Legal considerations directly influence industrial-organizational (I-O) psychology. To understand the consequences of selection measures, practitioners must consider how the law guides use of these measures. However, I-O psychologists should take caution using legal interpretation and recommendations without proper training or supervision. The focal article shows how a minor misinterpretation can lead to overbroad generalizations and misguided recommendations. Melson-Silimon, Harris, Shoenfelt, Miller, and Carter’s (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) misconstrual of case law and official Equal Employment Opportunity Commission (EEOC) guidance has the potential to incite unnecessary concern in I-O practitioners and researchers.
Hence, we clarify federal jurisprudence shaping the application of the Americans with Disabilities Act, as amended (ADA, 1990), to personality-based selection instruments. By providing a more precise understanding of the relevant case law, we will show that Melson-Silimon et al.’s (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) exploration of the relationship between normal personality and clinical personality has little bearing on whether personality measures constitute ADA “medical examinations.” Although we believe they draw flawed legal inferences, the focal article provides an adequate overview of the ADA’s statutory scheme and the EEOC’s proposed test for identifying ADA “medical examinations.” Thus, we do not repeat an overview here.
Beyond that, substantial errors exist in Melson-Silimon et al.’s (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) legal interpretations, but a clarification of each is beyond the scope of this response. We thus encourage readers to exercise caution and scrutiny when considering the legal conclusions in the focal article. In this commentary, we intend only to provide a broad clarification of the relevant legal landscape, which guides determination of illegal medical examinations. This analysis leads us to question the basis for recommendations in the focal article. We also address concerns of disparate impact from a practitioner perspective. We limit the scope of our commentary to federal law. Relevant antidiscrimination state laws exist and vary by state. We do not intend this response as legal advice. Readers with specific questions should consult independent legal counsel.
ADA medical examinations must be designed for clinical use
Based on the current state of ADA jurisprudence, courts will not classify normal personality assessments as ADA “medical examinations.” This is clear regardless of the evidence that normal personality relates to personality disorders. Case law analysis necessitates consideration of the jurisdiction and legal authority of the court issuing a given opinion. Melson-Silimon et al. (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) rely on three cases to conclude that all personality assessments could constitute ADA medical examinations. Only one of the cited cases has any binding legal authority over other courts. That case was the Seventh Circuit case, Karraker v. Rent-A-Center, Inc. (hereafter, Karraker, 2005). Not that they render any novel or inconsistent holdings, but the other two cases were district court cases and thus do not serve as a useful source of general case law. Even Karraker’s reasoning binds courts only within the Seventh Circuit. However, with no U.S. Supreme Court ruling on the issue, a U.S. Court of Appeals opinion is the next best source for gleaning case law information. With limited space, we focus on clarifying Karraker.
Legal interpretation requires precise examination of each word of a judicial holding. The Karraker (2005) court was careful to limit its holding to the issue of whether the Minnesota Multiphasic Personality Inventory (MMPI) was a “medical examination” under the ADA. The Karraker court precisely found that “because it is designed [emphasis added], at least in part, to reveal mental illness and [emphasis added] has the effect of hurting the employment prospects of one with a mental disability, we think the MMPI is best categorized as a medical examination” (p. 837). The word “designed” makes the intent of the assessment developer crucial. The use of the conjunctive “and” is also highly informative to legal interpretation. It makes both elements required. Hence, an adverse effect is not enough to render a measure an ADA medical examination. The measure must also be designed for a clinical purpose. Indeed, the Karraker court explained at the outset that the “case largely turn[ed] on whether the MMPI test is designed [emphasis added] to reveal a mental impairment” (p. 835).
Although they acknowledge that assessments designed to measure mental disorders are medical examinations, Melson-Silimon et al. (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) go on to misconstrue the holdings apparent in the case law: “Finally, and perhaps most important for the current discussion, measures of normal personality traits are permitted under the ADA, but any measure that might yield insight [emphasis added] to medical disorder or impairment is expressly prohibited prior to a conditional job offer” (p. 124). This improperly conflates unintentional discrimination with intentional discrimination, both of which have distinct legal and practical implications. As is clear from the Karraker (2005) holding, yielding insight on a medical disorder is insufficient to prohibit use of a measure for preemployment contexts. Unfortunately, Melson-Silimon et al. use this inaccurate legal interpretation to justify the exploration of the relationship between normal and clinical personality that fuels their concerns and recommendations.
Prominent employment law scholars and casebook editors, Estreicher, Harper, and Tippet (Reference Estreicher, Harper and Tippett2016), provide further support for our interpretation of Karraker (2005). Citing this case, Estreicher et al. (Reference Estreicher, Harper and Tippett2016) explained that psychological measures are permitted unless they were “designed to detect mental disability” (p. 558). Indeed, the purpose of the provision of the ADA that prohibits the use of preemployment medical examinations was to deter intentional disability discrimination. Hence, the requirement that the developer must have designed it intentionally to measure mental impairments is dispositive.
For example, an assessment is likely designed to measure mental impairments when it is developed to diagnose disorders (e.g., EEOC, 1995), and such use is endorsed in its manual. On the other hand, evidence to the contrary would likely exist when an assessment publisher subjects its measure to ongoing revisions to specifically remove any disability-related items. Note, however, courts have not yet provided specific guidelines to discern the purpose or intent of an assessment developer. Rather, they simply state the purpose of the assessment and move on. This is likely because, as with the MMPI in Karraker (2005), the purpose of assessment is made explicit by publishers’ instructions and manuals and is thus beyond in-depth judicial analysis.
It is also important to distinguish the intent of the assessment developer from the intent of the employer. As we demonstrated, the assessment developer’s intent was a driving component of the Karraker (2005) holding. Although relevant, an employer’s intent is not dispositive to classifying a measure as an ADA medical examination (see Kroll v. White Lake Ambulance Auth., 2012) but is required for finding disparate treatment discrimination. Yet, neither a discriminatory effect nor a correlation is sufficient to render a measure an ADA medical examination.
In addition, Melson-Silimon et al. (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) reviewed an EEOC guidance that is not applicable to their inferences or arguments. They referenced an EEOC publication from 2000 that expressly limits its scope to medical examinations during employment (cf. EEOC, 2000, para. 1). Job-related medical examinations are permitted during employment, whereas they are prohibited preemployment and pre-offer (ADA, 1990). As noted in Karraker (2005), internal promotions to new positions also fall within the preemployment context. Thus, this conversation should focus on the preemployment stage. Melson-Silimon et al. would have found information contradictory to their premise had they consulted the proper publication.
The EEOC guidance on preemployment medical examinations and inquiries contemplates that the ADA does not prohibit inquiries that tend to elicit medical information, as long as some potential responses would not contain medical information. The EEOC (1995) states, “On the other hand, if there are many possible answers to a question and only some of those answers would contain disability-related information, that question is not ‘disability-related’” (“The Pre-Offer Stage” section, para. 3). This suggests that the ADA prohibition on preemployment medical examinations and inquiries does not concern measures or items that might yield medical information.
The EEOC (1995) goes on to define an ADA medical examination as one that “seeks [emphasis added] information about an individual’s physical or mental impairments” (“What is a Medical Examination?” section, para. 1). The use of the word “seek,” again, points to the purpose, design, or intent of the given measure. Otherwise, they would have used a word like “yield,” “provide,” or “relate,” as does Melson-Silimon et al. (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019). Additionally, none of the factors in the EEOC’s proposed test for classifying medical examinations makes any mention of a measure that might yield insight to medical information, but they do expressly mention the design. This further supports the conclusion that measures designed for normal personality are not ADA medical examinations.
Although an accurate look at the extant case law is comforting to normal personality practitioners, future holdings could alter the application of the ADA to personality measures. However, a holding consistent with Melson-Silimon et al.’s (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) interpretation is highly unlikely. Judges tend to avoid holdings that would open the floodgates to future increases in litigation or that would impact an entire industry. A legal precedent that normal personality assessments constitute ADA medical examinations because they yield insight on psychological impairments would open the floodgates, rendering virtually all selection measures illegal preemployment measures. For example, general mental ability tests would tend to reveal information about cognitive disorders, and physical ability tests would tend to reveal information about physical impairments. An in-depth discussion of the empirical relationships of these additional constructs is beyond our scope. The point is that nearly any psychological or physical measure could provide information about the person’s psychological or physical health.
Along these lines, the EEOC (1995) pointed out that “physical fitness tests are not ‘medical’ examinations” (footnote 17), but that such a test that tends to screen out applicants with disabilities should still be job-related so as to comply with “other parts of the ADA” (footnote 17), meaning the prohibition against disparate impact. This reiterates the notion that merely yielding disability-related information is insufficient. This also highlights the crucial distinction between analysis of ADA medical examinations and disparate impact. There is no need for courts to render anything that might yield disability-related information a medical examination because the disparate impact theory already regulates such unintended effects. Indeed, a holding to this effect would make the disability disparate impact framework obsolete. Thus, a court finding normal personality assessments as constituting ADA medical examinations based on their relationship to personality disorders would lack any supporting precedent or judicial rationale.
Concerns for unintentional discrimination are not novel
Although the focal article’s legal discussion focuses on whether personality assessments constitute ADA medical examinations, some of Melson-Silimon et al.’s (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) empirical claims concerning clinical-normal personality correlations implicate consideration of disparate impact or unintentional discrimination. The concerns presented are not novel and apply to all psychological measures used in employment contexts. I-O psychology practitioners and publishers of prominent workplace personality measures understand the concept of disparate impact. The best test publishers and I-O psychologists actively ensure their assessments are job-related when they design an assessment for the workplace. They document information on the validation and job relevance for any measure with the potential to cause disparate impact. They have always considered ADA requirements on multiple fronts, including reasonable accommodations and item content.
Furthermore, personality measures provide incremental validity through the prediction of contextual performance, giving insight on the likelihood of relevant contextual behavior in the workplace (Borman, Penner, Allen, & Motowidlo, Reference Borman, Penner, Allen and Motowidlo2001). Practitioners demonstrate these relationships with performance through criterion-related validation studies, showing how personality measures predict key components of a given job. This has been an essential step in the I-O field for decades and will continue to play a role in the applied use of personality in the workplace. Although we acknowledge the focal authors’ concerns, we do not believe normal personality measures are currently in legal jeopardy. In fact, since 1978, plaintiffs have challenged personality assessments for disparate impact the least often of other common selection measures (Williams, Schaffer, & Ellis, Reference Williams, Schaffer and Ellis2013).
Conclusion
Therefore, the call to reconsider normal personality is unwarranted based on a misinterpretation of case law and overgeneralization of legal outcomes. We believe I-O professionals should focus on the misuse of assessments and methods not designed for the workplace. As demonstrated by Karraker (2005), there are key differences between clinical and normal measures of personality. Organizations need to understand the difference to avoid potential legal complaints based on the use of a measure like the MMPI in preemployment screening.
Despite our disagreement with the legal conclusions, we do agree with other recommendations in the focal article. Specifically, organizations should carefully consider implications of utilizing medical or clinical professionals in their employment decision-making processes. Courts will critique situations involving clinicians to a greater extent. Also, organizations and practitioners should use work-specific and job-related measures. However, this is not new. Test publishers deliberately make high-quality personality assessments for work-specific applications without items that directly ask about medical information, such as the Hogan Personality Inventory (Hogan & Hogan, Reference Hogan and Hogan2007) and the Hogan Development Survey (Hogan & Hogan, Reference Hogan and Hogan2009).
You have
Access
Legal considerations directly influence industrial-organizational (I-O) psychology. To understand the consequences of selection measures, practitioners must consider how the law guides use of these measures. However, I-O psychologists should take caution using legal interpretation and recommendations without proper training or supervision. The focal article shows how a minor misinterpretation can lead to overbroad generalizations and misguided recommendations. Melson-Silimon, Harris, Shoenfelt, Miller, and Carter’s (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) misconstrual of case law and official Equal Employment Opportunity Commission (EEOC) guidance has the potential to incite unnecessary concern in I-O practitioners and researchers.
Hence, we clarify federal jurisprudence shaping the application of the Americans with Disabilities Act, as amended (ADA, 1990), to personality-based selection instruments. By providing a more precise understanding of the relevant case law, we will show that Melson-Silimon et al.’s (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) exploration of the relationship between normal personality and clinical personality has little bearing on whether personality measures constitute ADA “medical examinations.” Although we believe they draw flawed legal inferences, the focal article provides an adequate overview of the ADA’s statutory scheme and the EEOC’s proposed test for identifying ADA “medical examinations.” Thus, we do not repeat an overview here.
Beyond that, substantial errors exist in Melson-Silimon et al.’s (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) legal interpretations, but a clarification of each is beyond the scope of this response. We thus encourage readers to exercise caution and scrutiny when considering the legal conclusions in the focal article. In this commentary, we intend only to provide a broad clarification of the relevant legal landscape, which guides determination of illegal medical examinations. This analysis leads us to question the basis for recommendations in the focal article. We also address concerns of disparate impact from a practitioner perspective. We limit the scope of our commentary to federal law. Relevant antidiscrimination state laws exist and vary by state. We do not intend this response as legal advice. Readers with specific questions should consult independent legal counsel.
ADA medical examinations must be designed for clinical use
Based on the current state of ADA jurisprudence, courts will not classify normal personality assessments as ADA “medical examinations.” This is clear regardless of the evidence that normal personality relates to personality disorders. Case law analysis necessitates consideration of the jurisdiction and legal authority of the court issuing a given opinion. Melson-Silimon et al. (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) rely on three cases to conclude that all personality assessments could constitute ADA medical examinations. Only one of the cited cases has any binding legal authority over other courts. That case was the Seventh Circuit case, Karraker v. Rent-A-Center, Inc. (hereafter, Karraker, 2005). Not that they render any novel or inconsistent holdings, but the other two cases were district court cases and thus do not serve as a useful source of general case law. Even Karraker’s reasoning binds courts only within the Seventh Circuit. However, with no U.S. Supreme Court ruling on the issue, a U.S. Court of Appeals opinion is the next best source for gleaning case law information. With limited space, we focus on clarifying Karraker.
Legal interpretation requires precise examination of each word of a judicial holding. The Karraker (2005) court was careful to limit its holding to the issue of whether the Minnesota Multiphasic Personality Inventory (MMPI) was a “medical examination” under the ADA. The Karraker court precisely found that “because it is designed [emphasis added], at least in part, to reveal mental illness and [emphasis added] has the effect of hurting the employment prospects of one with a mental disability, we think the MMPI is best categorized as a medical examination” (p. 837). The word “designed” makes the intent of the assessment developer crucial. The use of the conjunctive “and” is also highly informative to legal interpretation. It makes both elements required. Hence, an adverse effect is not enough to render a measure an ADA medical examination. The measure must also be designed for a clinical purpose. Indeed, the Karraker court explained at the outset that the “case largely turn[ed] on whether the MMPI test is designed [emphasis added] to reveal a mental impairment” (p. 835).
Although they acknowledge that assessments designed to measure mental disorders are medical examinations, Melson-Silimon et al. (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) go on to misconstrue the holdings apparent in the case law: “Finally, and perhaps most important for the current discussion, measures of normal personality traits are permitted under the ADA, but any measure that might yield insight [emphasis added] to medical disorder or impairment is expressly prohibited prior to a conditional job offer” (p. 124). This improperly conflates unintentional discrimination with intentional discrimination, both of which have distinct legal and practical implications. As is clear from the Karraker (2005) holding, yielding insight on a medical disorder is insufficient to prohibit use of a measure for preemployment contexts. Unfortunately, Melson-Silimon et al. use this inaccurate legal interpretation to justify the exploration of the relationship between normal and clinical personality that fuels their concerns and recommendations.
Prominent employment law scholars and casebook editors, Estreicher, Harper, and Tippet (Reference Estreicher, Harper and Tippett2016), provide further support for our interpretation of Karraker (2005). Citing this case, Estreicher et al. (Reference Estreicher, Harper and Tippett2016) explained that psychological measures are permitted unless they were “designed to detect mental disability” (p. 558). Indeed, the purpose of the provision of the ADA that prohibits the use of preemployment medical examinations was to deter intentional disability discrimination. Hence, the requirement that the developer must have designed it intentionally to measure mental impairments is dispositive.
For example, an assessment is likely designed to measure mental impairments when it is developed to diagnose disorders (e.g., EEOC, 1995), and such use is endorsed in its manual. On the other hand, evidence to the contrary would likely exist when an assessment publisher subjects its measure to ongoing revisions to specifically remove any disability-related items. Note, however, courts have not yet provided specific guidelines to discern the purpose or intent of an assessment developer. Rather, they simply state the purpose of the assessment and move on. This is likely because, as with the MMPI in Karraker (2005), the purpose of assessment is made explicit by publishers’ instructions and manuals and is thus beyond in-depth judicial analysis.
It is also important to distinguish the intent of the assessment developer from the intent of the employer. As we demonstrated, the assessment developer’s intent was a driving component of the Karraker (2005) holding. Although relevant, an employer’s intent is not dispositive to classifying a measure as an ADA medical examination (see Kroll v. White Lake Ambulance Auth., 2012) but is required for finding disparate treatment discrimination. Yet, neither a discriminatory effect nor a correlation is sufficient to render a measure an ADA medical examination.
In addition, Melson-Silimon et al. (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) reviewed an EEOC guidance that is not applicable to their inferences or arguments. They referenced an EEOC publication from 2000 that expressly limits its scope to medical examinations during employment (cf. EEOC, 2000, para. 1). Job-related medical examinations are permitted during employment, whereas they are prohibited preemployment and pre-offer (ADA, 1990). As noted in Karraker (2005), internal promotions to new positions also fall within the preemployment context. Thus, this conversation should focus on the preemployment stage. Melson-Silimon et al. would have found information contradictory to their premise had they consulted the proper publication.
The EEOC guidance on preemployment medical examinations and inquiries contemplates that the ADA does not prohibit inquiries that tend to elicit medical information, as long as some potential responses would not contain medical information. The EEOC (1995) states, “On the other hand, if there are many possible answers to a question and only some of those answers would contain disability-related information, that question is not ‘disability-related’” (“The Pre-Offer Stage” section, para. 3). This suggests that the ADA prohibition on preemployment medical examinations and inquiries does not concern measures or items that might yield medical information.
The EEOC (1995) goes on to define an ADA medical examination as one that “seeks [emphasis added] information about an individual’s physical or mental impairments” (“What is a Medical Examination?” section, para. 1). The use of the word “seek,” again, points to the purpose, design, or intent of the given measure. Otherwise, they would have used a word like “yield,” “provide,” or “relate,” as does Melson-Silimon et al. (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019). Additionally, none of the factors in the EEOC’s proposed test for classifying medical examinations makes any mention of a measure that might yield insight to medical information, but they do expressly mention the design. This further supports the conclusion that measures designed for normal personality are not ADA medical examinations.
Although an accurate look at the extant case law is comforting to normal personality practitioners, future holdings could alter the application of the ADA to personality measures. However, a holding consistent with Melson-Silimon et al.’s (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) interpretation is highly unlikely. Judges tend to avoid holdings that would open the floodgates to future increases in litigation or that would impact an entire industry. A legal precedent that normal personality assessments constitute ADA medical examinations because they yield insight on psychological impairments would open the floodgates, rendering virtually all selection measures illegal preemployment measures. For example, general mental ability tests would tend to reveal information about cognitive disorders, and physical ability tests would tend to reveal information about physical impairments. An in-depth discussion of the empirical relationships of these additional constructs is beyond our scope. The point is that nearly any psychological or physical measure could provide information about the person’s psychological or physical health.
Along these lines, the EEOC (1995) pointed out that “physical fitness tests are not ‘medical’ examinations” (footnote 17), but that such a test that tends to screen out applicants with disabilities should still be job-related so as to comply with “other parts of the ADA” (footnote 17), meaning the prohibition against disparate impact. This reiterates the notion that merely yielding disability-related information is insufficient. This also highlights the crucial distinction between analysis of ADA medical examinations and disparate impact. There is no need for courts to render anything that might yield disability-related information a medical examination because the disparate impact theory already regulates such unintended effects. Indeed, a holding to this effect would make the disability disparate impact framework obsolete. Thus, a court finding normal personality assessments as constituting ADA medical examinations based on their relationship to personality disorders would lack any supporting precedent or judicial rationale.
Concerns for unintentional discrimination are not novel
Although the focal article’s legal discussion focuses on whether personality assessments constitute ADA medical examinations, some of Melson-Silimon et al.’s (Reference Melson-Silimon, Harris, Shoenfelt, Miller and Carter2019) empirical claims concerning clinical-normal personality correlations implicate consideration of disparate impact or unintentional discrimination. The concerns presented are not novel and apply to all psychological measures used in employment contexts. I-O psychology practitioners and publishers of prominent workplace personality measures understand the concept of disparate impact. The best test publishers and I-O psychologists actively ensure their assessments are job-related when they design an assessment for the workplace. They document information on the validation and job relevance for any measure with the potential to cause disparate impact. They have always considered ADA requirements on multiple fronts, including reasonable accommodations and item content.
Furthermore, personality measures provide incremental validity through the prediction of contextual performance, giving insight on the likelihood of relevant contextual behavior in the workplace (Borman, Penner, Allen, & Motowidlo, Reference Borman, Penner, Allen and Motowidlo2001). Practitioners demonstrate these relationships with performance through criterion-related validation studies, showing how personality measures predict key components of a given job. This has been an essential step in the I-O field for decades and will continue to play a role in the applied use of personality in the workplace. Although we acknowledge the focal authors’ concerns, we do not believe normal personality measures are currently in legal jeopardy. In fact, since 1978, plaintiffs have challenged personality assessments for disparate impact the least often of other common selection measures (Williams, Schaffer, & Ellis, Reference Williams, Schaffer and Ellis2013).
Conclusion
Therefore, the call to reconsider normal personality is unwarranted based on a misinterpretation of case law and overgeneralization of legal outcomes. We believe I-O professionals should focus on the misuse of assessments and methods not designed for the workplace. As demonstrated by Karraker (2005), there are key differences between clinical and normal measures of personality. Organizations need to understand the difference to avoid potential legal complaints based on the use of a measure like the MMPI in preemployment screening.
Despite our disagreement with the legal conclusions, we do agree with other recommendations in the focal article. Specifically, organizations should carefully consider implications of utilizing medical or clinical professionals in their employment decision-making processes. Courts will critique situations involving clinicians to a greater extent. Also, organizations and practitioners should use work-specific and job-related measures. However, this is not new. Test publishers deliberately make high-quality personality assessments for work-specific applications without items that directly ask about medical information, such as the Hogan Personality Inventory (Hogan & Hogan, Reference Hogan and Hogan2007) and the Hogan Development Survey (Hogan & Hogan, Reference Hogan and Hogan2009).
Author ORCIDs
Chase A. Winterberg, 0000-0003-2483-140X