The Validity & Reliability of Employment Testing

by Diane Chinn; Updated September 26, 2017
Exam

Businesses use employment testing to help make decisions about hiring and promoting employees. Employers use a variety of tests, including personality, intelligence, job skills, knowledge, physical ability, situational judgment and language proficiency tests. Title VII of the Civil Rights Act of 1964 prohibits the use of employment tests “designed, intended or used to discriminate because of race, gender, religion, age or national origin.” Test validity and reliability are two metrics used to ensure that employment tests are not discriminatory.

Validity

Validity measures the degree to which a test actually measures what it claims to measure. Validity is determined by research conducted by test publishers, using the guidelines established by the Equal Employment Opportunity Commission and professional organizations such as the Society for Industrial Organizational Psychologists. For example, the Five Factor Model of Personality is considered a good predictor of overall job performance by human resources professionals. Various FFM tests were developed and validated in the early 1990s and underwent a second round of validity testing in 2003 to ensure that they continue to accurately reflect personality factors related to employment.

Types of Validity

The EEOC has approved three types of validity tests. Content validity relates to job-function testing, including mathematics, typing and certification tests for professional organizations, such as registered nurses. Content validity identifies specific behaviors, knowledge and tasks required for a given job. For a test to be valid, the content must be directly linked to the job for which a person is applying. Criterion validity determines whether a test accurately predicts on-the-job performance. The applicant’s test results are compared to subsequent job performance using performance appraisals, productivity and attendance records. Construct validity identifies which dimensions of a test relate to one another. For example, honesty and dependability are not the same but are related as a part of a person’s personality and behavior.

Reliability

Reliability means that a test produces consistent results over time. If a person takes a test today and then takes the same test six months from now, the test would be considered reliable if the results of both tests are similar. For example, if a person scores highly on honesty on the first test, it is expected that the honesty score on the second test would also be high.

A Good Employment Test

A good employment test must be valid and reliable. It must measure what it claims to measure and do so consistently. It is directly related to the job for which a person is being considered, and it measures one or more important job characteristics. It is presented in a format and style appropriate to the test taker’s level of education. A good test also gives decision makers information not obtained through other methods, such as applications, resumes, interviews, reference checks and work samples. Finally, it complies with EEOC requirements and does not discriminate against people based on age, gender, national origin or religion.

About the Author

Diane Chinn is a freelance writer with more than 15 years experience in many areas, including business and technical communications. She has a Bachelor of Arts in psychology from California State University and a Master of Arts in human resources and industrial relations from the University of Minnesota. She is a Six Sigma Green Belt .

Photo Credits

  • gece33/iStock/Getty Images