Faced with an increasingly competitive business environment, many employers are turning to employment testing as a way to improve their workforces. In a recent survey, forty percent of Fortune 100 companies indicated that their employment selection systems included some form of psychological testing. A similar survey by the American Management Association showed that forty-four percent of its responding members used testing to select employees. While cognitive ability tests continue to be the most commonly used form of psychological testing in the workplace, personality tests are being used more and more frequently.
Personality tests are self-report measures of what might be called traits, temperaments, or dispositions. The number of personality measures available is enormous. Some personality measures focus on characterizing individuals within the normal adult range of functioning, while others focus on the identification of psychopathology. Many personality instruments such as the Minnesota Multiphasic Personality Inventory (MMPI) produce measures of a substantial number of personality characteristics; others concentrate on measuring single traits.
Employers use different types of personality tests for different purposes. For example, an insurance company might use a measure of extroversion-introversion to select applicants for a sales job so that their characteristics match successful incumbents in their sales force. Similarly, a police department might use the MMPI or a similar test to screen out applicants for mental instability or psychopathology.
Perhaps the most commonly used personality tests are honesty or integrity tests. They are used predominantly in the retail and financial services industries for low-paying entry level jobs in settings where employees have unsupervised access to money or merchandise. Integrity tests are designed to predict proneness for theft and other forms of counterproductive work behavior in job applicants. Because the number of employers using integrity tests in the workplace is growing rapidly, it is useful to discuss them in some detail.
With estimates of annual economic losses to American business from employee theft ranging from $15 billion to $25 billion per year and as much as 30 percent of all business failures attributable to employee theft, employers are interested in any device or technique that could detect or prevent employee theft. Prior to the passage of the Employee Polygraph Protection Act of 1988, the pre-employment polygraph was the method of choice for screening applicants in many industries. It is estimated that 70 to 80 percent of the four million polygraphs given annually were for pre-employment selection purposes.
The Employee Polygraph Protection Act changed everything. Except in limited circumstances, private employers were prohibited from using a pre-employment polygraph to screen applicants. With the polygraph no longer available, employers turned to paper and pencil integrity tests to perform the screening function. By 1990, it was estimated that 6,000 organizations administered up to 5 million integrity tests annually.
Although there is currently no federal regulation of integrity testing, a few states have restricted their use under their antipolygraph statutes. For example, the Rhode Island statute does not ban integrity tests, but it requires that such tests not constitute the "primary basis" of an employment decision. Massachusetts forbids the use of any written technique that provides a diagnostic opinion of honesty, a proscription so broad that it may prohibit not only integrity tests but also the use of application blanks, reference checks, structured interviews, and the validity scales of most psychological tests as well.
To understand the benefits and risks associated with integrity testing, it is useful to briefly review the different types of tests. There are three basic types of integrity tests: overt integrity tests, personality-oriented measures, and clinical measures.
Overt integrity tests were specifically designed to predict the predisposition of job applicants to engage in on-the-job theft and other counterproductive job behavior. They measure attitudes related to one or more of the following psychological constructs: tolerance of others who steal, projection about the extent of theft by others, acceptance of rationalizations for theft, interthief loyalty, antisocial beliefs and behaviors, and admissions of theft-related activities. Commonly used tests of this type include the London House PSI, the Employee Attitude Inventory EAI, the Reid Report, the Stanton Survey, and the Trustworthiness Attitude Survey.
These tests typically consists of two sections. The first is a measure of theft attitudes, and includes questions about beliefs concerning the frequency and extent of theft, punitiveness toward theft, ruminations about theft, and assessments of one's own honesty. The second involves requests for admissions regarding theft and other wrongdoing. Applicants are asked to describe the frequency and amount of theft and other illegal or counterproductive activity. The test items that make up this type of instrument are clearly assessing job-related content. (e.g., "Will everyone steal at work if the conditions are right?"; "Do you believe you are too honest to steal at work?"; "Do you think it is humanly possible for the average person to be completely honest on the job?"; etc.).
There has been a great deal of validity research showing that integrity test scores can predict theft behavior. Overt integrity tests have been found to predict the following theft criteria: (a) supervisors' ratings of employees' dishonesty, (b) applicants who are likely to get caught stealing once hired, (c) applicants that have a criminal history, (d) applicants who are likely to make theft admissions in an anonymous testing situation, and (e) theft admissions made in a pre-employment polygraph.
Longitudinal studies have also demonstrated the real world impact of using integrity tests to select employees. In one study, a group of convenience stores using the PSI integrity test to select employees experienced a 50% reduction in inventory shrinkage due to theft over an 18 month period. This impact-on-losses study was replicated in a home improvement center chain. Several other studies have shown that both applicants and employees from high-theft stores scored more poorly on integrity tests than did applicants and employees from low-theft stores. In a quasi-experimental study, employees working in supermarkets that used integrity tests reported significantly less theft by their coworkers than employees working in a matched group of supermarkets that did not screen their employees using integrity tests.
Personality-oriented measures are typically developed by psychologists, and are closely linked to normal-range personality assessment devices, such as the California Personality Inventory. Considerably broader in focus, they are not explicitly aimed at theft or theft-related behaviors. They aim to predict a broad range of counterproductive work behaviors using composite measures of personality dimensions, such as reliability, conscientiousness, adjustment, trustworthiness, and sociability. Commonly used tests of this sort are the Personal Outlook Inventory, Personnel Reaction Bank, the PDI Employment Inventory, and the Reliability Scale of the Hogan Personality Inventory. Many of the test items that make up instruments of this type are not clearly job-related. (e.g., "My mate/spouse does not understand my sexual needs."; "I have had a child out of wedlock"; "I have thought about losing my mind."; "When you were growing up, you had a lot of disagreements with your parents." ).
Although personality-oriented measures have not been validated explicitly against measures of employee theft like overt integrity tests, they have been shown to predict theft and other such counterproductive behaviors as absenteeism, tardiness, and disciplinary problems.
Clinical measures were developed for other uses, such as identifying psychopathology. However, these measures can be used by a trained psychologist to form a diagnostic opinion about an applicant's integrity. The MMPI is good example. It consists of 550 questions covering, among other things, health, psychosomatic symptoms, family and marital issues, phobias and delusions, and sexual, religious, political and social attitudes. As with personality-oriented measures, many of the items that make up instruments of this type are not clearly job-related. (e.g., "My sex life is satisfactory."; "I am very seldom troubled by constipation."; "Once in a while I think of things too bad to talk about."; "People should try to understand their dreams and be guided by or take warning of them.")
Title VII. Under Title VII of the Civil Rights Act of 1964, it is unlawful for an employer to fail or refuse to hire any individual, or otherwise discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of race, color, religion, sex, or national origin. With respect to employment testing, Section 703(h) of the Act provides that "[n]otwithstanding any other provision of this subchapter, it shall not be an unlawful practice for an employer . . . to give and to act upon the results of any professionally developed ability test provided that such test, its administration or action upon the results is not designed, intended or used to discriminate because of race, color, religion, sex or national origin."
Clearly, Title VII does not prohibit employers from using personality or integrity tests in the workplace. However, the statute is implicated when: (1) an employer uses tests to intentionally discriminate against minorities or (2) the tests have an adverse impact on minorities and are not job-related for the position in question and consistent with business necessity.
Disparate Treatment. No cases have found that an employer's use of personality testing in the workplace resulted in the disparate treatment of protected groups. However, there could be such a cause of action if the tests were used intentionally to exclude members of protected groups. This would usually involve evidence of comparative proof that members of a protected group were treated differently than majority group members.
Disparate Impact. The 1991 Civil Rights Act codified the disparate impact theory articulated by the U.S. Supreme Court in Griggs v. Duke Power Company and other cases predating Ward's Cove v. Antonio. To prove a prima facie case of disparate impact, the plaintiff must show that a specific employment practice has a significant adverse impact upon a protected group of which the plaintiff is a member. While the Supreme Court has not provided definitive guidance on the size of the disparity necessary to satisfy this burden, most courts have been satisfied by a showing that the disparity is sufficiently large that it is unlikely to have occurred by chance. Tests of statistical significance are generally used to do this. However, some courts have preferred the 80 percent rule contained it the EEOC's Uniform Guidelines on Employee Selection Procedures.2 Under this rule, adverse impact is established if the selection rate for any race, sex, or ethnic group is less than four-fifths (or 80 percent) of the rate for the group with the highest selection rate.
If a plaintiff establishes adverse impact, the burden shifts to the employer to prove that the challenged practice is job-related for the position in question and consistent with business necessity. The Uniform Guidelines recognize three methods by which an employer can show job-relatedness: content validity, criterion validity, and construct validity. Content validity is used when the test closely approximates the job, as in a typing exam for a typist position. This form of validation is often used for aptitude testing, but is not generally applicable to personality testing because such tests do not measure skills or job knowledge. Criterion validity compares success on the test with success on critical or important job duties using correlational analysis. It is the preferred method of validation under the Uniform Guidelines, but is not always technically feasible. Construct validity establishes a relationship between some aspect of satisfactory job performance and a specific trait, such as the human relations skills required to perform effectively as a customer service representative. This form of validation is the most appropriate for personality testing, because it focuses on the link between a particular trait and predicted job performance.
Even if the employer can satisfy this burden, it cannot escape liability if the plaintiff shows that there exists an alternative practice, which would have less disparate impact, and would equally serve the employer's legitimate interests and the employer refuses to adopt it. In determining whether the proposed practice equally serves the employer's legitimate interests, the relative costs and benefits of the selection practices at issue may be considered.
There are have been few disparate impact cases involving personality tests because such tests generally do not have an adverse impact on any protected group. Integrity tests may have the best record of any selection technique in demonstrating freedom from adverse impact. Neither the EEOC nor any state administrative agency charged with administering fair employment laws has ever determined that an integrity test demonstrated adverse impact. Thus, no complainant has ever successfully proved a prima facie case against their use. It is also noteworthy that the record shows few challenges--probably less then 50--for the millions of tests that have been administered.
Even if a complainant could demonstrate adverse impact, an employer would usually be able to justify the use of testing. Leading integrity test publishers have extensive statistical evidence demonstrating the validity of their instruments at predicting theft and job-relevant counterproductive behavior that would satisfy the employer's burden. Similar data may exist for other tests. However, in some cases an employer may need to undertake their own validity studies.
Section 106 of the Civil Rights Act of 1991, makes it an unlawful practice for an employer:
The use of personality tests that have race-specific or gender-specific norms could violate Section 106. Although very few personality measures offer norm scoring based on race or ethnicity, many do employ gender-specific norms. Employers should probably avoid tests of this sort in favor of integrity tests with unitary norms.
The ADA provides that an employer "shall not conduct a medical examination or make inquiries as to whether such applicant is an individual with a disability or as to the nature or severity of such disability." 42 U.S.C. ' 12112(d)(2)(A). According to EEOC guidance, employers may give psychological examinations to job applicants so long as the examination is not medical. Psychological examinations are medical if they provide evidence that would lead to identifying a mental disorder or impairment. The EEOC provides the following example:
A psychological test is designed to reveal mental illness, but a particular employer says it does not give the test to disclose mental illness (for example, the employer says it uses the test to disclose just tastes and habits). But the test also is interpreted by a psychologist, and is routinely used in a clinical setting to provide evidence that would lead to a diagnosis of a mental disorder or impairment (for example, whether an applicant has paranoid tendencies, or is depressed). Under these facts, this test is a medical examination.
The use of clinically oriented personality measures designed primarily to diagnose psychopathology would probably violate the ADA's prohibition on medical examinations. Employers should avoid the MMPI and other similar measures for selection purposes.
EEOC guidance recognizes that tests designed and used to measure things such as honesty, tastes, and habits are not medical tests. Thus, most integrity tests would not be considered medical examinations under the ADA.
An issue of growing legal concern for employers using personality testing is privacy. The concept of a right to employment privacy evolved from the broader concept of the right to privacy. Invasion-of-privacy claims can be predicated on the federal constitution, state constitutions and statutes, and common law.
Although the U.S. Constitution has no applicability to private employers, public employers must comply with its requirements, and some states such as California, have applied such privacy standards to private employers. The lower courts have recognized the constitutional right to privacy protects public employees. Thus, test questions administered to public employees must not be unreasonably intrusive and must be job-related.
A federal district court in New Jersey upheld the personality testing of firefighter applicants in McKenna v. Fargo. The court found that although the tests burdened the applicant's right to privacy, the city's interest in screening out applicants who could not withstand the pressures of the job was sufficient to justify the intrusion. McKenna establishes that personality test questions that inquire into an applicant's sexual, religious, and social attitudes implicate the constitutional right of privacy when there is state action. It also recognizes that government can justify the intrusion if it has a compelling interest such as the public safety function of firefighters.
Unlike the federal constitution, state constitutions provide a possible avenue for job applicants to challenge the employment practices of private employers. Ten states have constitutions that provide right to privacy protections, many of which are broader than the federal constitution. Probably the best known among them is California.
A California Court of Appeals found that certain portions of a personality test administered to job applicants for security guard positions violated the state constitutional right to privacy and state antidiscrimination laws. Saroka v. Dayton Hudson. The plaintiff had applied for a security guard position with Target Stores, and as a condition of employment was required to complete an inventory consisting of the MMPI and the California Psychological Inventory (CPA). The two tests are widely used to assess psychological balance and contain items dealing with highly personal topics such as religion, bodily functions, sexual behavior, and political beliefs.
The plaintiff alleged that the test questions probed into his private thoughts and innermost feelings and were not job-related. The court agreed, finding that certain questions invaded the applicants' privacy because they asked about sexual and religious preferences. To justify this invasion of privacy, Target needed to demonstrate a compelling interest and show that the test served a job-related purpose. The court acknowledged that Target had an interest in employing emotionally stable persons as store security officers. However, the court found that Target did not show how information about an applicant's religious beliefs or sexual orientation would have any bearing on their emotional stability. Hence, the questions were not deemed to be job-related.
Unlike personality-oriented and clinical measures of personality, overt integrity tests have generally not been challenged on privacy grounds. This is because these tests do not typically inquire into sensitive areas and are more clearly job-related. Moreover, leading test publishers continuously conduct follow-up research to ensure that their test items are not perceived to be invasive.
Although there are legal reasons for exercising caution in using personality testing in employment, there are also countervailing trends that argue strongly for the most extensive use of such tests. With the tort of negligent hiring now recognized in a majority of the states, employers have been forced to defend a growing number of suits seeking redress for crimes committed by employees, usually thefts or assaults that victimize customers or coworkers. In these cases, liability is predicated on the employer negligently placing an applicant with propensities, which should have been discovered by reasonable investigation, into an employment position where it was foreseeable that the hired individual posed a threat of injury to others.
Psychological testing becomes relevant because the plaintiffs generally allege that the employer should have been aware of the characteristics of the employee causing the harm. Testing provides probative evidence that the employer met its duty to investigate reasonably an applicant's fitness. Companies that adopt pre-employment integrity tests to screen job applicants can reduce their exposure to negligent hiring claims. This may be particularly important in states where it is difficult to obtain criminal records or other background information on applicants.
Although there are real benefits to using personality or integrity tests in employment, employers should carefully weigh the risks and benefits. It may be helpful to consult with an industrial-organizational psychologist and an employment attorney. Industrial-organizational psychologists can provide critical information related to test validity as well as utility analyses that show how testing would impact the company's profitability and efficiency. An experienced employment lawyer can alert employers to potential problems and refine their employment selection policies and procedures to minimize their exposure to liability.
1David J. Shaffer is a Partner and the head of the Employment law group [link to practice page]of the Washington office of Thelen Reid & Priest LLP. Ronald A. Schmidt is an attorney with the firm. Both work closely with experts in the fields of statistics, labor economics, and industrial-organizational psychology in defending employment and civil rights class action suits throughout the country.
2 In the opinion of the authors, the 80% rule is illegal in light of Lutheran Church-Missouri Synod v. FCC, 141 F.3d 344 (D.C. Cir. 1998).