Work Behavior: What It Means for Employers

What it means in practice

What the term means. "Work behavior" sounds like an everyday phrase about how people act at work, but in federal employment law it is a technical term. The Uniform Guidelines define it as "an activity performed to achieve the objectives of the job," involving "observable (physical) components and unobservable (mental) components," and consisting of "the performance of one or more tasks" (29 CFR 1607.16(Y)). Reconciling a ledger is a work behavior, and the mental steps behind it are part of it. Conduct rules, the everyday sense of the phrase, are covered later on this page.

The related defined terms. "Observable" means "able to be seen, heard, or otherwise perceived by a person other than the person performing the action" (1607.16(O)). A knowledge is "a body of information applied directly to the performance of a function," a skill is a present, observable competence to perform a learned psychomotor act, and an ability is "a present competence to perform an observable behavior or a behavior which results in an observable product" (1607.16(M), (T), (A)). A job analysis is "a detailed statement of work behaviors and other information relevant to the job"; a job description is only "a general statement of job duties and responsibilities" (1607.16(K), (L)).

Why it matters: selection procedures. UGESP applies to tests and other selection procedures used for decisions such as hiring, promotion, demotion, referral and retention (1607.2(B)). It does not require validity studies where no adverse impact results (1607.1(B)). How impact is measured is on the adverse impact page and the defense is on the business necessity page. Work behavior links the job to the test in all three validation strategies UGESP recognizes: criterion-related, content and construct (1607.5(A)-(B)).

Step one: the job analysis. A validity study should be based on a review of information about the job, normally including a job analysis, and "any method of job analysis may be used if it provides the information required for the specific validation strategy used" (1607.14(A)). For content validity, the analysis should cover the important work behaviors, their relative importance and any work products, and "should focus on the work behavior(s) and the tasks associated with them." If a behavior is not observable, it should analyze the observable aspects and work products (1607.14(C)(2)). The Q&A says a description of every task is not required, but the analysis should describe all important work behaviors, their relative importance and their level of difficulty.

Content validity: sampling the work. A content-valid procedure shows behaviors that "are a representative sample of the behavior(s) of the job," or samples its work product (1607.14(C)(4)). As a test less resembles a work behavior, "the greater the need for other evidence of validity." A test of a knowledge, skill or ability must define it operationally and show it is a necessary prerequisite to critical or important work behaviors. Content validity is not appropriate for traits or constructs "such as intelligence, aptitude, personality, commonsense, judgment, leadership, and spatial ability," or for skills learned on the job (1607.14(C)(1)). A content-valid procedure may be used if it represents a critical work behavior, one "necessary for performance of the job," or behaviors making up most of the important parts of the job (1607.14(C)(8)).

Criterion-related and construct validity. In a criterion-related study, test scores are compared statistically with job performance, and the criteria "should represent important or critical work behavior(s) or work outcomes" (1607.14(B)(3)). Production rate, error rate, tardiness, absenteeism and length of service may be used without a full job analysis if their importance is shown. Because of possible bias, supervisory rating techniques and instructions "should be carefully developed" (1607.14(B)(2)); see performance review. A construct study needs a job analysis naming the critical work behaviors and underlying constructs, plus empirical evidence linking them (1607.14(D)(2)-(3)).

Comparing jobs and borrowing studies. An employer may rely on another user's or a publisher's criterion-related study only if, among other conditions, the incumbents of both jobs "perform substantially the same major work behaviors," shown by job analyses of both (1607.7(B)(2)). For construct validity, if neither the observed work behaviors nor the observed work products of two jobs are the same, the agencies "will presume that the work behavior(s) in each job are different" (1607.14(D)(4)(b)). A test's reputation, name, promotional literature or sellers' testimonials never substitute for validity evidence (1607.9(A)), and the Q&A says an employer whose tests are run by an agency or consultant remains responsible.

Pass/fail versus ranking. Evidence enough for pass/fail use "may be insufficient" for ranking (1607.5(G)). A content-valid procedure supports ranking only if a higher score is likely to mean better job performance (1607.14(C)(9)). Cutoff scores should normally be reasonable and consistent with normal expectations of acceptable proficiency within the work force (1607.5(H)).

Example 1: lifting, reading and typing. These are the agencies' own Q&A examples. If a warehouse job analysis shows lifting 50 pounds is essential, but not that lifting more improves performance, a 50-pound lift test can be justified by content validity for pass/fail use, but ranking by weight lifted would be inappropriate. For a machine operator, reading simple instructions is essential though not a major part of the job, so it is a critical behavior and a content-valid reading test should be pass/fail. For a typing pool job made up almost entirely of typing, ranking on an error-adjusted words-per-minute test is appropriate.

Example 2: a bookkeeper test. As our own illustration, suppose a job analysis lists reconciling monthly bank statements as a critical work behavior for a bookkeeper. A test asking candidates to reconcile a sample statement in the job's format closely approximates that behavior; the Q&A says paper-and-pencil work samples fit best where, as in bookkeeping, the work itself is done that way. A personality inventory measuring "dependability" could not rest on content validity, because the Q&A says traits are not directly observable, though promptness and regularity of attendance in a prior work record may be supported by content validity.

Documenting work behaviors. Where a total selection process has adverse impact, users should keep validity documentation for each component with adverse impact (1607.15(A)(3)). A content validity report should completely describe the work behaviors, tasks and work products, give measures of criticality or importance, relate each knowledge, skill and ability to each work behavior, and identify the work behaviors each part of the procedure samples (1607.15(C)(3), (5)). A report missing "essential" items is considered incomplete unless the user shows they were unavailable or irrelevant (1607.15(A)(3)(c)).

Work behaviors and the ADA. The EEOC's appendix to its ADA rule says UGESP is "similarly inapplicable" to the ADA (appendix to 29 CFR 1630.10). The ADA asks instead about essential functions, "the fundamental job duties" of the position, not marginal ones (29 CFR 1630.2(n)(1)). Evidence includes the employer's judgment, job descriptions written before advertising or interviewing, time spent, the consequences of not requiring the function, bargaining agreements and incumbents' experience (1630.2(n)(3)); the statute says such a pre-advertising description "shall be considered evidence" (42 U.S.C. 12111(8)). The appendix says the first question is whether the employer actually requires the function: if no one in the job has ever had to type, typing is likely not essential. Criteria that screen out people with disabilities must be job related and consistent with business necessity, and criteria unrelated to an essential function are not (1630.10 and appendix). See Americans with Disabilities Act.

Production standards and test administration. The appendix also says the essential-functions inquiry does not second-guess production standards. An employer requiring typists to type 75 words per minute accurately, or hotel workers to clean 16 rooms a day, need not justify the level, but must impose it "in fact, and not simply on paper." Tests must measure the skill they claim to measure, not an impaired sensory, manual or speaking skill (29 CFR 1630.11); the EEOC's example is an oral test for an applicant who disclosed dyslexia beforehand.

The everyday meaning: conduct standards. In HR practice, "work behavior" also means conduct. The EEOC's 2008 technical assistance says the ADA gives employers "wide latitude to develop and enforce conduct rules." Rules against violence, threats, stealing or destruction of property always meet the job-related and consistent with business necessity standard, and employers may prohibit insubordination, require respect toward customers and bar yelling, cursing or shoving between coworkers. If a disability causes a violation, discipline is allowed if the rule meets that standard and others are held to it too. Rules need not be in a handbook if they meet the standard and are applied consistently, though the EEOC warns ad hoc rules are harder to defend. See insubordination, disciplinary action and employee handbook.

Federal agencies and job analysis. Each federal employment practice must be based on a job analysis identifying basic duties and responsibilities, the knowledges, skills and abilities required, and the factors important in evaluating candidates, with a rational relationship between job performance and the practice (5 CFR 300.103(a)-(b)). OPM calls job analysis "the foundation for all assessment and selection decisions," and its six-step checklist has subject matter experts rate tasks and competencies and link the two. These rules bind federal agencies, not private employers.

O*NET work activities. O*NET OnLine, published under DOL's Employment and Training Administration, describes work activities as "general types of job behaviors occurring on multiple jobs." It is a descriptive resource, not a legal standard; UGESP looks to analyses of the user's own job (1607.7(B)(2)).

Federal enforcement posture in 2026. On June 9, 2026, the Justice Department announced an Office of Legal Counsel opinion telling the EEOC its disparate-impact guidelines are unconstitutional; OPM summarizes it as also finding UGESP's validation framework goes beyond what Title VII requires. OPM's July 31, 2026 interim final rule removed UGESP references from federal personnel rules but kept the job-analysis and rational-relationship requirements. DOL rescinded 41 CFR part 60-3 effective October 26, 2026, saying this does not affect other agencies' use of UGESP. As of October 29, 2026, 29 CFR part 1607 is still in the eCFR, and a Federal Register search found no EEOC rescission.

Where states differ. State laws can add selection rules. As one example, the California Civil Rights Department says regulations effective October 1, 2025 make clear that an automated-decision system may violate California law if it harms applicants or employees based on protected characteristics, and require employment records, including automated-decision data, to be kept at least four years. Check your state agency.

Recordkeeping. Application forms, test papers and other hiring and promotion records must be kept one year from the record or the personnel action, whichever is later, and after a Title VII, ADA or GINA charge, all relevant records, including other candidates' test papers, until final disposition (29 CFR 1602.14). For example, a test paper from an applicant rejected on March 2, 2026, with no charge filed, is kept until at least March 2, 2027. Users with 100 or fewer employees may meet UGESP recordkeeping with yearly counts of hires, promotions, terminations and applicants by sex and, where appropriate, race and national origin, plus the procedures used (1607.15(A)(1)).

Why it matters and common mistakes

How Remindax helps

Related terms

See all terms in Workforce Trends & Culture or browse the full HR Glossary.