptitudes commonly measured in employee selection.
Achievement Tests. Achievement tests usually measure a specific skill or body of knowledge. Examples include spelling, reading, drafting, word processing, or welding. Theoretically, achievement tests measure previously acquired skills and abilities in contrast to ability and aptitude tests that measure the potential to learn. However, the distinctions between ability, aptitude, and achievement tests in the practice of employment selection are often blurred. A special kind of achievement tests is a performance test, in which the applicant actually performs a part of several parts of the job for which he or she is under considerations. The performance is then rated or graded according to pre-established criteria. A work sample test asks experienced applicants to perform part of the job, repairing a piece of equipment, for example. Simulations of job performance are often used in the selection of supervisors or managers. Job situations such as leading a group discussion,
Personality Tests. Personality tests measure enduring traits or the predisposition of an individual to behave in a certain way regardless of external circumstances. Personality tests are as often used in professional development assessments as in employment testing. Typically, tests measuring traits such as self-confidence, independence, conscientiousness, or people orientation are used to understand an individual’s management or supervisory style. Personality tests are also frequently used as part of screening procedures for law enforcement personnel and personnel who have unrestricted access to nuclear facilities.
Personality tests are not as strong a predictor of job performance as ability tests, work samples, or highly structured interviews but may add to the predictive power of an assessment battery. A number of personality tests are commercially available that have not been professionally developed. Test users should be sure to
The effectiveness of a selection procedure, whether a test, interview, performance simulation, or reference check, is termed its validity and can be expressed statistically by a validity coefficient ranging from -1 to +1. The higher the validity coefficient the better the predictor. If there is no relationship between a test and job performance the validity coefficient is 0.0 and using such a test would be like throwing darts at names on a dartboard to select employees.
In practice it is unusual to find validity coefficients of greater than +. 5 because of difficulties in measuring both job performance and differences in job applicants. Validity coefficients of .3, .4, or .5 are useful predictors of job performance.
The typical selection interview as conducted in industry has an average validity of .38 compared with random hiring validity flip of a coin) of 0.0. The average validity of general cognitive ability testing is .51. The validity of a carelessly conducted interview would be much lower than .38. Thus we can state that ability tests scores predict job performance ratings roughly 30% more effectively than the unstructured interview.
Several researchers have used a sophisticated statistical technique termed
“meta-analysis” to quantitatively collect and compare the validity
of different selection techniques from hundreds of research studies. The following
table presents some of the findings from a meta-analytic study of the effectiveness
of common selection techniques (Schmidt and Hunter 1998).
Selection Technique |
Validity |
Ability Tests |
.51 |
Employment Interviews (structured) |
.51 |
Employment Interviews (unstructured) |
.38 |
Assessment Centers |
.37 |
Conscientiousness Tests |
.31 |
Reference Checks |
.26 |
Job Experience (years) |
.18 |
Years of Education |
.10 |
Graphology |
.02 |
We can conclude from a comparison of these validity coefficients that many widely used selection techniques such as reference checks or years of education are much less powerful predictors of future
Getting Started in Testing. Tests are usually part of a selection process that may include recruiting, screening, testing and assessment, interviews, reference checks, and the employment decision. This first step in using psychological testing as part of a selection system is to conduct a job analysis. Much more detailed and scientific than the typical job description, job analysis identifies the key knowledge, skills, abilities, and organizational factors which will guide the selection approach.
Summary. There are large differences in employee job performance and success in training. A good selection technique can yield large increases in employee and organizational productivity. Psychological testing, particularly ability testing, is the best-known predictor of employee job performance and success in training. The use of psychological testing and assessment in selecting employees makes sense both on business and scientific grounds. As experts in employment testing and assessment, West Falmouth Associates can help you analyze job requirements and
References
Schmidt, F.L., & Hunter, J.E. (1998). The validity and utility of selection methods in personnel psychology: Practical and theoretical implications of 85 years of research findings. Psychological Bulletin, 124, 262-274.
© 2007 West Falmouth Associates. All rights reserved.