Employment selection involves finding the best person for any particular job.

The pay off for doing this correctly is huge.  If you rank order the employees in any organization from least productive to most productive, the top 25% of the employees will contribute 4 or 5 times as much to profitability as the bottom 25%.  The goal is to hire more top performers and fewer bottom performers.

Psychologists have been doing selection research for over 100 years, and know a lot about how to select high quality employees.  The bottom line:  there is one right way and many wrong ways to do selection.

Doing selection correctly means doing it in ways that are defensible from a scientific and a legal perspective.

A scientifically defensible selection process is one for which there is evidence that people who do better in the selection process also do better on the job.

A legally defensible selection process is one in which all applicants, regardless of age, gender, or ethnicity, have an equal chance of passing.  That is, a legally defensible selection process is one that doesn’t discriminate against applicants on the basis of age, gender, or ethnicity.



Five selection methods are commonly used in business today:

  1. Interviews
  2. Cognitive Ability Measures
  3. Integrity Tests
  4. Personality Inventories
  5. Assessment Centers.

Interviews.  Research shows that interviews are the most frequently used selection process in modern business.  Research also shows that unstructured interviews are almost completely indefensible on scientific or legal grounds.  Structured interviews, which use a standard script tailored to the demands of the job in question, and a standard way to interpret answers, are in fact defensible on scientific and legal grounds.  The problem with structured interviews concerns “interview drift”—interviewers get bored with the standard procedure and drift back into unstructured interviews and all the problems that they entail.

Cognitive Ability Measures.  Most psychologists believe that measures of cognitive ability predict performance in any job and are, therefore, scientifically defensible.  Research shows that white job applicants get somewhat higher scores than black job applicants on standard measures of cognitive ability.  This difference is, in fact, discrimination.  However, the fact that measures of cognitive ability typically predict job performance is used to offset the problem of discrimination.  More importantly, however, cognitive ability measures tell us whether a person can do a job, but they don’t tell us whether a person will do a job.

Integrity Tests.  Integrity tests are typically used to screen entry-level employees for honesty, dependability, and willingness to follow rules.  These tests have two attractive features.  First, they are scientifically defensible—they predict job performance.  And second, they are legally defensible—they don’t discriminate against minority job applicants.  The problem concerns their narrow focus—they only evaluate a person’s willingness to follow rules.  They have nothing to say about potential for customer service work, for working as part of a team, for exercising leadership or taking initiative, or for thinking creatively.

Personality Inventories.  Well-developed inventories of normal personality such as the California Psychological Inventory (CPI), the 16 PF, and the Hogan Personality Inventory (HPI) have a number of attractive features.  They are scientifically defensible—they predict job performance about as well as measures of cognitive ability.  They are legally defensible—they don’t discriminate against minorities, women, or older workers.  They predict a wide range of performance outcomes from integrity to sales ability.  The draw back to tests of this type is that perhaps 2500 of them are commercially available, and most of them lack scientific credibility—they don’t predict job performance.  Here it is a case of caveat emptor.

Assessment Centers.  Assessment centers are the gold standard of employment selection.  They provide a comprehensive appraisal of a person’s talents, skills, and abilities and can be used to select talented employees at any level.  Their major drawback is their high costs and the time it takes to complete them.

Finally, it is worth noting that Joyce and Robert Hogan pioneered the use of measures of normal personality for employee selection beginning with research papers published in the early 1970’s.  They did this at a time when studying personality and real world performance would put a scholar’s career at risk.