Integrity Testing

Paper-and-pencil tests designed to measure integrity, honesty, dependability, and related constructs have been in existence since at least the 1950s and have long been used in the retail sales, banking, and food service industries. Following the 1988 Employee Polygraph Protection Act, a federal law that restricted the use of the polygraph (i.e., the so-called lie detector), integrity testing grew substantially and spread into a wider range of industries and applications.

Although the individual tests differ in a number of specifics, there are a number of features common to virtually all integrity tests. In particular, integrity tests usually include items that refer to one or more of the following areas: (a) direct admissions of illegal or questionable activities, (b) opinions regarding illegal or questionable behavior, (c) general personality traits and thought patterns thought to be related to dishonesty (e.g., the tendency to constantly think about illegal activities), and (d) reactions to hypothetical situations that may or may not feature dishonest behavior.

Overt Versus Personality-Based Tests

A distinction is usually drawn between tests that inquire directly about integrity, asking for admissions of past theft, or asking about the degree to which the examinee approves of dishonest behaviors (often labeled overt tests), and tests that indirectly infer integrity on the basis of responses to questions that are not obviously integrity-related (often labeled personality-based tests. However, the distinction between overt and personality-based tests is not always a simple one. Many overt tests include items, scales, and so forth that are not obviously related to honesty, and many personality-based tests contain items that might alert the respondent to the true purpose of the test.

Overt tests usually include direct or indirect measures of perceptions of norms relating to honesty. The rationale for this type of measure is that individuals who believe dishonesty, theft, and so on are common and widely accepted are more likely to engage in these behaviors themselves. Even questions that ask for admissions of wrongdoing are probably best understood as measures of perceived norms. Individuals who admit to theft, dishonesty, and rule breaking when responding to an integrity test often believe that the behaviors they are admitting to are in fact widespread and at least implicitly accepted.

Personality-based tests often measure constructs such as thrill seeking, socialization, and resistance to authority. More recent tests have incorporated broad dimensions of personality that have been empirically linked to theft and counterproductive behavior, notably conscientiousness, agreeableness, and emotional stability.

Integrity tests (both overt and personality-based) appear to measure a range of constructs, and there is disagreement over the assumption that a general honesty or integrity factor underlies these tests. Various tests use unique scoring systems, and it is unwise to assume that tests are interchangeable.

Validity of Integrity Tests

Early reviews of integrity test validity came to discouraging conclusions, but since the late 1980s, both narrative reviews and meta-analyses of integrity test validity have suggested that integrity tests are useful for a variety of purposes.

In discussing validity evidence, it is important to identify the specific criteria used in different studies. Some studies have validated integrity tests against measures of counterproductive behavior, whereas others have validated these tests against measures of general job performance. For example, scores on integrity tests show average correlations of .21 and .33 with measures of job performance and counterproductivity, respectively (correcting for unreliability and a variety of statistical artifacts, the estimated population correlations are .34 and .47, respectively). These two criteria are clearly not independent; employees who engage in a wide variety of counterproductive behavior are unlikely to be good performers. Nevertheless, there are important differences between the two criteria and, more important, differences in the criterion-related validity of integrity tests for predicting the two.

Construct Validity

There are several challenges in evaluating the construct validity of integrity tests. It is exceedingly difficult to define honesty, integrity, or whatever attributes these tests are designed to measure. Different tests seem to focus on very different attitudes, beliefs, or behaviors. For example, there are a number of definitions of employee theft. Researchers often distinguish between trivial and nontrivial theft; conclusions about the extent of theft depended largely on whether taking articles of little value (e.g., pencils, paper, supplies) was included in one’s definition of theft. Gold-bricking, taking long lunch breaks, using company time to carry out personal business, and similar activities are sometimes labeled time theft.

A related issue has been a source of controversy for nearly a century—the question of whether honesty is a distinct trait. Most researchers believe that situa-tional factors have a strong impact on the tendency to engage in honest versus dishonest behaviors and that labeling a person as honest or dishonest is a serious oversimplification.

Informed consent is a potentially serious issue in integrity testing. Integrity test publishers often advise against informing examinees of their test scores. This implies that if an individual is denied employment on the basis of a score on an integrity test, he or she should not be so informed. The Standards for Educational and Psychological Testing and the Ethical Principles of Psychologists make it clear that psychologists involved in integrity testing are obliged to inform examinees of the risks and consequences of taking the test versus refusing to take the test, the purpose and nature of the test, and the way in which test scores will be used.

Finally, there are serious concerns over the way in which integrity tests are scored and in which scores are reported. It is common to use some sort of dichotomous scoring (e.g., pass-fail) in integrity testing. More sophisticated tests sometimes report test scores in terms of a small number of zones: high danger, moderate danger, average danger, or low danger of theft, substance abuse, and so on. Test scores that are reported on a pass-fail basis are inherently suspect because they blur potentially meaningful differences between individuals in each of the two categories (e.g., it is unlikely that all individuals who fail present the same risks).

Overall Assessment of Integrity Tests

Integrity testing was the focus of a great deal of controversy and debate in the late 1980s, when these tests first emerged into the mainstream. The American Psychological Association appointed a task force to examine integrity testing; their general conclusions provide a useful framework for evaluating integrity tests. In their 1991 report, they noted that integrity tests shared many of the strengths and weaknesses of other psychological tests and that well-developed tests showed acceptable evidence of reliability and validity to justify their use. There are still valid concerns about the use of bad tests and unwise testing practices when assessing integrity, but these concerns are not categorically different from concerns that can be raised in contexts where most psychological tests are used. Integrity tests are not broadly accepted as potentially useful assessment tools.

Integrity testing was once regarded as a field at the fringe of respectability. Integrity test publishers were regarded as secretive and (ironically) unreliable. In the last 20 years, there has been considerable progress in integrity testing and a very healthy interchange between test publishers and researchers interested in understanding the causes and consequences of integrity in the workplace. It now seems clear that the controversies over integrity testing in the 1980s and 1990s have proved beneficial to both test developers, who have improved both their products and the process by which tests are developed and researched, and the users of integrity tests, who have increasing access to empirically validated measures that provide information about the dependability of job applicants.

References:

  1. APA Task Force. (1991). Questionnaires used in the prediction of trustworthiness in pre-employment selection decisions: An APA Task Force Report. Washington, DC: American Psychological Association.
  2. Murphy, K. (1993). Honesty in the workplace. Monterey, CA: Brooks/Cole.
  3. Ones, D. S., Viswesvaran, C., & Schmidt, F. L. (1993). Comprehensive meta-analysis of integrity test validities: Findings and implications for personnel selection and theories of job performance. Journal of Applied Psychology, 78, 679-703.
  4. 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.
  5. Wanek, J. E., Sackett, P. R., & Ones, D. S. (2003). Towards an understanding of integrity test similarities and differences: An item-level analysis of seven tests. Personnel Psychology, 56, 873-894.

See also: