The Uniform Guidelines on Employee Selection Procedures (Guidelines) were published in August 1978 as the result of a joint effort involving the governmental organizations responsible for enforcing equal employment opportunity laws: the Equal Employment Opportunity Commission (EEOC), Department of Labor (including the Office of Federal Contract Compliance Programs [OFCCP]), Civil Service Commission, and Department of Justice. The Guidelines define discrimination in the context of employment selection procedures, which include tests, interviews, simulations, minimum requirements, or other tools used to make employment decisions. The Guidelines outline the validity and adverse impact evidence that the enforcement agencies would consider when evaluating a discrimination claim under Title VII of the Civil Rights Act of 1964, and Executive Order 11246. The Guidelines apply to all public and private employers covered by these two laws.
This review provides an overview of the Guidelines, including their purpose, major topics covered and not covered, and conclusions and implications. This review should not be construed as legal advice or a recommended interpretation of the Guidelines. Although this review takes into account some of the developments occurring since the original 1978 publication date, a comprehensive summary of such developments is not possible. In addition, certain topics that are no longer relevant or that the Guidelines address only tangentially have been excluded from this review. A References: section is provided listing sources for more detail about topics within the Guidelines and for information about relevant events occurring since the Guidelines’ publication.
Purpose
The stated purpose of the Guidelines is to codify a single set of standards to aid organizations in complying with federal laws prohibiting employment discrimination based on race, sex, or ethnicity. They are designed to provide a framework for determining the proper use of employment selection procedures. The Guidelines are not designed to apply to other forms of discrimination (e.g., age-based, disability-based).
Topics Covered By the Guidelines
Definition and Documentation of Adverse Impact
The Guidelines define discrimination as an employer’s use of a selection procedure that has an adverse impact on members of any race, sex, or ethnic group. Discrimination in this form would require an employer to produce appropriate validation evidence. That is, validation evidence, although valuable for other purposes, is not required by the Guidelines if adverse impact does not exist. The Guidelines also advocate reasonable efforts to consider adverse impact while comparing alternative selection procedures that are of approximately equal validity.
The Guidelines recommend that employers collect and retain documentation regarding selection procedures’ impact on different race, sex, and ethnicity groups. Federal enforcement organizations (e.g., EEOC, OFCCP) provide current guidance about specific methods for gathering this information. The Guidelines also introduce the four-fifths rule of thumb regarding adverse impact. Under this rule, adverse impact exists if the selection rate for a particular race, sex, or ethnic group is less than 80% that of the group with the highest selection rate.
General Standards for Validity Studies
The Guidelines espouse a three-component model of validity: criterion-related, content, and construct. More recent documents such as the 2003 Principles for the Validation and Use of Personnel Selection Procedures (Principles) and the 1999 Standards for Educational and Psychological Testing (Standards) instead view validity as a unitary concept incorporating numerous sources of evidence to justify interpretation of a selection procedure’s results. Despite this discrepancy, most of the Guidelines’ standards for validity studies retain relevance and are reviewed in the following text.
The Guidelines define suitable criterion-related validity evidence as data demonstrating that a selection procedure has a statistical relationship with job performance. Content validity evidence should demonstrate overlap between the content of the procedure and the content of the job. The Guidelines define construct validity as evidence that a selection procedure measures characteristics that have in turn been linked to successful job performance. However, because the Principles and Standards no longer recognize construct validity as a distinct form of evidence, it is excluded from the remainder of this review.
Key design features of acceptable validity studies under the Guidelines include adequate documentation of validity, accuracy and standardization, and correspondence between methods used during the validity research and those used operationally. The Guidelines also promote the concepts of setting cutoff scores based on acceptable proficiency, avoiding selection procedures that focus on characteristics learned during a brief orientation program, and avoiding in most circumstances the use of selection procedures to evaluate an applicant’s suitability for a higher-level job. The Guidelines also outline the conditions necessary to use a selection procedure in an interim manner pending validation data collection and the importance of periodically reviewing validation studies for currency.
Alternative Validation Strategies
The Guidelines outline several circumstances in which alternative validation strategies are possible, and the standards that these strategies must nonetheless meet. The Guidelines suggest that formalizing and quantifying a selection procedure can allow a user to conduct appropriate validity research. They also advocate elimination of adverse impact as a primary objective, even if a formal validity study is not technically feasible.
The Guidelines state that validity studies conducted by other users or test publishers may provide acceptable validity evidence, but the ultimate responsibility for adhering to the Guidelines resides with the end user. They also outline conditions (i.e., validity evidence, job similarity, and fairness evidence) necessary to transport this external validity evidence to a new setting. A recurring theme is the careful consideration of variables that may substantially affect validity when relying on other studies. Examples include differences in work behaviors, criterion measures, and experience levels.
The Guidelines propose multiorganization studies as an approach to meet validation standards that an individual organization could not otherwise fulfill. They describe unacceptable forms of validity evidence such as nonempirical information obtained from an external test provider. The Guidelines also convey a stance regarding professional supervision of validation activities: Such supervision is encouraged, but it does not alleviate the need for documented validity evidence. They also clarify that employment agencies retain responsibility for following the Guidelines as a developer and a user of selection procedures.
Technical Standards for Validation Studies
Given the level of detail within the Guidelines for this topic, a thorough review is not possible. Rather, key concepts for each subtopic are briefly described. This topic comprises four subtopics, one for each of the Guidelines’ three forms of validity: criterion-related, content, and construct, as well as a preceding section emphasizing the importance of a job analysis regardless of the specific validation approach chosen. As noted earlier, construct validity is no longer considered a distinct validation approach and is omitted from this review.
For criterion-related validity studies, the Guidelines outline several design considerations: technical feasibility (e.g., sample size), job analysis-based and uncontaminated criterion measures, sample representativeness, and the general (albeit not exclusive) suitability of a .05 significance level for evaluating selection procedure-performance relationships. However, the Guidelines eschew stating a minimum acceptable magnitude for these relationships, and recommend also considering adverse impact in the final choice of a selection procedure. The Guidelines define fairness as a selection tool’s ability to predict job performance equally well regardless of race, sex, or ethnic group; several considerations for conducting fairness studies are also summarized.
The Guidelines emphasize that a content validation approach should be limited to selection procedures that provide a representative sample of the job’s content. A core feature of the Guidelines’ standards on this topic is use of job analysis to delineate core knowledges, skills, and abilities (KSAs), and compilation of evidence linking both the selection procedure and job content to these KSAs. Additional content validity considerations include reliability of the selection procedure, justification of prior training or experience requirements, and evidence that higher selection procedure scores are linked to higher performance prior to use of a rank-order method for com-paring applicants.
Documentation of Impact and Validity Evidence
The Guidelines provide a detailed description of documentation requirements, adverse impact of selection procedures in most cases, and criterion-related or content validity evidence if adverse impact is detected. The Guidelines also propose documentation requirements for the alternative validation strategies noted earlier. Because a comprehensive summary of documentation requirements is not possible in this review, referral directly to the Guidelines on these issues is recommended.
Topics Not Covered By the Guidelines
It is important to recognize that the Guidelines do not address several important topics that were either nonexistent or insufficiently advanced when the guidelines were published; each of these topics is reviewed within both the Principles and the Standards. One set of such topics comprises alternative validation approaches: validity generalization, synthetic or job component validity, and meta-analysis. Other absent topics deal with sources of validity evidence: internal structure, response process, consequences of testing, and convergent or discriminant validity. The Guidelines’ definition of fairness is also limited to prediction bias and does not include other forms of bias, such as differential item functioning between groups. Finally, the Guidelines do not cover utility analysis as a broader approach for evaluating the potential usefulness of a particular selection procedure.
Conclusions and Implications
Along with the Principles and Standards, the Guidelines represent one of three primary documents with direct implications for employment practice and litigation related to selection procedures. However, as the earliest of these documents, the Guidelines omit certain developments in employment selection as noted previously; they have also been subject to more than 25 years of interpretation within the courts. Recent reviews also suggest that direct references to the Guidelines within recent civil rights cases have been rare, and their influence on litigation may more commonly be indirect by means of interim court decisions. Therefore a degree of caution is recommended to avoid overreliance on the Guidelines without consideration of intervening case law or the more recent developments incorporated by the Principles and Standards. Despite these cautions, the Guidelines nonetheless represent the current official standpoint of the agencies charged with enforcing Federal employment laws regarding adverse impact, validity, and other issues relating to employment selection. As such, the Guidelines retain prominence as an important reference document in the area of employee selection.
References:
- American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (1999). Standards for educational and psychological testing. Washington, DC: American Psychological Association.
- Equal Employment Opportunity Commission, Civil Service Commission, Department of Labor, & Department of Justice. (1978). Uniform guidelines on employee selection procedures. Federal Register, 43(166), 38295-38309.
- Gatewood, R. D., & Field, H. S. (1994). Human resource selection (3rd ed.). Toronto, Canada: Harcourt Brace.
- Guion, R. M. (1998). Assessment, measurement, and prediction for personnel decisions. Mahwah, NJ: Lawrence Erlbaum.
- Jeanneret, R. (2005). Professional and technical authorities and guidelines. In F. J. Landy(Ed.), Employment discrimination litigation: Behavioral, quantitative, and legal perspectives (pp. 47-100). San Francisco: Jossey-Bass.
- Society for Industrial and Organizational Psychology. (2003). Principles for the validation and use of personnel selection procedures (4th ed.). College Park, MD: Author.
See also: