The use of personality tests by applied psychologists in employee selection is well documented (e.g., Ones, Dilchert, Viswesvaran, & Judge, Reference Ones, Dilchert, Viswesvaran and Judge2007). In spite of this rich history and the fact that there have been “limited legal challenges in the past [as relate to the ADA]” (Melson-Silimon, Harris, Shoenfelt, Miller, & Carter, Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019, p. 119), the focal article presents a case for an overly cautious, regressive, and, in the end, less effective framework to assess potential employees. We posit that avoiding personality tests that assess constructs tangentially related to personality disorders (PDs) and PID-5 dimensions is not only limiting within applied settings but that it is counter to the more expansive and sophisticated reconceptualization process underway in the larger field of psychology to better capture the nuances of individual personality. Furthermore, the current body of case law is too limited in quantity and applicability to necessitate such a restrictive approach.
As noted in the focal article, with each permutation of the Diagnostic and Statistical Manual of Mental Disorders (DSM), the field of psychology displays an improved understanding of the human condition, increasingly employing models that offer room for more varied and nuanced aspects of personality. Most notably, there has been increased emphasis on recognizing that personality characteristics may fall along a continuum (from normal to pathological). This move has been afoot for some time (i.e., circa 2013; DSM-5) with regard to anxiety, depression, and their associated disorders. However, only recently has a similar scheme been offered for traits commonly associated with PDs, which have long been viewed as the more pervasive and intractable of disorders. However, the fact that such models have been employed for some time regarding anxiety and depression provides guidance as to how employment practices might be affected by the newer models of personality disorders.
The focal article suggests that because the modern (i.e., normal–pathological) continuum view of personality traits has now extended to traits associated with PDs, we should now operate with a sense of (assessment) trepidation. However, this downplays the fact that there have been limited negative ramifications like the ones foreshadowed by the focal article related to anxiety and depression. Not to mention that, in comparison to anxiety and depression, personality disorders are exceedingly rare, ranging in prevalence rates from a high of 7.9% (obsessive compulsive disorder) to a low of .06% (dependent personality disorder; APA, 2013). Thus, it seems highly unlikely that the use of assessment tools with any individual would unintentionally tap into pathological personality traits noted in the alternative model of PD (AMPD) of DSM-5. At a minimum, it does not seem effective to choose a less comprehensive or less valid personality test out of fear that the variables of interest (KSAOs) may “approximate” PDs. Further, although the focal article attempts to make an argument that a number of FFMs have been shown via meta-analyses to have some relation to PDs, the point is overstated, as the estimates of association range from weak (.02) to moderate (.47), with the majority being less than .20. The most frequently used personality factor, conscientiousness, had a mean correlation with personality disorders of only –.08 (Saulsman & Page, Reference Saulsman and Page2004).
Furthermore, a personality score alone is insufficient to identify and formally diagnose personality disorders. DSM-5 diagnostic criteria require evidence of impairments in personality functioning in self and interpersonal domains along with pathological trait scores. Evidence of impairments is most often from clinical interviews and other-reports (e.g., family). Extreme test scores, in isolation, are not sufficient to conclude a diagnosable condition such as PD, and no clinical psychologist (diagnostic specialist) would make a diagnosis based on one score from one measure. In fact, Dr. Koransky, the clinical psychologist expert witness in Karraker v. Rent-A-Center (2005), argued that the personality scale used by Rent-A-Center “does not diagnose or detect any psychological disorders” (p. 837). Instead, “an elevated score on the personality scale [was] one of several [emphasis added] symptoms which may contribute [emphasis added]” to a diagnosis of paranoid personality disorder (p. 837). This argument was understood by the lower court in Karraker (2005), which ruled that the nonmedical use of the MMPI was appropriate; and, interestingly, to date, there have been minimal affirmed legal challenges under the ADA that are directly based on the utilization of a personality test like the MMPI.
On appeal, the judgment against Rent-A-Center resulted not from the potential for a test score to identify individuals with PDs but rather from poor justification by Rent-A-Center concerning the (nonmedical) use of the MMPI and insufficient validity evidence linking the test with job competencies. The 7th Circuit specified that while psychological tests “‘designed to identify a mental disorder or impairment’ qualify as medical examinations,” “psychological tests ‘that measure personality traits such as honesty, preferences, and habits’”—which can logically be tied to job requirements—“do not” (p. 835; citing EEOC Enforcement Guidance, 1995). Recall that some 10 years earlier in Thompson v. Borg-Warner (1996), the defendant successfully justified its use of personality testing as it related to applicant character and personality traits and their relation to job duties. Thus, case law suggests that personality tests are acceptable under the ADA when they are clearly linked to KSAOs—in other words, we argue that the requirements for establishing the validity of selection tests have not changed.
We also think it important to clarify the applicability of the case law discussed in the focal article. The Thompson (1996) case is a district court (lowest level federal court) of California case that ultimately settled. Thus, this case has minimal persuasive authority, at best. Barnes (1997) is an 11th Circuit (mid-level federal appellate court) case. The 11th Circuit appellate court binds only federal district courts in Alabama, Florida, and Georgia. Finally, Karraker (2005) is a 7th Circuit (mid-level federal appellate court) case. The 7th Circuit appellate court binds only federal district courts in Illinois, Indiana, and Wisconsin. The Soroka (1991) case was brought as a constitutional privacy law claim under the state court system—a court system entirely distinct from the federal court system that analyzes ADA claims (see Figure 1). Thus far, the cases involving personality test challenges under the ADA are too few and too limited in their binding applicability to draw firm conclusions on what the law states in this area. Further, the vast majority (approximately 95%) of cases settle before trial (Hirby, n.d.). This includes several prominent cases involving ADA claims based on personality test items, such as the 2015 litigation against Target, where the retail giant was forced to shell out $2.8 million per its settlement agreement with the EEOC (Zillman, Reference Zillman2015). Because the terms of the settlement and the factual details of the case are typically kept confidential when parties settle, the problematic test items, inventories, and circumstances of the employee selection process are never released. Thus, we should be cautious about suggesting a blanket limitation on the use of personality test items in employee selection.
Figure 1.Structure of the United States Federal and State court systems.
Furthermore, the EEOC guidance discussed in the focal article should be taken as exactly that—guidance. Although the EEOC is authorized to issue interpretive and procedural guidance to advise employers on how to comply with the laws it enforces, the EEOC cannot issue regulations with the force of law (Wern, Reference Wern1999). There is also some evidence that the Supreme Court defers to the EEOC less frequently than other federal agencies (only 54% of the time, compared to 72% of the time for other federal agencies; Wern, Reference Wern1999). Therefore, although the EEOC guidance cited in the focal article should not be ignored, it should not be taken as a source of binding legal precedent either.
The focal article also fails to adequately appreciate the larger and more significant paradigm shift that has taken place within the field of clinical psychology at large from a destigmatization standpoint. It is indeed meaningful that this convergence of the normal and pathological models has finally extended to PDs, which have long been viewed as largely intractable and fully (over)pathologized in a categorical sense, with little space for nuanced aspects of associated personality traits. We would argue that this actually has the potential to normalize and destigmatize a person’s mode of being (e.g., higher scores on a trait) that was once universally and solely viewed as pathological. This is actually a “win” for people with disabilities, as it has the potential to open doors once viewed as closed because of the perception of pathology. With the expansion of traits to a singular continuum, the very notion of “normal” and “disordered” is shifting and so is the meaning of test scores and their resultant inferences. Although there are certainly some personality characteristics that presumably represent a poor KSAO fit with some occupations, in truth, these have always existed, and will always loom over any selection processes.
In summary, it seems premature to leap to the conclusion that there is “a seemingly inevitable collision between the practice of personality testing for employment purposes and the scientific understanding of personality models” (Melson-Silimon et al., Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019, p. 120). Few ADA challenges have been brought concerning personality models’ correlations with anxiety and depression—which are far more prevalent in the population than the personality disorders discussed here. Further, FFM personality tests have only limited correlations with, and are insufficient for, the diagnosis of personality disorders, and the applicability of case law concerning personality tests and PDs is limited and far from resolved. As always, careful job analysis and validation of selection procedures should be undertaken. As long as personality is validity linked with job-related KSAOs, we see little cause for a change of course at this time.
You have
Access
The use of personality tests by applied psychologists in employee selection is well documented (e.g., Ones, Dilchert, Viswesvaran, & Judge, Reference Ones, Dilchert, Viswesvaran and Judge2007). In spite of this rich history and the fact that there have been “limited legal challenges in the past [as relate to the ADA]” (Melson-Silimon, Harris, Shoenfelt, Miller, & Carter, Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019, p. 119), the focal article presents a case for an overly cautious, regressive, and, in the end, less effective framework to assess potential employees. We posit that avoiding personality tests that assess constructs tangentially related to personality disorders (PDs) and PID-5 dimensions is not only limiting within applied settings but that it is counter to the more expansive and sophisticated reconceptualization process underway in the larger field of psychology to better capture the nuances of individual personality. Furthermore, the current body of case law is too limited in quantity and applicability to necessitate such a restrictive approach.
As noted in the focal article, with each permutation of the Diagnostic and Statistical Manual of Mental Disorders (DSM), the field of psychology displays an improved understanding of the human condition, increasingly employing models that offer room for more varied and nuanced aspects of personality. Most notably, there has been increased emphasis on recognizing that personality characteristics may fall along a continuum (from normal to pathological). This move has been afoot for some time (i.e., circa 2013; DSM-5) with regard to anxiety, depression, and their associated disorders. However, only recently has a similar scheme been offered for traits commonly associated with PDs, which have long been viewed as the more pervasive and intractable of disorders. However, the fact that such models have been employed for some time regarding anxiety and depression provides guidance as to how employment practices might be affected by the newer models of personality disorders.
The focal article suggests that because the modern (i.e., normal–pathological) continuum view of personality traits has now extended to traits associated with PDs, we should now operate with a sense of (assessment) trepidation. However, this downplays the fact that there have been limited negative ramifications like the ones foreshadowed by the focal article related to anxiety and depression. Not to mention that, in comparison to anxiety and depression, personality disorders are exceedingly rare, ranging in prevalence rates from a high of 7.9% (obsessive compulsive disorder) to a low of .06% (dependent personality disorder; APA, 2013). Thus, it seems highly unlikely that the use of assessment tools with any individual would unintentionally tap into pathological personality traits noted in the alternative model of PD (AMPD) of DSM-5. At a minimum, it does not seem effective to choose a less comprehensive or less valid personality test out of fear that the variables of interest (KSAOs) may “approximate” PDs. Further, although the focal article attempts to make an argument that a number of FFMs have been shown via meta-analyses to have some relation to PDs, the point is overstated, as the estimates of association range from weak (.02) to moderate (.47), with the majority being less than .20. The most frequently used personality factor, conscientiousness, had a mean correlation with personality disorders of only –.08 (Saulsman & Page, Reference Saulsman and Page2004).
Furthermore, a personality score alone is insufficient to identify and formally diagnose personality disorders. DSM-5 diagnostic criteria require evidence of impairments in personality functioning in self and interpersonal domains along with pathological trait scores. Evidence of impairments is most often from clinical interviews and other-reports (e.g., family). Extreme test scores, in isolation, are not sufficient to conclude a diagnosable condition such as PD, and no clinical psychologist (diagnostic specialist) would make a diagnosis based on one score from one measure. In fact, Dr. Koransky, the clinical psychologist expert witness in Karraker v. Rent-A-Center (2005), argued that the personality scale used by Rent-A-Center “does not diagnose or detect any psychological disorders” (p. 837). Instead, “an elevated score on the personality scale [was] one of several [emphasis added] symptoms which may contribute [emphasis added]” to a diagnosis of paranoid personality disorder (p. 837). This argument was understood by the lower court in Karraker (2005), which ruled that the nonmedical use of the MMPI was appropriate; and, interestingly, to date, there have been minimal affirmed legal challenges under the ADA that are directly based on the utilization of a personality test like the MMPI.
On appeal, the judgment against Rent-A-Center resulted not from the potential for a test score to identify individuals with PDs but rather from poor justification by Rent-A-Center concerning the (nonmedical) use of the MMPI and insufficient validity evidence linking the test with job competencies. The 7th Circuit specified that while psychological tests “‘designed to identify a mental disorder or impairment’ qualify as medical examinations,” “psychological tests ‘that measure personality traits such as honesty, preferences, and habits’”—which can logically be tied to job requirements—“do not” (p. 835; citing EEOC Enforcement Guidance, 1995). Recall that some 10 years earlier in Thompson v. Borg-Warner (1996), the defendant successfully justified its use of personality testing as it related to applicant character and personality traits and their relation to job duties. Thus, case law suggests that personality tests are acceptable under the ADA when they are clearly linked to KSAOs—in other words, we argue that the requirements for establishing the validity of selection tests have not changed.
We also think it important to clarify the applicability of the case law discussed in the focal article. The Thompson (1996) case is a district court (lowest level federal court) of California case that ultimately settled. Thus, this case has minimal persuasive authority, at best. Barnes (1997) is an 11th Circuit (mid-level federal appellate court) case. The 11th Circuit appellate court binds only federal district courts in Alabama, Florida, and Georgia. Finally, Karraker (2005) is a 7th Circuit (mid-level federal appellate court) case. The 7th Circuit appellate court binds only federal district courts in Illinois, Indiana, and Wisconsin. The Soroka (1991) case was brought as a constitutional privacy law claim under the state court system—a court system entirely distinct from the federal court system that analyzes ADA claims (see Figure 1). Thus far, the cases involving personality test challenges under the ADA are too few and too limited in their binding applicability to draw firm conclusions on what the law states in this area. Further, the vast majority (approximately 95%) of cases settle before trial (Hirby, n.d.). This includes several prominent cases involving ADA claims based on personality test items, such as the 2015 litigation against Target, where the retail giant was forced to shell out $2.8 million per its settlement agreement with the EEOC (Zillman, Reference Zillman2015). Because the terms of the settlement and the factual details of the case are typically kept confidential when parties settle, the problematic test items, inventories, and circumstances of the employee selection process are never released. Thus, we should be cautious about suggesting a blanket limitation on the use of personality test items in employee selection.
Structure of the United States Federal and State court systems.
Furthermore, the EEOC guidance discussed in the focal article should be taken as exactly that—guidance. Although the EEOC is authorized to issue interpretive and procedural guidance to advise employers on how to comply with the laws it enforces, the EEOC cannot issue regulations with the force of law (Wern, Reference Wern1999). There is also some evidence that the Supreme Court defers to the EEOC less frequently than other federal agencies (only 54% of the time, compared to 72% of the time for other federal agencies; Wern, Reference Wern1999). Therefore, although the EEOC guidance cited in the focal article should not be ignored, it should not be taken as a source of binding legal precedent either.
The focal article also fails to adequately appreciate the larger and more significant paradigm shift that has taken place within the field of clinical psychology at large from a destigmatization standpoint. It is indeed meaningful that this convergence of the normal and pathological models has finally extended to PDs, which have long been viewed as largely intractable and fully (over)pathologized in a categorical sense, with little space for nuanced aspects of associated personality traits. We would argue that this actually has the potential to normalize and destigmatize a person’s mode of being (e.g., higher scores on a trait) that was once universally and solely viewed as pathological. This is actually a “win” for people with disabilities, as it has the potential to open doors once viewed as closed because of the perception of pathology. With the expansion of traits to a singular continuum, the very notion of “normal” and “disordered” is shifting and so is the meaning of test scores and their resultant inferences. Although there are certainly some personality characteristics that presumably represent a poor KSAO fit with some occupations, in truth, these have always existed, and will always loom over any selection processes.
In summary, it seems premature to leap to the conclusion that there is “a seemingly inevitable collision between the practice of personality testing for employment purposes and the scientific understanding of personality models” (Melson-Silimon et al., Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019, p. 120). Few ADA challenges have been brought concerning personality models’ correlations with anxiety and depression—which are far more prevalent in the population than the personality disorders discussed here. Further, FFM personality tests have only limited correlations with, and are insufficient for, the diagnosis of personality disorders, and the applicability of case law concerning personality tests and PDs is limited and far from resolved. As always, careful job analysis and validation of selection procedures should be undertaken. As long as personality is validity linked with job-related KSAOs, we see little cause for a change of course at this time.