Job Analysis, Personnel Selection, and the ADA - Study Notes

Introduction

The ADA will change how employers screen and hire applicants; central concept is the notion of essential functions for hiring under the ADA.
Key questions addressed: meaning of essential functions under the ADA, how to conduct a job analysis to determine essential functions, and how essential functions influence selection.
The ADA (Title I) prohibits discrimination against a qualified person with a disability in employment. A qualified person with a disability is defined as "an individual with a disability who .., with or without reasonable accommodation, can perform the essential functions of such a position" (EEOC 1992, p. 11).
Determining essential functions is necessary for fair employment practices under the ADA.
Although the ADA does not require a job analysis to determine essential functions, it is difficult to determine them without some job analysis.
Sparks (1988) emphasized that job analysis is essentially required to justify tests before the courts (see also Thompson & Thompson 1982).
Purpose of the paper: explore the meaning of “essential function” as outlined by the ADA, discuss how conventional job analysis can provide information about essential functions, and consider the role of essential functions in selection.
The authors note potential challenges for job analysis under the ADA and offer directions for future research.

Select In vs. Select Out – Changing Perspectives

Historical perspective: since Taylor and Russell (1939) and Brogden (1949), HR has used tests to identify applicants predicted to excel; the rationale is that modest test-criterion correlations can yield productivity gains in favorable conditions.
Traditional view (select out): by rejecting a large proportion of applicants who could perform the job at some minimal level, employers gain a competitive advantage. This is described as a status quo, or "select out" philosophy.
The ADA promotes a "select in" philosophy: focus on helping applicants qualify for the job rather than excluding them; the ADA requires reasonable accommodation in testing and in how work is done (Solomon 1992).
Implication: employers must consider what applicants can do and how to accommodate so that all work gets done; the ADA may aid in learning how to accomplish this task.
Intent of the ADA: to provide access to jobs for millions currently denied work.

Using Statistical Information

HR often relies on statistical information for decision making; the ADA pushes toward case-by-case consideration.
Implications for HR practice include: much of EEOC-derived evidence has been statistical (e.g., applicant flow data for adverse impact).
Under the ADA, case-by-case examination becomes generally illegal to ask if an applicant is disabled on the application or in an interview; the ADA permits inviting individuals with disabilities to identify themselves on an application to comply with federal regulations (e.g., section 503 of the Rehabilitation Act).
Consequences: applicant flow records may be suspect or useless for showing adverse impact; even if data were available, disabilities vary widely within classes (e.g., visual impairment).
Example illustrating heterogeneity: three individuals with different visual impairments (nearsighted, colorblind, totally blind) would perform very differently on a task like operating a pharmacy cash register or sorting pills; thus, noting adverse impact for a visually impaired group may be unhelpful.

Proactive vs. Reactive Postures

Various sources (e.g., Frierson 1992) suggest the ADA requires a reactive rather than proactive posture for accommodations.
Process: the applicant identifies a problem in testing or task performance; employer and applicant determine the facts.
If the problem is not related to an essential function, the employer must accommodate.
If the problem relates to an essential function, accommodations are discussed; if a reasonable accommodation is found, the assessment of the applicant’s abilities is considered unbiased.
If no reasonable accommodation is found, the employer may find another job for the applicant or may reject the applicant; this is similar for current employees.
Theoretically, employers should be allowed to select the best-qualified applicants; e.g., if a disabled applicant, with accommodation, performs more poorly on a test than a nondisabled applicant, the employer should choose the nondisabled applicant if the test is related to job performance and consistent with business necessity.
Employers would like to have policy about which disabilities can be accommodated; without clear policy, recruiters and hiring managers face uncertainty about investment in applicants who cannot perform essential functions.
An early attempt in job analysis for ADA selection tried to create a matrix with common disabilities as rows and essential functions as columns, but the authors concluded this is not possible due to wide variation within disability categories and the enormous number of disabilities.
Example: three people with disabilities (e.g., vision impairments) could have very different capabilities for the same task; a fixed matrix would misrepresent capabilities and accommodations.
This variability encourages a shift away from rigid matrices toward case-by-case assessment and accommodation planning.

Determining Essential Functions

The ADA requires employers to focus on essential functions to determine if a person with a disability is qualified.
The ADA provides guidelines for determining essential functions:
(a) the position exists to perform the function,
(b) there are a limited number of other employees available to perform the function, or among whom the function can be distributed,
(c) the function is highly specialized and the person is hired for special expertise or ability to perform it.
Evidence to consider in determining essential functions includes:
(a) the employer's judgment,
(b) a written job description prepared before advertising or interviewing,
(c) the amount of time spent performing the function,
(d) the consequences of not requiring a person in this job to perform the function,
(e) the terms of a collective bargaining agreement,
(f) work experience of people who have performed the job in the past and current experience of people who perform similar jobs,
(g) other relevant factors (e.g., nature of the work operation, organizational structure) (EEOC 1992, pp. 13–18).
Although the ADA does not require a job analysis, a job analysis that yields information about these factors and results in a written job description is clearly helpful.
The authors describe a concrete job-analysis effort for two pharmacy assistant roles in a large retail drug store to illustrate the process and challenges. The roles included: (i) ringing up merchandise, stocking, answering the phone; (ii) covering for (i) and assisting the pharmacist by filling prescription forms, generating labels, and fetching drugs.
Method: task inventories were created after observations and interviews with ten incumbents per job and with supervising pharmacists; drafts reviewed by HR managers; after revision, large samples of incumbents and pharmacists responded; incumbents and pharmacists often disagreed on task judgments.
Incumbents provided information on:
(a) whether they performed the task,
(b) relative time spent on the task,
(c) consequences of error in task performance.
Essential tasks were defined using four criteria:
(a) the pharmacists thought the task was essential,
(b) the task had serious consequences of error,
(c) no one was available to cover the task,
(d) over $70\%$ of incumbents performed the task and spent above-average time on it.
The authors acknowledge a potential misalignment: the list of essential tasks was their policy decision, not necessarily management’s, and not guaranteed to be EEOC-acceptable. They highlight a concern that someone other than management determined essentiality, which could raise governance and legal concerns.
Policy capturing: to model how management judgments combine to determine essentiality, four steps: use multiple regression to model judgments; treat tasks as objects; independent variables as task aspects (centrality, coverage, time spent, etc.); the output is an explicit policy describing how task elements combine to determine essentiality. This can aid consistency and reliability in designating essential tasks.
However, in this case, explicit policy was deemed potentially risky legally: labeling a task as essential could be seen as company policy, exposing the firm to liability if misapplied.
Additional steps in the analyzed project included:
- Estimating sensory and motor requirements for each essential task, using the U.S. Employment Services Handbook for Analyzing Jobs, with an added cell for "twisting" (important for lower back disability) (Hollenbeck, Ilgen, & Crampton 1992).
- Making a table of sensory and motor requirements for essential tasks as typically performed; the authors later viewed this step as problematic because it could imply excluding disabled individuals and does not reflect accommodations. They argue it is better to evaluate disabled individuals on a case-by-case basis and to consult disability-accommodation groups.
The authors also had pharmacists rate importance of knowledge, skills, and abilities (KSA) for completing the job; such judgments are useful for general selection but not specifically for ADA purposes. These judgments were made about the job as a whole rather than per task; there is a practical constraint given the number of possible KSAs and tasks; extrapolating from task to ability involves a leap beyond typical expertise of either managers or analysts.
Defining “Essential” and “Reasonable”: the ADA invites litigation because both terms are vaguely defined. Considerations for reasonable accommodation include:
(a) the nature and cost of the accommodation,
(b) the overall financial resources of the facility or facilities,
(c) staffing levels of the facility,
(d) the effect on expenses and resources,
(e) the overall financial resources of the company,
(f) geographic/administrative/fiscal relationships of facilities to the employer,
(g) the impact of the accommodation on other employees’ ability to do their jobs.
The salary of the disabled employee cannot be considered when determining accommodation.
The authors conclude that essential tasks result from a combination of information from conventional job analysis; the combination approach is defensible, but others might choose differently, yielding different lists of essential tasks. A clear policy on essential functions and reasonable accommodations could reduce attorney fees and settlements.
They note a tendency among some firms’ legal counsel to advise “hiding the ball” by avoiding explicit policies and records, which complicates the court’s ability to assess employer practices.
The technology of decision analysis is proposed as a method to formulate policy in collaboration with disabled groups; evidence of successful use includes applications in land use management, highway design, and modifications to nuclear power plants (Keeney & Raiffa 1976; Mumpower & Adelman 1980; Brannick & Darling 1991). A more likely scenario, they argue, is a sequence of court decisions shaping more explicit legal definitions of essential and reasonable accommodations.

How to Describe Jobs for the ADA

Under the ADA, job descriptions must enable identification of essential functions for communication to applicants, lawyers, etc.
A wide range of job-analysis approaches can be used (e.g., Gael 1988; McCormick 1979).
Key decisions involve: types of descriptors (work activities, worker activities, machines, etc.), data sources (job analysts, incumbents, supervisors, etc.), and data collection methods (observation, interview, open-ended or closed-ended questionnaires, etc.).
The authors’ rationale:
- Work-oriented techniques yield information about what the job does (e.g., operating a cash register).
- Worker-oriented techniques yield information about the KSAs necessary to perform the work (e.g., manual dexterity).
- For ADA compliance, both types of information are useful for selection.
The two job roles studied were pharmacy assistants (not the pharmacist):
- Role 1: cash register operation, stocking, answering phones.
- Role 2: covers Role 1 but for shorter periods; supports the pharmacist by filling prescription forms, generating labels via computer, and fetching drugs/containers.
A task inventory was created after observations/interviews with ten incumbents per job and interviews with supervising pharmacists; after revisions, a large sample of incumbents and pharmacists responded.
Incumbents and pharmacists often disagreed on task judgments.
Task inventory data collection by incumbents included:
(a) whether they performed each task,
(b) the relative time spent on the task,
(c) the consequences of errors in task performance.
How to determine essential tasks (criteria):
- A task was deemed essential if it met one of four conditions:
  (i) the pharmacists considered it essential,
  (ii) the task had serious consequences of error,
  (iii) no one was available to cover the task,
  (iv) more than 70% of incumbents performed the task and spent above-average time on it.
The authors argue that the policy (the essential tasks list) was their own; management might have produced a different list, and EEOC could potentially disagree.
They discuss potential methods for eliciting policy:
- Policy capturing using multiple regression to model judgments by management about task aspects (centrality, coverage, time spent, etc.) and the resulting essentiality.
- The aim is a transparent and reliable basis for determining essential tasks, though legal concerns persist about creating explicit task-essentiality policies.
They took two additional steps:
- Estimating sensory and motor requirements for each essential task using the U.S. Employment Services Handbook, adding a twist (twisting) to account for lower back disability (Hollenbeck, Ilgen, & Crampton 1992).
- Pharmacists estimated the importance of KSA requirements for completing the job. These judgments were about the job as a whole rather than per task, which the authors argue may be less actionable for ADA purposes due to the volume of potential KSAs and the difficulty of linking task-level performance to general KSAs.
Importantly, the authors acknowledge a practical constraint: managers may not be willing or able to provide the thousands of judgments necessary to map every task to every possible ability; analysts may lack expertise to infer abilities from tasks.

Defining "Essential" and "Reasonable" (and ADA Policy Implications)

The ADA is intentionally vague about what constitutes essential vs. non-essential, and what constitutes a reasonable accommodation.
The factors to consider in reasonable accommodation (undue hardship considerations) include:
(a) nature and cost of the accommodation,
(b) overall financial resources of the facility or facilities involved,
(c) staffing levels,
(d) effect on expenses/resources,
(e) overall financial resources of the employer,
(f) geographic/administrative/fiscal relationships of facilities to the employer,
(g) impact of the accommodation on the ability of other employees to perform their jobs.
The salary of the disabled employee cannot be considered when deciding accommodation.
The authors emphasize that essential tasks result from integrating information from conventional job analysis; different combinations can yield different lists of essential tasks, underscoring the role of judgment and policy decisions.
A clear policy on determining essential functions and reasonable accommodations could help reduce attorney fees and settlements; however, pre-ADA practice sometimes encouraged avoiding explicit records or policies (e.g., to minimize EEOC scrutiny).
The authors advocate using decision analysis techniques to negotiate policies with disabled groups, as such methods have been successfully applied in other domains; court decisions are likely to eventually provide more explicit definitions of essential and reasonable accommodations.

Selection and Reasonable Accommodation in Testing

The ADA requires reasonable accommodation in the hiring process. An example: reading a paper-and-pencil test to a blind applicant.
Accommodating tests raises several validation questions:
(a) Does changing the test medium alter psychometric properties?
(b) Are validation studies feasible with disabled groups?
(c) Should disabilities be grouped for validation purposes?
Standards for Educational and Psychological Testing (1985) state that claims about psychometric properties of a test cannot be generalized to a modified version unless data show no adverse effects.
Research by ETS and U.S. Office of Personnel Management examined test characteristics within different disabled populations: hearing-impaired individuals taking regular or cassette versions, learning-disabled individuals taking cassette or large-type or braille versions, visually-impaired individuals taking braille or regular versions, etc.
Psychometric characteristics studied include reliability, speededness, and differential item performance (DIF); overall findings show comparable reliability and DIF across groups, but differences in performance level across groups (e.g., learning disabled and hearing-impaired groups scoring lower on some tests; hearing-impaired regular students scoring higher on the math section of the SAT than nondisabled peers).
Limitations: in groups with little within-group variability (e.g., legally blind, pre-lingually deaf), results are clearer; in groups with substantial variability (e.g., learning disabilities, physically disabled), results are more ambiguous.
The discussion reinforces the need to evaluate validation evidence carefully when accommodations alter test formats or populations; generalization from standard versions to modified versions requires data demonstrating no adverse impact on psychometric properties.

Challenges for Conventional Job Analysis and Future Research

1) Snapshot limitation: conventional job analysis captures the job as it exists and offers little information about accommodation; task inventories and other inventories provide limited guidance on shifting tasks across positions; traditional analyses are not designed to track ongoing changes in jobs.
2) Limited knowledge on accommodating disabled workers: Kochhar & Armstrong (1988) summarize current understanding; effective accommodation analyses must focus on aspects of the job where disabled workers have difficulty, including physical attributes (e.g., hand strength), mental attributes (sensation, cognition, affect), the workplace (tools/equipment), and the environment (air quality, noise, temperature). This implies the need for repeated analyses each time a new accommodation is requested and calls for more efficient methods.
3) Job vs. position focus: conventional job analysis focuses on the job (an abstract entity held by multiple people) whereas the ADA requires case-by-case decisions based on the specific position; differences in workload and staffing across stores or contexts mean that one store may accommodate a disabled applicant while another cannot; methods must account for position-to-position variation within a job title.
4) Organization of accommodation resources: a method to organize resources for accommodation is needed—identifying non-essential tasks and how they can be redistributed.
5) Task criticality vs. what happens if left undone: the authors did not explicitly ask what would happen if a task were left undone, though they used judgments of criticality. There is a need for a closed-ended format or alternative approach to reliably assess the consequences of omitting a task.
6) Subtleties of teamwork and organizational fit: job analysis often misses subtler aspects like team integration or company image; questions arise about social competence and nonstandard accommodations (e.g., placing someone with severe facial disfigurement in a public-contact role). Future research should develop techniques that incorporate these subtle dimensions when evaluating essential functions and accommodations.

Practical Implications and Real-World Relevance

ADA compliance requires a shift from exclusionary screening toward inclusive evaluation and accommodations; this includes a case-by-case approach to testing and job performance modifications.
Effective communication of essential functions through written job descriptions remains central for compliance, litigation risk management, and HR decision-making.
Organizations should consider policy development and documentation (policy on essential functions and reasonable accommodations) to reduce legal disputes and to improve consistency in hiring decisions.
Decision-analytic approaches offer a structured means to develop shared policies with disability stakeholder groups and to articulate the basis for essential-function determinations.
Validation and measurement considerations in testing with accommodations emphasize maintaining psychometric properties and ensuring fairness across disabled and nondisabled groups; evolving standards and research can guide best practices for validating modified assessments.

Takeaways for Exam Preparation

Essential functions are the core determinant of who is a qualified applicant under the ADA, and they must be identified through thoughtful job analysis informed by policy decisions.
The ADA encourages a select-in approach, emphasizing accommodation and inclusion, while balancing business necessity and safety concerns.
Job analysis under the ADA must consider multiple sources of evidence (managerial judgments, written descriptions, time spent, consequences of errors, staffing constraints, and organizational context).
A combination of work-oriented and worker-oriented analysis is useful for describing jobs in ADA context, but limits exist in predicting accommodations for the wide variability of disabilities.
Accommodation planning should be case-by-case rather than relying on rigid matrices that attempt to map disabilities to fixed job functions.
Validation of tests and accommodations requires careful consideration of psychometric properties and the potential impact of modifications on reliability and validity; group-level inferences may be limited for disabled populations.
Future research directions include improving methods to track job changes, automate accommodation planning, and incorporate nuanced outcomes related to team fit and organizational culture in ADA-friendly job analyses.

Key Concepts and Terms (Glossary)

Essential functions: the fundamental job duties that are necessary for job performance; central to determining if a person with a disability can perform the job with or without reasonable accommodation.
Reasonable accommodation: adjustments or modifications enabling a qualified individual with a disability to participate in the hiring process or perform the job; not to impose undue hardship on the employer.
Undue hardship: significant difficulty or expense relative to the size, resources, and operating context of the employer; factors include cost, resources, staffing, and impact on other employees.
Case-by-case determination: assessing disability accommodations and essential functions for each individual and position, rather than applying broad group-based rules.
Policy capturing: a methodological approach to elicit and quantify management judgments to form explicit organizational policies.
Task inventory: a catalog of tasks identified for a job, used to determine essential tasks and inform job descriptions.
Sensory and motor requirements: physical capabilities required for tasks (e.g., twisting, grip strength); notes about accommodations often caution against rigid tabulations due to the availability of alternative methods.
KSAs (Knowledge, Skills, Abilities): attributes used in assessing job performance; under ADA, KSAs may be inferred from essential functions but linking per-task KSAs to overall ability can be challenging.
Validity and reliability in testing: psychometric properties evaluated when tests are modified for accommodations; changes to test medium may affect measurement properties; evidence is needed to support maintained validity.
Disparate impact/adverse impact: traditionally examined with group data; ADA emphasizes individualized assessment, which complicates reliance on group-based adverse-impact analyses.

References (Selected Works Cited)

Equal Employment Opportunity Commission. 1992. A Technical Assistance Manual on the Employment Provisions (Title I) of the Americans with Disabilities Act.
Taylor, H. C., & Russell, J. T. 1939. The Relationship of Validity Coefficients to the Practical Effectiveness of Tests in Selection: Tables and Discussion. J. Appl. Psych.
Sparks, C. P. 1988. Legal Basis for Job Analysis. In The Job Analysis Handbook for Business, Industry, and Government (Vol. 1).
Thompson, D. E., & Thompson, T. A. 1982. Court Standards for Job Analysis in Test Validation. Personnel Psychology.
Kochhar, D. S., & Armstrong, T. J. 1988. Designing Jobs for Handicapped Employees. In The Job Analysis Handbook (Vol. 1).
Levine, E. L., Thomas, J. N., & Sistrunk, F. 1988. Selecting a Job Analysis Approach. In The Job Analysis Handbook (Vol. 1).
Campion, M. A. 1988; Campion, M. A., & Thayer, P. W. 1985. Interdisciplinary Approaches to Job Design; Development and Field Evaluation of an Interdisciplinary Measure of Job Design.
Nester, M. A. 1984; Nester, M. A., & Sapinkopf, R. C. 1982. Employment Testing for Handicapped People; A Federal Employment Test Modified for Deaf Applicants.
Heaton, S. M., Nelson, A. V., & Nester, M. A. 1980. Guide for Administering Examinations to Handicapped Individuals for Employment Purposes.
Hedging in the discussion: Keeney, K. L., & Raiffa, H. 1976. Decisions with Multiple Objectives: Preferences and Value Tradeoffs.
Hollenbeck, J. R., Ilgen, D. R., & Crampton, S. M. 1992. Lower Back Disability in Occupational Settings: A Review of the Literature from a Human Resource Management View.
Kept as references in the article; context for ADA-related job analysis and testing research.