Read
Work Analysis
Work analysis is any systematic process for gathering, documenting, and analysing information about the work performed in an organisation. It covers three dimensions:
The Work
Tasks, responsibilities, and outputs — what is actually done in the role.
The Worker
Knowledge, skills, abilities, and other personal characteristics (KSAOs) required to perform it well.
The Context
Physical and psychological conditions in the immediate work environment, plus the broader organisational and external environment.
Why it matters for selection
In personnel selection, work analysis ensures that selection systems are directly grounded in the requirements of the role. This work-relatedness is what makes a selection system valid — meaning it genuinely predicts performance — and gives it practical utility for the organisation. It also makes selection decisions legally defensible, since they can be shown to be based on demonstrable job requirements rather than arbitrary criteria.
Four traditional selection-related applications
Work analysis serves four distinct purposes within personnel selection. Each addresses a different question about how to build or evaluate a selection system.
Predictor Development
What should we measure, and how?
This application involves two sequential phases: first identifying what attributes are needed to do the job, then selecting the right tools to measure those attributes.
Phase 1 — Inferring person requirements
Work analysis is used to make inferences about the KSAOs required for effective performance — not only which attributes matter, but at what level of proficiency they are needed. These inferences form the specification against which candidates will be assessed.
Phase 2 — Linking measures to KSAOs
Once the required KSAOs have been identified, appropriate assessment tools are matched to them — for example, a spatial reasoning test for a role requiring spatial ability, or a structured interview probing a specific skill. The work analysis findings justify why each measure was chosen.
See an example
Clinical psychologist role. Work analysis of a clinical psychologist role reveals that formulating diagnoses and treatment plans are central activities. From these, the analyst infers the required KSAOs: diagnostic reasoning ability, knowledge of psychometric assessment tools, and skill in communicating complex findings to distressed clients (Phase 1). A case vignette exercise is then chosen to measure diagnostic reasoning, and a structured interview with simulated client interactions to assess communication skill (Phase 2).
Criterion Development
What does good performance actually look like?
Before a selection procedure can be evaluated, you need a clear definition of success. Work analysis provides that definition.
Defining the content and context of performance
Work analysis identifies the specific activities, behaviours, and outcomes that constitute effective performance, as well as the work setting in which performance occurs. This is necessary before any performance measure can be meaningfully constructed.
Criteria for evaluating selection tools
The performance measures developed through this application serve as criteria — the standard against which individual selection tools or an entire selection system can be formally validated. A predictor is only useful if it predicts these work-analysis-derived criteria.
See an example
Clinical psychologist role. Work analysis of a clinical psychologist role identifies that the core activities include conducting risk assessments, delivering CBT sessions, and writing clinical reports. From this, "good performance" is defined concretely: reliable risk assessment judgements (checked against expert consensus), client improvement on standardised outcome measures such as the PHQ-9 and GAD-7, and report quality as rated through peer review. These become the criteria against which the selection procedure will be validated.
Domain Sampling
What is the full universe of the job?
Some selection approaches — particularly content-valid tests — require a precise definition of the job's content domain so that test items can be drawn from it representatively.
Defining the work domain
Work analysis maps the full universe of important tasks, activities, responsibilities, and work behaviours that make up the job. This domain definition determines what a representative sample looks like — essential when building tests whose validity rests on content representativeness rather than statistical correlation with a criterion.
Associated worker requirements
Alongside the task domain, work analysis maps the KSAOs associated with each task. This ensures that any test drawn from the domain targets the right underlying attributes, not just task surface features.
See an example
Clinical psychologist role. Work analysis maps the full task domain of a clinical psychologist: administering psychometric batteries, formulating case conceptualisations, delivering evidence-based interventions, liaising with multidisciplinary teams, managing risk, and writing clinical reports. A selection exercise designed to have content validity must sample representatively across this domain — so interview questions and written tasks are drawn from the full range, not just the most visible or easily tested activities.
Validity Evidence Extension
Can we transfer validity evidence from elsewhere?
Conducting a full validation study in every new setting is often impractical. Work analysis enables organisations to extend or transfer existing evidence to new contexts through three recognised approaches.
Validity generalisation (meta-analysis)
Drawing on accumulated research across many studies to show that a selection procedure has demonstrated validity for a given type of work broadly, and can therefore be expected to be valid in the new setting. Work analysis confirms that the target job is sufficiently similar to those studied.
Synthetic (job-component) validity
Validity is built up by combining evidence for each component or KSAO of a job, rather than validating a whole system against a single criterion. Work analysis identifies the components; validity evidence for each is then assembled from existing research.
Validity transportability
Demonstrating that validity evidence from one job or setting can be transferred to a new one because the jobs are sufficiently similar in their work requirements. Work analysis of both the original and new jobs provides the evidence that the transfer is justified.
See an example
Clinical psychologist role. An NHS trust wants to select clinical psychologists but cannot run a full local validation study. Using validity generalisation, it draws on accumulated research showing that structured clinical interviews and case vignette exercises predict diagnostic accuracy across comparable settings. Work analysis of the trust's own role confirms it is sufficiently similar in its requirements to the jobs already studied — providing the justification needed to use those instruments without conducting a new study from scratch.
Work analysis methods
Methods differ in how they compile, analyse, and present work analytic information. The most fundamental distinction is between qualitative and quantitative approaches.
Work analysis processes
Work analysis processes can be broadly differentiated in terms of whether they are primarily qualitative or quantitative.
Qualitative
Qualitative approaches build up work information from scratch — typically through observing or interviewing job incumbents to determine the specific tasks performed — generally one job at a time.
- Produced through observation, interviews, or open-ended inquiry
- Generates detailed, narrative descriptions
- Customised to individual jobs or specific work within an organisation
- Flexible — can capture nuance and context that structured surveys miss
Quantitative
Quantitative approaches use structured questionnaires or surveys built from pre-established lists of work descriptors — such as work behaviours, worker functions, or KSAOs — rated by subject matter experts (SMEs).
- SMEs are typically incumbents, supervisors, or job analysts
- Rating scales quantify judgments along dimensions such as frequency, importance, complexity, and consequences of error
- Enables statistical comparison across jobs and organisations
- Examples include the Position Analysis Questionnaire (PAQ)
Work analysis content
Work analysis content refers to the types of work descriptors used and the level of analysis or detail they represent. McCormick (1979) identified three broad descriptor categories, each defined by a different frame of reference.
Work-oriented
Frame of reference: The work itself
The frame of reference is the work itself — its purpose, the steps involved, the tools and materials required, and the conditions under which it is carried out.
Examples: Tasks, activities, duties, responsibilities, working conditions, work outputs
Worker-oriented
Frame of reference: What workers do
The frame of reference is what workers actually do in order to carry out the work — focusing on human behaviour rather than the work product itself.
Examples: Worker functions, processes, behaviours
Attribute requirements
Frame of reference: What workers need
The frame of reference is what workers need to bring to the work — the attributes required to do the job.
Examples: Skills, knowledge, abilities, and temperaments or dispositions
Subject matter experts (SMEs) are the people whose judgments provide the raw data in quantitative work analysis. They are typically incumbents who currently perform the job, supervisors who oversee it, or trained job analysts — each bringing a different vantage point on what the work actually involves.
Data collection
An important issue in any data collection process involving human sources is the risk of distortion — whether intentional or not. People's accounts of their own work can be shaped by self-interest, selective memory, or simply the difficulty of articulating tacit knowledge.
Best practice: Use more than one data collection method wherever possible — for example, conducting interviews first and then following up with a structured questionnaire. This allows findings from different methods to be cross-checked for convergence, increasing confidence in the resulting work information.
Four inferential leaps
Applying work analysis to employee selection requires a chain of inferential steps — each one moving from what the work analysis reveals to what the selection system needs to achieve. Gatewood et al. (2008) identified four such leaps.
Work content → Worker attributes
Translating information about what the work involves into the attributes — KSAOs — a worker needs to perform it. This is the core inference that links the job description to the person specification.
See an example
Clinical psychologist role. Work analysis reveals that the job requires conducting clinical assessments, formulating case conceptualisations, and delivering evidence-based interventions. From this, the analyst infers that the role requires diagnostic reasoning ability, knowledge of psychometric tools, and the ability to communicate findings clearly to clients and colleagues — these become the required KSAOs.
Work content → Criterion measures
Translating work content information into measures of work performance that can serve as criteria for evaluating selection tools. Defining what "good performance" looks like requires a clear picture of what the work actually involves.
See an example
Clinical psychologist role.Because the work involves delivering effective treatment and producing accurate clinical documentation, "good performance" can be defined as: client improvement on validated symptom measures (e.g. PHQ-9), quality of case formulations as rated by supervisors, and timely completion of clinical reports. These become the criterion measures against which selection tools are evaluated.
Worker attributes → Selection instruments
Translating the identified KSAOs into actual assessment tools designed to measure them — for example, choosing an ability test, a structured interview, or a work sample based on the attributes identified through work analysis.
See an example
Clinical psychologist role. Diagnostic reasoning is assessed via a case vignette exercise; knowledge of psychometric tools via a written knowledge test; communication skills via a structured interview with role-play scenarios. Each instrument maps directly to a KSAO identified through work analysis.
Selection instruments → Performance measures
The inferential connection between what a selection tool measures and the performance criteria it is intended to predict. This leap is the subject of validation — demonstrating that the instrument actually relates to on-the-job performance.
See an example
Clinical psychologist role. A validation study checks whether scores on the case vignette exercise correlate with supervisor ratings of case formulation quality and client symptom improvement at six-month follow-up. If they do, the inference holds — the instrument is measuring something that genuinely predicts the performance that matters in this role.
Cognitive task analysis
Emerging in the 1970s, cognitive task analysis (CTA) is a set of methods for identifying the cognitive skills and mental demands required to perform a task proficiently (Militello & Hutton, 1998).
What it adds
Standard task analysis captures the observable, physical elements of work — what people do, in what sequence, using what tools. CTA supplements this by focusing on the mental operations that are not directly observable: the decisions being made, the knowledge being applied, the cues being noticed, and the reasoning that connects them. For roles where cognitive demands are high — medical diagnosis, emergency response, engineering — this hidden layer of the work is often the most critical to capture.
Standard vs cognitive task analysis: A standard task analysis might record that a nurse "assesses a patient's condition" — CTA would probe what they are actually thinking: which cues they attend to, what prior knowledge they draw on, how they distinguish normal from abnormal findings, and how they decide what to do next.
Background: hierarchical task analysis
Developed in the 1960s, hierarchical task analysis (HTA) breaks a task down into subtasks at whatever level of detail is required. Like standard task analysis, it describes what people do — but it goes further by explicitly incorporating the conditions that trigger each action and the feedback that signals whether the goal has been achieved (Annett, 2003).
Each subtask is composed of four parts:
Subtask goal
What the subtask is trying to achieve — the specific outcome or end state the worker is working towards.
Input conditions
The conditions that must be present before the subtask begins — the triggers or prerequisites that initiate the action.
Action or operation
The actual steps or operations carried out by the worker in order to achieve the subtask goal.
Feedback
Information received by the worker that signals whether the goal has been attained — what tells them the subtask is complete.
How HTA differs from standard task analysis: Standard task analysis captures the goal and the action. HTA explicitly adds the input conditions (what triggers the task) and the feedback mechanism (how the worker knows the task is done) — making the analysis more complete and more useful for training and error diagnosis.
The future of work analysis
Traditional work analysis is a largely bottom-up endeavour: it begins with workers, observes what they do, and builds upward. An emerging perspective argues this is no longer sufficient — and that work analysis needs to be reconceived as a strategic organisational tool.
Strategic work analysis
A systematic effort to identify or define current or anticipated work and worker requirements that are strategically aligned with an organisation's mission and goals.
Strategic work analysis (SWA) reconceptualises work analysis as a strategic tool with a strong organisation development (OD) component. Rather than simply documenting what a job currently involves, SWA asks what the work needs to look like given where the organisation is heading — making it as much a top-down process (grounded in macro-organisational strategy and context) as a bottom-up one (grounded in what workers actually do).
What this rethinking requires:
Broader scope
Work analysis needs to become strategic, multistep, multifaceted, and interdisciplinary — not a bounded data-collection exercise but an ongoing organisational process.
Rethinking who does it
The qualifications and organisational roles of those conducting work analysis need to reflect its expanded purpose — including people with strategic and OD expertise, not only HR specialists.
A new goal
The aim shifts from gathering information about work to generating insight, meaning, and knowledge about it — transforming work analysis from a descriptive tool into a source of strategic understanding.
Test yourself
Four quizzes, each one a step up from the last — from definitions through to scenario judgement.