A Summary of Personnel Selection Methods From: Noe, R. A., Hollenbeck, J. R., Gerhart, B., and Wright, P. M., (2006).

Human Resource Management: Gaining a Competitive Advantage, Fifth Edition. (Table 6.5, p. 245), New York: McGraw-Hill Irwin.
METHOD Interviews RELIABILITY Low when unstructured and when assessing nonobservable traits Low, especially when obtained from letters High test-retest, especially for verifiable information High VALIDITY Low if unstructured and nonbehavioral GENERALIZABILITY Low UTILITY Low, especially because of expense LEGALITY Low because of subjectivity and potential interviewer bias; also, lack of validity makes jobrelatedness low Those writing letters may be concerned with charges of libel May have adverse impact; thus often develop separate scoring keys based on sex or race Often have adverse impact on women and people with disabilities; need to establish job-relatedness Often have adverse impact on race, especially for African Americans, though decreasing over time Low because of cultural and sex differences on most traits, and low job-relatedness in general High because of low adverse impact and high job-relatedness Insufficient history of litigation, but will undergo scrutiny May be challenged on invasionof-privacy grounds

Reference checks Biographical information Physical ability tests Cognitive ability tests Personality inventories

Low because of lack of range in evaluations High criterion-related validity; low in content validity Moderate criterionrelated validity; high content validity for some jobs Moderate criterionrelated validity; content validation inappropriate Low criterion-related validity for most traits; content validation inappropriate High criterion and content validity Insufficient independent evidence High

Low Usually job-specific, but have been successfully developed for many job types Low; pertain only to physically demanding jobs High; predictive for most jobs, although best for complex jobs Low; few traits predictive for many jobs

Low, although not expensive to obtain High; inexpensive way to collect vast amounts of potentially relevant data Moderate for some physical job; may prevent expensive injuries and disability High; low cost and wide application across diverse jobs in companies Low, although inexpensive for jobs where specific traits are relevant High, despite the relatively high cost to develop Insufficient independent evidence Expensive, but may yield high payoffs for health related costs

High

High

Work-sample tests Honesty tests Drug tests

High

Insufficient independent evidence High

Usually job-specific, but have been successfully developed for many job types Insufficient independent evidence High