Chapter 9 — Analyzing Jobs and Work

Chapter 9 opens with eight learning goals, numbered 9.1 through 9.8. As with every chapter in this book, the numbering scheme matters — instructors and quizzes sometimes reference goal numbers directly.

1. 9.1 Explain what job or work analysis is and the many purposes for which it is used.
2. 9.2 Distinguish the terms task, duty, position, job, and job family from each other.
3. 9.3 Distinguish work-oriented from worker-oriented descriptors.
4. 9.4 Identify seven key choices that exist in the analysis of work.
5. 9.5 Establish legally defensible minimum qualifications.
6. 9.6 Construct appropriate interview questions for analyzing work.
7. 9.7 Know when to use personality-based job analysis.
8. 9.8 Compare and contrast job or work analysis to competency modeling.

The chapter opens with a provocation. More than two decades ago, Bridges (1994a, 1994b) proclaimed "The End of the Job," arguing that organizing work into jobs "is a social artifact that has outlived its usefulness." Bridges's logic cascaded downward: if jobs are obsolete, so are positions ("too fixed"), roles ("too unitary, single purposed"), and even skills and competencies (they "will become too obsolete"). His prediction was that post-job workers would become self-employed contract workers hired onto projects and teams — citing firms like Intel and Microsoft that organize work around projects, with people working 6 to 10 projects at once, sometimes for different employers simultaneously.

Cascio and Aguinis push back with a simple empirical observation: go to any company's website, and it invites applications — for jobs. Some employees do work on 6 to 10 projects at once, but typically for only one employer. This is not to say work is static — fluid, competitive organizations constantly force adaptation to strategic initiatives like empowerment, reengineering, automation, intranet-based self-service HR, high-performance work practices that push authority downward, and alternative arrangements such as virtual teams and telework. Technologies like wireless communication, e-mail, and teleconferencing have made the "anytime, anywhere" workplace real.

Machine learning, automation, and the Internet are genuinely changing how work gets done, with more activities performed by intelligent systems rather than humans. Yet the job, as a way to organize and group tasks and responsibilities, has not disappeared — especially in large organizations. In the age of intelligent automation, organizations are instead reexamining what a "job" actually is: how it should be structured, reconfigured, and perhaps redefined, incorporating multilevel factors into job design. They ask probing questions: How should companies rethink the value of a job given machine-driven performance gains? What skills should they invest in? Which jobs should stay in-house versus be accessed via talent platforms or shared with peers or competitors? Answers to these questions, the chapter argues, lie in a process known as work analysis.

Work analysis is a broad term referring to any systematic process for gathering, documenting, and analyzing three features of work: (1) its content — tasks, responsibilities, or outputs; (2) worker attributes related to performance — knowledge, skills, abilities, or other personal characteristics (KSAOs); and (3) the context in which the work is performed — physical and psychological conditions.

The chapter treats "job analysis" and "work analysis" as applicable to changing structures of work — using either term is not meant to imply a focus on rigidly prescribed jobs. Conducted thoroughly and competently, work analysis provides a deeper understanding of individual jobs and their behavioral requirements, creating a firm basis for employment decisions.

The chapter grounds this in professional standards. The APA Standards (Standard 11.7) state that when empirical evidence of predictor-criterion relationships supports test use, systematic collection of job information should inform criterion-measure development — but there is no single preferred job analysis method, and job analysis is not limited to direct observation; large-scale job-analytic databases can also provide useful information. The SIOP Principles add that "analysis of work" now subsumes traditional job analysis methods plus competency modeling and other information about the work, worker, organization, and environment. There is no single preferred method — the right analysis depends on the nature of the work, available information, the organizational setting, the workers, and the purpose of the study. The central point is obtaining reliable, relevant job information addressing anticipated behaviors, activities, KSAOs, or competencies.

Courts vary in how much work analysis detail they require: some insist on extensive analysis (e.g., as content-related evidence of validity), while other purposes, such as validity generalization, may not require the same level of detail. Work analysis information can be collected for many uses and purposes (Figure 9.1 in the text catalogs these), and it can underpin an organization's structure and design by clarifying roles — patterns of expected behavior tied to organizational position — from mail clerk to chairperson of the board, avoiding overlap and duplication of effort and promoting efficiency and harmony.

Talent management, like any specialty, has its own jargon. Although some terms are used interchangeably in casual conversation, technically there are distinct differences among them — and this is one of the chapter's most quizzable sections (learning goal 9.2). The definitions build on each other in a strict hierarchy, from the smallest unit of work up to an entire working life.

• Element — the smallest unit into which work can be divided without analyzing the separate motions, movements, and mental processes involved. Example: removing a saw from a tool chest prior to sawing wood.
• Task — a distinct work activity carried out for a distinct purpose. Examples: running a computer program, cleaning cooking utensils, unloading a truckload of freight.
• Duty — a large segment of the work performed by an individual, comprising any number of tasks. Examples: conducting interviews, counseling employees, providing information to the public.
• Position — one or more duties performed by a given individual in a given firm at a given time (e.g., clerk typist–level three). There are as many positions as there are workers.
• Job — a group of positions that are similar in their significant duties (e.g., two or more mechanics–level two). A job may involve only one position depending on organization size — a local garage may employ only one mechanic–level two.
• Job family — a group of two or more jobs that either call for similar worker characteristics or contain parallel work tasks as determined by job analysis.
• Occupation — a group of similar jobs found in different organizations at different times (e.g., electricians, machinists). A vocation is similar but is a term more likely used by a worker than an employer.
• Career — a sequence of positions, jobs, or occupations that one person engages in during their working life.

A wide variety of methods exist for collecting information about jobs and work, and they vary on multiple dimensions, which creates choices. The chapter's governing rule: methods for analyzing work must align with the purpose for which the information is collected. It is not true that a single type of work analysis data can support any talent management activity — the level of detail needed to build a pay structure (job evaluation) is usually not detailed enough for a human-factors engineer redesigning a person–machine interface. First define the purpose, then choose a method that fits.

The chapter then names seven choices confronting the work analyst — though the range narrows once the specific purpose is identified. These map directly to learning goal 9.4 and are a near-certain glossary/quiz target.

1. Activities or attributes? Work-oriented techniques focus on what gets done (tasks); worker-oriented techniques focus on how the work gets done (KSAOs — personality, values, attitudes). Competency modeling incorporates the full range of KSAOs.
2. General or specific? This concerns the level of detail needed. A brief description for pay-survey comparisons needs far less detail than developing preemployment assessments based on critical KSAOs.
3. Qualitative or quantitative? Work can be described narratively (qualitatively) or via numerical ratings on fixed scales — time, frequency, importance, criticality (quantitatively). Qualitative methods suit career planning; comparing different types of work requires quantitative data.
4. Taxonomy based or blank slate? Taxonomy-based approaches (the PAQ and F-JAS) apply relatively general work activities across a broad range of work. Blank-slate approaches — trained observers or incumbents developing lists specific to a job — allow greater detail than taxonomy approaches.
5. Observers or incumbents and supervisors? Trained analysts may observe work directly and distill qualitative or quantitative judgments. Alternatively, incumbents and supervisors may identify and rate activities or attributes themselves; with enough respondents, this allows checking rating consistency and identifying response-pattern clusters.
6. Single-job or multiple-job comparison? Sometimes the focus is one specific job (e.g., an entry-level test for bank tellers); other times the focus is documenting similarities and differences across jobs — to justify a shared selection system, to justify using a system for the same job across organizations, or to build job families and career paths.
7. Descriptive or prescriptive? Work analysis typically describes a job as it currently exists. When a job doesn't yet exist, analysts must prescribe activities or attributes for the soon-to-be-created job — this approach is called strategic work analysis, covered later in the chapter.

Work analysis consists of defining work in terms of its component tasks, specifying what employee behaviors are necessary to perform those tasks, and then developing hypotheses about the personal characteristics needed to perform those behaviors. Two elements stand out: task requirements and people requirements — this section covers the former.

Many characteristics of work are "givens" to employees: the equipment used, the arrangement of the work space, the division of labor, and the procedures, methods, and standards of performance. From this information, the analyst produces a job description — a written statement of what a worker actually does, how, and why. Job descriptions serve several practical purposes: they tell employees what tasks they're expected to perform, they aid the interactive process some disability-accommodation laws require, they describe minimum qualifications, and they help justify exempt versus nonexempt status under wage and hour laws.

The Five Elements of a Traditional Job Description

• Job title — for internal bookkeeping and reporting to government agencies.
• Activities and procedures — descriptions of tasks performed, materials used, machinery operated, formal interactions with other workers, and the nature/extent of supervision given or received.
• Working conditions and physical environment — heat, lighting, noise level, indoor/outdoor setting, physical location, hazardous conditions.
• Social environment — e.g., the number of individuals in the work group and the amount of interpersonal interaction required.
• Conditions of employment — hours of work, wage structure, method of payment, benefits, place in the formal organization, and opportunities for promotion and transfer.

Behavioral Job Descriptions — the Modern Alternative

Some organizations are moving toward behavioral job descriptions: broader abilities that are easier to alter as technologies and customer needs change. Instead of listing narrow skills like "writing, speaking, and making presentations," a behavioral job description uses broader statements such as "actively listens, builds trust, and adapts his or her style and tactics to fit the audience." These behaviors don't change even as the means of executing them evolve with technology. Workers today are increasingly expected to draw inferences and render diagnoses, judgments, and decisions — often under severe time constraints — rather than execute simple, predictable procedures.

Job Specifications

Job specifications represent the KSAOs deemed necessary to perform a type of work — for example, keen vision (usually 20/20 uncorrected) required of astronauts and test pilots. In many cases, though, job specifications are not rigid; they serve as guidelines for recruitment, selection, and placement, depending on the acceptable performance level and the degree to which some abilities can substitute for others.

Job specifications identify personal characteristics — educational background, experience, training — valid for screening, selection, and placement. The chapter's central question: how are minimal qualifications (MQs) properly set?

The Levine, May, Ulm, and Gordon (1997) Methodology

Levine et al. (1997) developed a methodology for determining MQs in the context of a court case challenging the use of MQs of unknown validity but high adverse impact. Because a court ultimately approved it, and because it aligns with sound professional practice, the chapter walks through it step by step.

1. Working independently with a draft list of tasks and KSAs for a target job, separate groups of subject matter experts (SMEs) rate tasks and KSAs on a set of four scales. Because ratings are aggregated as means or percentages, there is no need for SME consensus.
2. Tasks and KSAs meeting the criteria on those scales form the domains of tasks and KSAs from which MQs are derived. SMEs then suggest appropriate types or amounts of education, work experience, and other qualifications.
3. Work analysts prepare a draft set of MQ profiles — each a statement of education, training, or experience presumably needed to perform the target job at a satisfactory level.
4. A new set of SMEs is convened to: (a) establish a description of a barely acceptable employee; (b) decide if the MQ profile list is complete or needs editing; and (c) rate the finalized profiles on two scales — level and clarity.
5. Profiles meeting the level and clarity criteria are linked back to the tasks and KSAs in the established domains, using two further scales (one for tasks, one for KSAs). Each profile must meet the linkage criterion to demonstrate content-oriented evidence of validity. In Levine et al.'s (1997) own study, six of the nine MQ profiles met this bar.

Buster, Roth, and Bobko (2005) — Eight Recommendations for Practice

Buster et al. (2005) presented a related, federal-court-approved method for developing content-oriented evidence of validity specifically for education- and experience-based MQs. Their eight practice recommendations:

1. Begin with a structured analysis of work that identifies critical tasks and KSAs, noting which KSAs are needed on day one of the work (entry-level KSAs).
2. Distribute a list of tasks and KSAs associated with the work at the first MQ-development meeting.
3. Emphasize that the point of reference for the MQs is an individual newly appointed to the work.
4. Instruct those generating potential MQs to think about alternative MQs (e.g., a professional certification).
5. Use straightforward, targeted MQs, since they can be rated more easily and reliably.
6. Have SMEs rate the list of MQs independently.
7. Have SMEs link all potential MQs back to KSAs or tasks.
8. Bracket potential MQs with both easier and more difficult statements.

A meta-analysis of 46 studies and 299 reliability estimates examined average inter- and intrarater reliability of job analysis ratings. Interrater reliability is the degree to which different raters agree on the components of a target work role or job. Intrarater reliability measures stability — rate-rerate consistency for the same work at different times.

Across 119 studies, task data showed higher inter- and intrarater reliabilities than generalized work activity data (.77 versus .60, and .72 versus .58, respectively). Analysts showed the highest interrater reliability and incumbents the lowest, regardless of data specificity. Within task data, scales measuring relative value (importance, difficulty) had similar, relatively high interrater reliability; scales involving temporal judgments (frequency, time-spent) had similar, relatively low interrater reliability (Dierdorff & Wilson, 2003).

Job descriptions are valid to the extent they accurately represent job content, environment, and employment conditions. Job specifications are valid to the extent that people possessing the believed-necessary characteristics actually perform more effectively than those lacking them.

Sixteen Sources of Inaccuracy — Social and Cognitive

Morgeson and Campion (1997) noted that many job analysis processes rest on fallible human judgment, and identified 16 potential sources of inaccuracy falling into two categories. Social sources apply mainly where groups, rather than individuals, judge the analysis of work — pressures to conform can distort results when group consensus is required. Cognitive sources reflect our limited capacity to process information — demand for large numbers of ratings, or very fine distinctions, can cause information overload. These sources are more likely to distort ratings of subjective, diffuse attributes (many KSAOs) than ratings of discrete, observable tasks — "Do you do this at work?" requires far less subjectivity than rating "criticality."

In a later study, Morgeson, Delaney-Klinger, Mayfield, Ferrara, and Campion (2004) examined self-presentation bias — an attempt by individuals to control the impression others form of them. Self-presentation may inflate ratings, particularly for ability statements.

The amount of descriptive information available to raters also significantly affects accuracy. Student raters given more detailed work information were consistently more accurate (relative to averaged incumbent ratings) than those given only a job title. Data from raters naive about the work showed little agreement with data from work-content experts (Harvey & Lozada-Larsen, 1988).

Numerous methods exist for describing jobs and work, differing in their assumptions, breadth of coverage, and precision. Some are work oriented, some worker oriented. In practice, several methods should be combined so the end product is a valid, comprehensive picture of job duties, responsibilities, and behaviors.

Direct Observation and Job Performance

Observation of incumbents and the analyst's own performance of the work are two data-gathering methods, recorded narratively or on a checklist/worksheet. Both assume jobs are relatively static. Job observation suits jobs requiring manual, standardized, short-cycle activities; job performance suits jobs the analyst can learn readily. Observations must include a representative sample of behaviors — an eight-hour observation of nurses tending sleepy postoperative patients may say little about how nurses "cope with emergencies," even though that activity is crucial to the job.

The analyst must be unobtrusive without hiding — the measuring process itself should not distort what's being measured. The chapter's cautionary anecdote: an analyst riding along with a police officer jumped out of the patrol car during a robbery-in-progress call and accidentally positioned himself between the robbers and the police — the robbers used him as a shield to escape (they were apprehended later).

Observation and job performance are inappropriate for highly mental, concentration-heavy work (lawyer, computer programmer, architect) but work well for many other jobs. Functional job analysis (FJA) is a technique for recording observed tasks, identifying exactly what the worker does and the results of that behavior — what gets done. An FJA worksheet captures a firefighter's salvage-and-overhaul operations, for example, under headings for WHAT (performs what action, to whom or what), WHY (to produce or achieve what), and HOW (tools, equipment, instructions, and whether the task is prescribed or left to discretion).

Under WORKER FUNCTIONS, the analyst describes the orientation and level of activity with data, people, and things — every job involves some mix of information/ideas (data), clients/coworkers/superiors (people), and machines/equipment (things), with numeric complexity scales developed by the U.S. Department of Labor for each.

Because these level and orientation scales apply to all tasks, they let analysts compare every task and every type of work on a common basis. High-resolution digital cameras offer one solution to the "unobtrusive observer" problem — video can be reviewed and coded offline, following an employee through a facility as if walking behind them. But video has real limits: you can't hear what employees say, can't track interactions among employees, ceiling-mounted cameras can't tell what a person is looking at, and an employee stepping out of frame leaves a gap unless multiple cameras cover overlapping fields.

Interview

The interview is probably the most commonly used technique for establishing tasks, duties, and behaviors — for standardized and nonstandardized activities, physical and mental work alike. Because the worker acts as their own observer, they can report activities that wouldn't otherwise be observed, including those spanning long time periods, and information unavailable from any other source. Viewed as a "conversation with a purpose," interview success depends partly on interviewer skill.

The chapter (citing McCormick, 1979) lists six criteria for checking interview question appropriateness — directly relevant to learning goal 9.6:

• The question should be related to the purpose of the analysis.
• The wording should be clear and unambiguous.
• The question should not "lead" the respondent — it should not imply a specific answer is desired.
• The question should not be "loaded" such that one response seems more socially desirable than another.
• The question should not ask for knowledge or information the interviewee doesn't have.
• There should be no personal or intimate material the interviewee might resent.

Workers often view interviewers with suspicion and are wary of divulging job information, so the analyst should create a comfortable atmosphere for open discussion. The major stumbling block is distortion — whether from outright falsification or honest misunderstanding. If a worker believes the analysis may influence wages, they may exaggerate some responsibilities and minimize others, so interviews require time and adroit questioning to elicit valid information.

As a check, it's wise to interview several incumbents plus immediate supervisors who know the work well. Both high- and low-performing incumbents and supervisors tend to give similar information, as do members of different demographic subgroups — though this may hold only for simple, not complex, jobs. Multiple interviews account for features of work made dynamic by time, people, and situations, but piecing together dissimilar interview results into one comprehensive picture remains difficult — additional techniques often supplement interviewing.

SME Panels

Panels of 6 to 10 subject matter experts (SMEs) are convened for two main purposes: (a) developing information on tasks or KSAOs for constructing work analysis questionnaires, and (b) in test development, establishing linkages between tasks and KSAOs, KSAOs and test items, and tasks and test items. The full SME group usually represents a 10–20% sample of incumbents and supervisors, representative of race, gender, location, shift, and assignment composition. A key SME attribute is experience — failing to include a broad cross-section of experience can distort ratings. Representative SME panels produce results very similar to broad field surveys.

SMEs are encouraged to discuss issues and resolve disagreements openly, prompted by questions such as: "Think of workers you know who are better than anyone else at (a particular task). Why do they do so well?"; "What kinds of KSAOs would you want a worker assigned to this task to have?"; "What do you expect workers to learn in training that would make them effective?"; "What KSAOs distinguish good workers from poor ones?"

When SMEs establish linkages for test development, quality-control statistics confirm the meaningfulness of their judgments — for example, questionnaires may include repeat items and "carelessness" items (items inappropriate for the job under study). High interrater agreement plus near-zero endorsement of carelessness items are important data-quality checks.

Questionnaires, Task Inventories, and Checklists

Questionnaires are usually standardized, requiring respondents to check applicable items or rate items' relevance to the job. They are cheaper and quicker than other methods and can be completed at the respondent's leisure (online or on paper), avoiding lost production time. With many workers in each job, questionnaires give breadth of coverage that would be exorbitantly expensive to obtain otherwise.

Drawbacks: questionnaires are time-consuming and expensive to develop, ambiguities that an interview could clarify go uncorrected, follow-up is hard, and the impersonal format loses the rapport of face-to-face contact — which can hurt respondent cooperation and motivation. On the plus side, the structured-questionnaire approach lets work analysis information be quantified and statistically analyzed.

Task inventories and checklists are questionnaires that collect information about a particular type of work — an analyst checks or rates each task/activity item for importance, frequency, difficulty, time to learn, or relationship to overall performance. These data are computer-analyzable, but checklists tend to ignore task sequencing and relationships to other jobs, making an overall perspective hard to obtain from checklist data alone. If the goal is assessing relative task importance, a unit-weighted, additive composite of ratings on task criticality, difficulty of learning, and relative time spent may best predict average task importance across SMEs.

The Position Analysis Questionnaire (PAQ)

Task inventories are work oriented and make static assumptions about jobs, so behavioral implications are hard to establish. Worker-oriented information, by contrast, describes how a job gets done and focuses on generalized worker behaviors. The Position Analysis Questionnaire (PAQ) is a worker-oriented instrument based on statistical analyses of job elements, well suited to quantitative analysis. It consists of 194 items across five categories: information input (where/how the worker gets needed information); mental processes (reasoning, planning, decision-making); work output (physical activities and tools/devices used); relationships with other persons; and work context (physical/social conditions).

The PAQ's average item reliability is a respectable .80 (similar results with a German-language form). A meta-analysis of 83 PAQ studies found average interrater reliability of .66 and average intrarater reliability of .82. Personal and organizational factors have little effect on PAQ results — similar profiles emerge regardless of analyst gender, whether the incumbent portrayed the work as interesting, or how much information was presented. However, as with other methods, PAQ ratings from expert versus job-naive raters are not equivalent, and there are no shortcuts: rating each PAQ dimension as a whole, versus rating individual items and combining them, produces near-zero convergence.

McCormick et al. (1972) argue structured, worker-oriented instruments hold real potential for finding common denominators linking dissimilar jobs — you cannot compare butchering, baking, and candlestick-making in strictly technological terms, but their commonalities may emerge when analyzed via generalized human behaviors (worker-oriented elements).

The Fleishman Job Analysis Survey (F-JAS)

The Fleishman Job Analysis Survey (F-JAS) is one of the most thoroughly researched work analysis approaches, describing work in terms of the abilities required to perform it. Its ability-requirements taxonomy aims to capture the fewest independent ability categories describing performance across the widest range of tasks: 21 cognitive abilities (e.g., oral comprehension, deductive reasoning, number facility), 10 psychomotor abilities (e.g., reaction time, control precision, finger dexterity), 9 physical abilities (e.g., gross body coordination, static strength, stamina), 12 sensory/perceptual abilities (e.g., depth perception, visual color discrimination, hearing sensitivity), and 21 social/interpersonal abilities (e.g., persuasion, dependability, social sensitivity). The methodology further identifies 33 types of general occupational knowledge and skill requirements (e.g., customer and personal services, administration and management, building and construction).

Rating scales define each ability, distinguish it from related abilities, and provide task examples at different ability levels — the chapter's example is cognitive ability 10, "Number Facility." Interrater reliabilities for F-JAS scales generally run in the mid-.80s, with considerable construct and predictive validity evidence supporting the scales. The companion Handbook of Human Abilities integrates ability definitions with the kinds of tasks/jobs that require each ability and published tests that measure it.

The Critical-Incidents Technique

This is the same method discussed earlier in connection with performance management (Chapter 5). The critical-incidents approach collects anecdotes of work behavior — from supervisors, employees, or others familiar with the job — describing especially good or especially poor performance. Each anecdote captures: (a) what led up to the incident and its context; (b) exactly what the individual did that was effective or ineffective; (c) the perceived consequences of the behavior; and (d) whether those consequences were within the employee's control.

An analyst typically gathers a broad sample of observations across many employees — depending on the work, hundreds or thousands of incidents may be needed to adequately cover the behavioral domain. Incidents are then categorized by the job dimensions they represent and assembled into a checklist format, and in their entirety provide a composite picture of a job's behavioral requirements.

Other Supplementary Sources

Additional sources can usefully supplement the methods above: examining training materials (manuals, standard operating procedures, equipment blueprints) can reveal the skills, abilities, and behaviors required to learn the work and operate essential equipment. Technical conferences of experts selected for broad knowledge and experience, and diaries in which incumbents record their tasks day by day, are also useful.

The Job Analysis Wizard (JAW)

The Job Analysis Wizard (JAW), developed at Lucent Technologies, is a Web-based tool capitalizing on computer technology and sophisticated search-and-retrieval methods. Its features:

• Thousands of elements organized into broader work- and worker-related dimensions — e.g., "work requirements" (with second-level work context, generalized work behaviors, tools and equipment) and "worker requirements" (with second-level abilities, knowledge, skills, education, certifications, languages, work styles).
• Fuzzy logic as a decision aid for placing new dimensions (e.g., emerging technologies) into the JAW taxonomy — comparing quantitative ratings of a new element against existing patterns to recommend placement near similar elements (e.g., a new programming language placed near other programming languages).
• Full automation of the work analysis process, with the ability to surface information on past products supporting business initiatives.
• Electronic surveys completed by incumbents, supervisors, and other SMEs anywhere with access to the internal website.
• Statistical filtering that creates linkage-matrix surveys linking key work dimensions (tasks, tools, equipment) to worker dimensions (knowledge, skills).
• High-quality graphic reports for easy interpretation, plus the ability to upload completed results for others to reuse and immediately identify preexisting materials (tests, interviews) relevant to a job.

Personality is the set of characteristics accounting for the consistent ways a person responds to situations. Interest in personality as a work-performance determinant has revived in recent years, driven by demonstrated positive relationships between personality characteristics and performance in some contexts. Although there is controversy about personality's value-added contribution relative to other predictors, some traits — conscientiousness in particular — are valid predictors across many occupations.

Personality-based job analysis (PBJA) is especially useful for cross-functional, difficult-to-define work that can't be described via simple tasks or discrete KSAs — work increasingly common in 21st-century organizations.

The PPRF and the Big Five

Perhaps the most credible peer-reviewed PBJA tool in the public domain is the Personality-Related Position Requirements Form (PPRF) (Raymark, Schmit, & Guion, 1997) — a worker-oriented method assessing the extent to which each Big Five personality trait is needed for a particular job. The Big Five is the most established, thoroughly researched personality taxonomy in work settings, comprising:

• Neuroticism — the degree to which someone is insecure, anxious, depressed, and emotional versus calm, self-confident, and cool.
• Extraversion — the degree to which someone is gregarious, assertive, and sociable versus reserved, timid, and quiet.
• Openness to experience — the degree to which someone is creative, curious, and cultured versus practical with narrow interests.
• Agreeableness — the degree to which someone is cooperative, warm, and agreeable versus cold, disagreeable, and antagonistic.
• Conscientiousness — the degree to which someone is hardworking, organized, dependable, and persevering versus lazy, disorganized, and unreliable.

The PPRF consists of behavioral indicators tied to the five traits; respondents (typically incumbents) rate how relevant each indicator is to the job, and averaged scores across respondents indicate how relevant each trait (or subdimension) is to that job.

Strategic or Future-Oriented Work Analysis (SWA)

Sometimes organizations need skill and ability information for jobs or positions that don't yet exist — jobs tied to new technology expected in three to five years, new plant start-ups with unusual work organization (e.g., "co-bots" in manufacturing), or reconfiguring existing work into process-based structures (e.g., credit issuance, procurement). Given the scale of change already underway in the world of work, strategic work analysis (SWA) is becoming ever more important. Competency models (discussed next) are inherently future oriented, but standard work analysis methods can also be adapted for this purpose.

Brannick et al. (2017) call for reconceptualizing work analysis as a strategic tool — requiring collaboration between those with a "micro" view of jobs and work (I/O psychologists, training specialists, industrial engineers) and those with a "macro" perspective (labor economists, sociologists, management consultants). This reframing underlies SWA: a systematic effort to identify or define current or anticipated work/worker requirements strategically aligned to an organization's mission and goals — a concept broad enough to subsume other terms in use, such as future-oriented job analysis, strategic job analysis, and competency modeling.

The chapter's worked example is Landis, Fogli, and Goldberg (1998): a large insurance company condensed 11 existing jobs into 3 new ones and hired consultants to develop valid selection tests for the new work. The consultants recognized at least three different perspectives on the new work — the organization's steering committee, an outside firm handling technological change (e.g., updated computer systems), and current organization members (supervisors, incumbents doing similar work, system-design experts, training coordinators) — and used SMEs from each group throughout the work-analysis procedure. This surfaced and addressed technology, work-design, and training changes early, and scheduled meetings gave the steering committee important feedback and early warning about employee concerns.

Work Analysis for Star Performers

In many types of organizations — academia, sports, financial services, entertainment, knowledge-based work — some individuals stand out, way out, in the quality or quantity of their performance relative to peers. Brannick et al. (2017) suggest two promising approaches to analyzing what star performers do. First, reverse the usual task-analysis procedure: instead of analyzing tasks and then comparing better versus poorer performers, begin with groups of better and poorer performers and systematically explore differences in what they do, how they do it, and what personal qualities distinguish the two groups.

Second, partner I/O psychologists or HR professionals with SMEs or professionals from other fields, such as economics or operations management, to broaden and deepen understanding of what drives high performance (e.g., organizational support quality, social network strength). Incorporated into a statistical model alongside individual-difference variables like ability, personality, or emotional intelligence, this broader predictor set can improve forecasts of successful performance.

Cognitive task analysis (CTA) is particularly appropriate for knowledge-based work — surgeons, financial analysts, scientific researchers. Its objective is to bridge the gap between what gets done and how it gets done, focusing on the cognitive skills or mental demands needed to perform a task proficiently, often by comparing novices to experts.

CTA involves three steps: (1) selecting participants (novices and experts); (2) eliciting knowledge, using any of more than 100 distinct methods; and (3) analyzing and representing that knowledge. The most common knowledge-elicitation methods in the workplace are structured interviews and think-aloud protocols (process tracing) while the person performs a task or solves a problem.

To be useful in employment settings, CTA requires explicit linkage of cognitive processes to required job knowledge, skill, or other characteristics. At the same time, it is laborious, time-consuming, and expensive — the chapter's guidance is that it makes the most sense where work requires effective performance under time pressure and where the consequences of error are severe.

From the Dictionary of Occupational Titles to O*NET

The U.S. Department of Labor published the Dictionary of Occupational Titles (DOT) in the 1930s to help the new public employment system link skill supply and skill demand during the Great Depression. The last DOT version (1991) described more than 12,000 jobs, but it lacked a cross-job organizing structure for comparing jobs, and its task-only focus didn't indicate what personal characteristics workers needed or the context in which the work occurred.

To fix this, the Department of Labor sponsored the Occupational Information Network (O*NET), incorporating decades of accumulated job/work information into a national occupational information system built on four design principles: (1) multiple descriptor domains providing "multiple windows" into the world of work; (2) a common language of work and worker descriptors spanning the entire spectrum of occupations; (3) descriptions based on a broad-to-specific taxonomy; and (4) a comprehensive content model integrating the first three principles.

• Multiple windows — multiple descriptor domains (tasks, abilities, skills, knowledge areas, work context) let users work with whichever descriptors suit their question, including how specific skills relate to different work activities.
• Common language — because job-specific information changes rapidly, O*NET uses more stable general descriptors (e.g., generalized work activities like "selling or influencing others" and "assisting or caring for others"), within which job-specific detail still fits.
• Taxonomies and hierarchies — hierarchical occupational classification lets information be summarized into fewer categories, spanning descriptors from specific skills to broader organizational/contextual factors like organizational climate.

The O*NET content model integrates these three principles into six major domains of cross-job descriptors, held in a relational database publicly accessible at www.onetonline.org. The system works in both directions: start with a skill/ability profile and find matching occupations, or start with an occupation and find others with similar characteristics. O*NET covers 974 occupations across the U.S. economy, and studies of adult literacy requirements and incumbent task/activity/knowledge/skill ratings provide construct-oriented validity evidence generally favorable to the system.

Competency Models

Competency models attempt to identify variables related to overall organizational fit and personality characteristics consistent with the organization's vision (e.g., drive for results, persistence, innovation, flexibility), written in terms operating managers can relate to. They are a form of work analysis focused on broader individual characteristics — the full range of KSAOs (motives, traits, attitudes, personality) needed for effective performance and characteristic of exceptional performers. Ideally, a competency model specifies a set of competencies necessary for successful performance, each with behavioral indicators tied to high performance.

Unfortunately, there is no consistent definition of "competency." Many competencies appearing in the literature and in competency models (e.g., "visioning") are ill-defined, with no clear meaning — a deficiency that transfers directly into any selection tool built on those constructs.

Competency Modeling vs. Job/Work Analysis — the Direct Comparison

This directly answers learning goal 9.8. A rigorous comparison found that competency approaches typically make a substantial effort to understand an organization's business context and competitive strategy, establishing a direct line-of-sight between individual competency requirements and broader organizational goals. Job or work analyses, by contrast, typically do not make this strategic connection — but their rigor and documentation more reliably withstand close legal scrutiny. As currently practiced, competency modeling is not a substitute or replacement for job or work analysis.

The unit of analysis for a competency model can range from a single job to an entire organization. When focused on a single job or job family, differences from traditional work analysis shrink. An organization-wide competency model is different in kind — it can reflect top managers' vision of what will be valued and rewarded in the future, functioning as part of an organizational-change effort. In that sense, competency modeling is more prescriptive/future oriented, while job or work analysis is more descriptive. Neither approach is singular — both vary widely in practice — and no single type of descriptor content (competencies, KSAOs, work activities, performance standards) suits every purpose; purpose remains the key consideration in choosing an approach.

Cascio and Aguinis close every chapter with an "Evidence-Based Implications for Practice" list. For Chapter 9, it functions as a checklist of what you should be able to state confidently after finishing the reading.

• When collecting work-related information, define clearly the purpose for collecting it first.
• Because every method for collecting such data carries offsetting advantages and disadvantages, choose multiple methods that best suit the identified purpose.
• If using panels of subject matter experts, include a broad cross-section of experience.
• If using interviews, include both incumbents and supervisors, and train interviewers in interviewing techniques.
• If using personality-based job analysis, incorporate frame-of-reference training.
• Recognize that competency models are not substitutes for job or work analysis — both provide useful information across the practice continuum.
• When establishing minimum qualifications for education or experience, assess content-oriented evidence of validity using the methods described in this chapter.

Chapter 9 ends with ten discussion questions. Below, each is paired with a concise model answer grounded directly in the chapter's content — useful both for self-testing and as a starting point if any of these questions resurface in a graded discussion or quiz.

1. Describe some of the choices that need to be made in deciding how to analyze jobs and work. How would you choose an appropriate technique in a given situation?

The analyst must decide across seven dimensions: activities versus attributes (work oriented versus worker oriented), general versus specific detail, qualitative versus quantitative data, taxonomy-based versus blank-slate approaches, observers versus incumbents/supervisors as the information source, single-job versus multiple-job comparison, and descriptive versus prescriptive (strategic) analysis. Choosing the right technique starts by defining the purpose of the analysis — a pay-structure comparison needs far less detail than a preemployment test built on critical KSAOs — and then selecting the method(s) whose strengths match that purpose, since no single method serves every purpose well.

2. Develop an outline for a work analysis workshop with a panel of subject matter experts.

A workshop outline would open by defining the purpose of the analysis and selecting 6 to 10 SMEs representing a broad cross-section of experience, race, gender, location, shift, and assignment. It would proceed through generating and discussing task and KSA lists, prompting discussion with questions like "What distinguishes your best performers?", rating tasks/KSAs on the relevant scales, and — if linking tasks to test items — embedding quality-control checks such as repeat items and carelessness items to verify the meaningfulness of the SME judgments.

3. Your boss asks you to incorporate personality characteristics into work analysis. How would you proceed?

I would use a personality-based job analysis (PBJA) tool such as the Personality-Related Position Requirements Form (PPRF), which assesses how much each Big Five trait — neuroticism, extraversion, openness, agreeableness, conscientiousness — matters for the job, using behavioral indicators rated by incumbents. Critically, I would first run all raters through frame-of-reference training, since research shows PBJA ratings can otherwise be inflated by raters' own personality characteristics and by general rating-inflation bias.

4. What is strategic work analysis, and why might it subsume other terms that are currently in use?

Strategic work analysis (SWA) is a systematic effort to identify or define current or anticipated work or worker requirements that are strategically aligned with an organization's mission and goals, requiring collaboration between "micro"-level specialists (I/O psychologists, engineers) and "macro"-level specialists (economists, consultants). It subsumes terms like future-oriented job analysis, strategic job analysis, and competency modeling because all of these share the same underlying reframing — analyzing work not just as it exists today but as it should exist to serve future organizational strategy.

5. How might you conduct a future-oriented work analysis?

Following the Landis, Fogli, and Goldberg (1998) example, I would assemble SMEs representing every relevant perspective on the future work — organizational leadership, any outside technology partners driving the change, and current employees or systems experts closest to the work — and use them throughout the analysis rather than relying on a single viewpoint. I would schedule checkpoint meetings so leadership gets early warning about employee concerns, and I would treat the resulting profiles as prescriptive statements of what the not-yet-existing job should require, not descriptions of a job that already exists.

6. What are the similarities and differences between competency modeling and job or work analysis?

Both aim to specify what a job requires and what workers need to bring to it, and both are subject to wide variation in practice. They differ in orientation: competency modeling explicitly links individual requirements to organizational strategy and vision and is more prescriptive/future oriented, especially at the organization-wide level, while job or work analysis is more descriptive of work as it currently exists and carries the rigor and documentation needed to survive legal scrutiny. Competency modeling is not a substitute for job or work analysis — the two serve complementary purposes across the practice continuum.

7. You have been asked to conduct a work analysis for astronauts on the international space station. Which technique(s) might be most appropriate in this situation, and why?

Given the highly cognitive, high-stakes, time-pressured nature of astronaut work, cognitive task analysis (CTA) would be central, comparing novice and expert performance via structured interviews and think-aloud protocols to capture the mental demands behind observable actions. This should be paired with the Fleishman Job Analysis Survey (F-JAS) to specify the precise cognitive, psychomotor, physical, and sensory abilities required, since astronaut work draws heavily on all four ability categories and the F-JAS was built to make such abilities comparable and well-defined.

8. How might you study the work of star performers in the field of human resource management? In other words, how would you conduct a work analysis of "HR management stars"?

Following Brannick et al.'s (2017) two suggested approaches, I would first identify groups of clearly better- and poorer-performing HR professionals and systematically compare what each group does, how they do it, and what personal qualities distinguish them, rather than starting from a generic task analysis. Second, I would partner I/O psychologists with professionals from adjacent fields such as economics or operations management to capture contextual drivers of star performance — like organizational support or social network strength — and fold those, along with individual-difference variables like personality and emotional intelligence, into a broader predictive model.

9. When does it make sense to use cognitive task analysis?

Cognitive task analysis makes the most sense for knowledge-based work where the real work happens mentally and isn't visible through direct observation — surgeons, financial analysts, scientific researchers. The chapter's specific guidance is to reserve it for situations where the work requires effective performance under time pressure and where the consequences of error are severe, since CTA is laborious, time-consuming, and expensive relative to other work analysis methods.

10. Go to the O*NET website (www.onetonline.org). Develop a profile of five skills or abilities, and find occupations that match it.

This question requires hands-on use of O*NET's skills-search tool rather than a textbook-derived answer. A workable approach: choose five skills or abilities relevant to a role you know well (for example, active listening, coordination, complex problem solving, time management, and quality control analysis), enter them into O*NET's advanced/skills search, and record which occupations the system returns as matches — then check whether those results make intuitive sense given the skills selected, since that is exactly the kind of "start with a skill profile, find matching occupations" flexibility the O*NET content model is designed to provide.

Every bolded or explicitly defined term in Chapter 9, in one line each, in the order the chapter introduces them.

A single table capturing the chapter's terminology ladder, its seven analyst choices, its named methods, and its most quotable claims — everything you need to answer a cold-call question about Chapter 9 without re-reading it.