Workforce Data Architecture, Quality and Governance

Introduction

Human resource management increasingly depends on data generated across the employee lifecycle. Recruitment platforms record applicants, human resource information systems (HRIS) maintain employee records, learning management systems (LMS) capture development activity, payroll systems process compensation, surveys reveal employee perceptions, and external sources provide labour-market and economic context. These systems can support evidence-based decisions only when their data are connected, accurately defined, responsibly governed and interpreted within context.

Workforce data architecture is therefore more than a technical design problem. It is an organisational capability involving technology, process, people, legal compliance and managerial accountability. A poorly governed data environment may produce contradictory headcount figures, incomplete diversity reports, unreliable turnover rates, duplicated employee records and unjustified automated decisions. Conversely, a coherent architecture allows HR professionals to answer strategic questions with greater speed, consistency and confidence.

This chapter examines the principal sources of workforce data, integration patterns, data definitions, lineage, quality management, access controls, ownership, master data and governance across the employee lifecycle. It also considers privacy, ethics, artificial intelligence and the limitations of treating workforce data as neutral or self-explanatory.

Learning Objectives

By the end of this chapter, students should be able to:

1. Distinguish the purposes and data structures of HRIS, ATS, LMS, payroll, survey and external data sources.
2. Explain how workforce data are integrated across the employee lifecycle.
3. Design useful data definitions, metadata and lineage records.
4. Diagnose and improve data quality using appropriate dimensions and controls.
5. Apply principles of data ownership, stewardship, master data management and access governance.
6. Evaluate ethical, privacy, regulatory and managerial risks in HR data practices.
7. Recommend a governance approach for a data-driven HR decision.

1. Workforce Data as an Organisational Asset

Workforce data are recorded representations of people, employment relationships, activities, outcomes and organisational context. Examples include employee identifiers, job families, employment status, salary, attendance, training completion, performance ratings, engagement responses and recruiting stages. Data become information when they are interpreted in context; they become evidence for decision-making when their quality, relevance and limitations are understood.

A central distinction is between operational data and analytical data. Operational data support transactions, such as issuing payroll or approving leave. Analytical data are structured and combined to examine patterns, relationships and trends. A payroll system may need to know the current pay rate for a worker, whereas an analytical dataset may require historical pay rates by month, currency, position and organisational unit.

Workforce data also have different levels of sensitivity. Public information, internal organisational information, personal data, special-category or highly sensitive data, and confidential business information should not receive identical controls. A home address, medical accommodation record, bank account and aggregated engagement score have different privacy and access implications.

Data do not speak independently. A turnover rate, for example, depends on the definition of a leaver, the population included, the time period, the treatment of acquisitions and the denominator used. A defensible decision therefore requires not only a number but also a documented meaning, source, calculation and limitation.

2. Major Workforce Data Sources

2.1 Human Resource Information Systems

An HRIS is the principal system for maintaining employee and employment information. Typical entities include person, employment relationship, position, job, organisation, location, manager, compensation and absence. Modern HR platforms may also provide workflow, self-service, talent and reporting capabilities.

The HRIS is often treated as the authoritative source for worker identity and employment status, but this authority should be explicitly defined rather than assumed. In a multinational enterprise, different systems may be authoritative for different attributes. For example, the HRIS may own legal name and employment status, while payroll owns payment-account details and a global identity service owns login credentials.

2.2 Applicant Tracking Systems

An applicant tracking system (ATS) manages recruitment requisitions, candidates, applications, stages, interviews, offers and recruitment outcomes. Its data are event-oriented: a candidate applies, is screened, interviewed, offered a role or rejected. Candidate data often contain personally identifiable information, assessment results and demographic information, requiring retention and access controls.

A frequent integration problem arises when a candidate becomes an employee. The ATS candidate record must be connected to, but not casually duplicated as, the HRIS worker record. A controlled conversion process should preserve relevant recruitment history while establishing a unique employee identifier and clear retention rules.

2.3 Learning Management Systems

An LMS records courses, learning objects, enrolments, completions, assessments, certifications and learning hours. It may also capture informal learning or links to external providers. Completion does not necessarily demonstrate competence or behavioural transfer; therefore, learning data should not be interpreted as a direct measure of performance without additional evidence.

2.4 Payroll and Time Systems

Payroll systems contain earnings, deductions, taxes, benefits, payment details and payroll-period results. Time and attendance systems may record scheduled hours, worked hours, overtime, absence and time approvals. Payroll data are financially and personally sensitive. They also present temporal complexity: gross pay may be reported by pay period, while headcount is often measured at a point in time and turnover over an interval.

2.5 Surveys and Listening Platforms

Employee surveys generate responses, scores, comments and participation metadata. Survey data require special attention to confidentiality, small-group disclosure and response bias. A technically anonymous survey may still be identifiable when demographic filters are combined with comments, job title or team membership. Survey results should therefore use minimum reporting thresholds and carefully designed suppression rules.

2.6 External Data

External data may include labour-market supply, salary benchmarks, inflation, unemployment, skills trends, commuting patterns, industry data and regulatory information. These sources differ in methodology, population, update frequency and licensing conditions. Benchmark comparisons are meaningful only when job definitions, geographic scope, currency, time period and population are reasonably comparable.

3. Architecture Across the Employee Lifecycle

A workforce data architecture describes how data are created, stored, transferred, transformed, accessed and retired. It includes systems of record, integration interfaces, databases, analytical platforms, reporting tools, identity services and governance processes.

A common pattern is a layered architecture:

1. Source layer: operational systems such as ATS, HRIS, LMS, payroll and survey platforms.
2. Integration layer: application programming interfaces (APIs), event streams, scheduled files or middleware that move data between systems.
3. Storage and analytical layer: a warehouse, lakehouse or governed data lake containing standardised and historical data.
4. Semantic layer: shared definitions, metrics, dimensions and business rules.
5. Consumption layer: dashboards, reports, analytical models and decision-support applications.
6. Governance and security layer: identity, access, quality, metadata, retention, audit and accountability controls applied across all layers.

Integration may be batch-based, near-real-time or event-driven. Batch integration is often sufficient for monthly workforce reporting and may be simpler to operate. Near-real-time integration is more appropriate for time-sensitive processes, such as provisioning access after a new hire is approved. Event-driven architecture publishes an event, such as “employment terminated,” so that authorised systems can respond. The correct choice depends on business urgency, reliability, cost, security and operational resilience.

A data warehouse commonly uses a dimensional design. A fact table records measurable events or states, such as payroll payments, applications or learning completions. Dimension tables describe entities such as employee, date, job, department and location. A critical design issue is history. If an employee moves from Department A to Department B, a historical report may need to retain the department that applied when an event occurred. Overwriting the old value can make historical analysis inaccurate.

The employee lifecycle creates particular integration points:

• workforce planning connects external labour data, vacancies and organisational plans;
• recruitment connects requisitions, candidates, selection decisions and offers;
• onboarding connects accepted offers, worker identity, position and access provisioning;
• employment connects payroll, time, performance, learning, mobility and absence;
• offboarding connects termination status, final pay, access removal and retention or deletion;
• alumni or rehire processes reconnect historical identities under controlled rules.

The objective is not to place every data item in one system. It is to establish reliable relationships, authoritative sources and traceable transformations.

4. Shared Definitions, Metadata and Data Lineage

4.1 Data Definitions

A data dictionary describes fields and their meanings. A useful definition includes the business term, precise meaning, data type, allowed values, owner, source system, update frequency, sensitivity classification and validation rules.

Consider “headcount.” It might mean the number of active employees on the last calendar day of a month, excluding contingent workers and including employees on paid leave. A different report may count all workers with an active assignment, including contractors. Both figures can be valid if their definitions are explicit; they become problematic when presented as though they are the same.

Metric definitions should specify:

• population and inclusion criteria;
• time reference, such as point-in-time or period-based;
• numerator and denominator;
• treatment of transfers, rehires, interns and contingent workers;
• data source and refresh date;
• exclusions, assumptions and known limitations.

For example, a basic period turnover rate can be represented as:

Here, “leavers” must be defined, and average headcount may be calculated as the mean of beginning and ending headcount or by averaging more frequent observations. The formula is meaningful only when the numerator and denominator refer to compatible populations and periods. It should not be interpreted as a probability that a particular employee will leave.

4.2 Metadata

Metadata are data about data. Structural metadata describe fields, tables, relationships and formats. Business metadata describe meaning, ownership and usage. Technical metadata record pipelines, refresh times, schemas and system errors. Privacy metadata identify classification, lawful purpose, retention and access constraints.

A mature organisation maintains a business glossary linked to technical catalogues. This prevents a common failure in which a dashboard uses a technically valid field whose business meaning is unclear.

4.3 Data Lineage

Data lineage records the origin, movement and transformation of data. It answers questions such as: Where did this figure originate? Which fields were joined? Which filters were applied? When was it last refreshed? Who approved the transformation?

Lineage may be represented at column level, table level or report level. For a diversity dashboard, lineage should identify the source demographic field, permitted categories, missing-value treatment, aggregation logic, suppression rules and publication date. Lineage is essential for auditability, error investigation, regulatory response and trust in analytics.

5. Data Quality Management

Data quality is fitness for a defined purpose, not an abstract property that exists independently of use. A field may be adequate for internal scheduling but unsuitable for statutory reporting or predictive modelling.

Important quality dimensions include:

• Accuracy: agreement with the real-world object or event.
• Completeness: required values are present.
• Consistency: related records and systems do not conflict.
• Timeliness: data are available and current enough for the decision.
• Validity: values conform to permitted formats, ranges or categories.
• Uniqueness: an entity or event is not incorrectly duplicated.
• Integrity: relationships between records remain valid.
• Fitness for purpose: data support the intended decision adequately.

A simple completeness indicator can be expressed as:

The result is a percentage. It measures presence, not correctness. A field containing “unknown” in every record may be technically populated but substantively incomplete.

Quality management should combine prevention, detection and correction. Prevention includes controlled forms, validation rules, reference data and user training. Detection includes reconciliation reports, duplicate checks, exception queues and statistical monitoring. Correction includes authorised updates, source-system fixes, documented manual remediation and root-cause analysis.

A practical quality control might reconcile HRIS active employees with payroll recipients. Differences may be legitimate—for example, unpaid leave or a delayed payroll cut-off—but each difference should be classified rather than automatically “fixed.” A quality score without investigation can create false confidence.

Common causes of poor quality include inconsistent job coding, free-text entry, system migrations, duplicated identities, weak integration keys, organisational reorganisations, local customisation and unclear accountability. The most effective response is usually to correct the process or system that creates the error, not repeatedly clean downstream extracts.

6. Master Data, Reference Data and Identity Resolution

Master data represent core entities shared across processes, such as person, worker, job, position, organisation and location. Reference data are controlled lists used to classify information, such as employment status codes, country codes, absence reasons and education categories.

Master data management (MDM) establishes authoritative records, identifiers, matching rules, survivorship rules and stewardship processes. A persistent person identifier is particularly important. Names, email addresses and job titles may change; a stable identifier should link records without exposing unnecessary personal information.

Identity resolution must distinguish a person from an employment relationship. One person may have multiple employments, contracts or assignments. A rehire may require a new employment record while preserving a controlled relationship to the same person record. Treating every employment as a new person can distort headcount, turnover and diversity analysis; treating every similar name as the same person can create severe privacy and payroll errors.

Reference data governance includes versioning and translation. If one country uses “full-time,” “part-time” and “temporary,” while another uses local categories, a global analytical model needs a documented mapping. Mapping should preserve local detail where necessary and should not imply equivalence without review.

7. Governance, Ownership and Stewardship

Data governance is the system of decision rights, policies, standards, roles and controls through which an organisation manages data. It is not merely an information-security committee and not a one-time technology project.

Typical roles include:

• Executive sponsor: provides authority, resources and strategic direction.
• Data owner: accountable for a data domain, its appropriate use, risk and policy decisions.
• Data steward: manages definitions, quality rules, issue resolution and operational coordination.
• System owner: responsible for the operation and functionality of a technology platform.
• Data custodian: implements technical storage, security, backup and access controls.
• Analyst or data consumer: uses data according to approved definitions and purposes.
• Privacy, legal and security specialists: advise on lawful processing, risk and safeguards.

Ownership should be assigned by data domain and decision right. A payroll director may own compensation data, while an HR operations leader owns employment status. Ownership does not mean unrestricted access. It means accountability for decisions and controls.

A governance council can approve enterprise definitions, prioritise quality issues, resolve conflicts between systems and review new uses of sensitive data. Its effectiveness depends on clear escalation routes and measurable outcomes. Policies that cannot be implemented in systems, workflows and training are unlikely to change behaviour.

8. Access, Privacy, Security and Ethical Use

Access should follow least privilege: users receive only the access necessary for an authorised role. Role-based access can distinguish HR administrators, line managers, analysts, payroll specialists and employees. Attribute-based controls may additionally consider geography, employment relationship, data sensitivity or purpose.

Important controls include authentication, encryption in transit and at rest, logging, periodic access reviews, segregation of duties, retention schedules, secure deletion and incident response. Sensitive data should be minimised. Collecting information merely because a system can store it increases exposure without necessarily improving decisions.

Privacy principles commonly relevant to workforce analytics include purpose limitation, data minimisation, transparency, accuracy, retention limitation, security and accountability. Applicable requirements vary by jurisdiction and may include employment law, data-protection law, records obligations, anti-discrimination law and sector-specific regulation. Legal compliance is a baseline, not a substitute for ethical judgement.

Ethical risks arise when data are used outside the context in which they were collected. Employees may participate in a wellbeing survey expecting aggregate feedback, not individual-level monitoring. Digital activity measures may reflect job design, disability, caregiving or network access rather than effort. Automated screening may reproduce historical discrimination if training data reflect biased recruitment decisions.

Before deploying an analytical or AI use case, HR leaders should ask:

1. What legitimate organisational purpose is being served?
2. Is the data necessary and proportionate?
3. Were affected people informed appropriately?
4. Could the use create disparate impact or unfair exclusion?
5. Can the result be explained and challenged?
6. Who is accountable for errors?
7. What human review, monitoring and appeal process exists?

AI systems require additional attention to model documentation, training-data provenance, validation, drift, security, explainability and human oversight. A model should not be treated as objective merely because it produces a numerical score. Predictions describe estimated patterns under assumptions; they do not establish causes or justify automatic decisions by themselves.

9. A Governance Framework for Implementation

A practical implementation can proceed through the following sequence:

Step 1: Establish decision priorities

Begin with decisions rather than technology. Examples include improving workforce planning, reducing avoidable recruitment delay or ensuring accurate regulatory reporting. Define the users, decisions, risks and required time horizon.

Step 2: Map the data landscape

Create an inventory of systems, data domains, interfaces, owners, classifications, retention periods and known quality problems. Map critical employee-lifecycle processes and identify points where records are created or changed.

Step 3: Define critical data elements

Prioritise fields whose failure would materially affect decisions, compliance or employee rights. Examples include worker identifier, employment status, start date, termination date, job code, organisation, pay basis and demographic categories.

Step 4: Agree definitions and authority

Create a glossary and assign each critical element an authoritative source, steward, allowed values, validation rule, refresh expectation and escalation route.

Step 5: Build quality controls

Implement validation at entry, automated reconciliation, duplicate detection, exception management and root-cause correction. Establish service levels for resolving high-risk errors.

Step 6: Design access and privacy controls

Classify data, define role-based access, document approved purposes, apply retention rules and test whether reports could reveal individuals through small groups or combinations of fields.

Step 7: Integrate and document

Use stable identifiers, controlled interfaces, schema versioning, error handling and lineage documentation. Do not rely on undocumented spreadsheets as permanent integration infrastructure.

Step 8: Monitor and improve

Track data-quality measures, access violations, unresolved issues, pipeline failures, definition disputes and user confidence. Governance should evolve as organisational structures, legislation and technology change.

10. Illustrative Case: Meridian Services

The following is a fictional illustrative case.

Background

Meridian Services, a 12,000-employee professional-services organisation, operated separate HRIS, ATS, LMS, payroll and survey platforms across six countries. Senior leaders received different monthly headcount figures from HR, finance and business-unit dashboards.

Problem

A workforce plan reported 11,640 employees, finance reported 11,410 payroll recipients and the HR dashboard reported 11,890 active workers. Recruitment analysts also reported that time to hire had improved, while regional HR teams believed vacancies were taking longer to fill. An internal review found inconsistent treatment of contractors, delayed termination updates, duplicate rehire records and multiple definitions of “filled requisition.”

Analysis

The organisation mapped its data flows and discovered that the HR dashboard counted active employment records, finance counted workers paid in the most recent payroll cycle, and the workforce plan included some contractors. None of the definitions was inherently invalid, but the labels implied that they measured the same population.

The ATS stored candidate status locally, while the HRIS created an employee record only after onboarding. Requisitions that were cancelled or placed on hold were included in some time-to-hire calculations. The LMS used email addresses rather than a persistent worker identifier, causing learning histories to fragment after name or email changes.

Approach and Solution

Meridian established a workforce-data governance council chaired by the HR operations executive, with finance, payroll, privacy, IT, recruitment and regional representation. It assigned owners for worker, job, organisation, payroll and recruitment domains. A persistent worker identifier was introduced, with controlled links between person, employment and application records.

The council approved separate definitions for active headcount, paid headcount, total workforce and employee headcount. Each dashboard displayed its population, point-in-time date, refresh time and exclusions. The ATS definition of time to hire specified the start and end events, excluded cancelled requisitions and reported median alongside average because a small number of very long searches distorted the average.

Meridian added required-field validation, duplicate checks, termination reconciliation and a monthly exception process. Survey reporting introduced minimum group-size thresholds and suppressed free-text extracts where combinations of attributes could identify individuals. Access reviews removed broad payroll access from general HR analysts.

Outcome

After three reporting cycles, the organisation had fewer unexplained discrepancies and faster resolution of data issues. Leaders stopped asking which headcount number was “the real one” and instead selected the measure appropriate to the decision. Recruitment discussions also improved because the metric definition made differences in vacancy status visible.

Learning Points

The case demonstrates that data governance is not achieved by purchasing a single platform. It requires shared definitions, authoritative sources, persistent identity, documented lineage, quality controls and accountable decision rights. It also shows that disagreement between figures may reflect different legitimate purposes, although undocumented differences undermine trust.

11. Challenges and Limitations

First, integration can create the illusion of precision. A highly connected environment may still contain biased, incomplete or poorly defined data. Second, standardisation can erase meaningful local differences, particularly across countries and employment systems. Global models should distinguish genuine comparability from forced uniformity.

Third, data quality has costs. Excessive validation can slow operational work, while insufficient validation creates downstream risk. Controls should be proportionate to the consequence of error. Fourth, historical data are often not comparable after reorganisations, acquisitions, changes in job architecture or revised survey scales. Analysts should preserve methodological notes and avoid presenting breaks in series as genuine workforce change.

Fifth, employee trust can be damaged by opaque monitoring. Transparency must address not only what is collected but why, how it is analysed, who sees it and how employees can challenge inaccuracies. Finally, governance can become bureaucratic if every decision requires excessive approval. Effective governance differentiates routine, low-risk use from novel, sensitive or high-impact use.

12. Contemporary Developments and Best Practices

Cloud HR platforms increasingly provide APIs, event streams and embedded analytics, but convenience does not remove the need for architecture and governance. Data mesh approaches may distribute analytical responsibility to domains while retaining enterprise standards for identity, security and interoperability. Lakehouse platforms may combine flexible storage with warehouse-style governance, but implementation quality remains more important than terminology.

Organisations are also adopting data contracts: explicit agreements describing schema, meaning, ownership, quality expectations and change procedures between data producers and consumers. Privacy-enhancing techniques, such as aggregation, pseudonymisation and controlled synthetic data, can reduce exposure but do not guarantee anonymity. Metadata catalogues and automated lineage tools can improve discoverability, although automated lineage may fail to capture informal spreadsheet transformations or human interpretation.

Best practice is characterised by the following principles:

• design governance around important decisions and risks;
• maintain one clear definition for each shared metric, while permitting clearly labelled specialised measures;
• assign ownership and stewardship to named roles;
• use persistent identifiers and controlled reference data;
• correct errors at their source whenever feasible;
• document lineage, assumptions and historical changes;
• apply least privilege, minimisation and retention controls;
• test data and models for bias, disparate impact and unintended use;
• involve HR practitioners, employees, privacy specialists, technology teams and business leaders;
• treat governance as a continuous organisational capability.

Chapter Summary

Workforce data architecture connects operational systems, integration processes, analytical platforms, definitions, security and governance across the employee lifecycle. HRIS, ATS, LMS, payroll, survey and external data sources serve different purposes and contain different risks. Reliable analytics depend on explicit definitions, metadata and lineage rather than on numerical output alone.

Data quality should be assessed through dimensions such as accuracy, completeness, consistency, timeliness, validity, uniqueness and integrity, always in relation to intended use. Master data management supports persistent identity, authoritative records and controlled reference values. Governance assigns decision rights to owners, stewards, custodians and users, while access controls, privacy safeguards, retention and auditability protect employees and the organisation.

The central managerial lesson is that data discrepancies are often symptoms of different definitions, processes or authorities. Effective HR analytics therefore requires both technical integration and institutional agreement. Ethical and lawful use also requires purpose limitation, proportionality, transparency, human accountability and attention to bias. A mature workforce data environment is not simply a collection of systems; it is a governed capability for making defensible decisions.

Key Terms

Data architecture: The structure through which data are created, stored, integrated, transformed, accessed and retired.

HRIS: A system that maintains workforce and employment information and supports HR administration.

Applicant tracking system: A platform for managing recruitment requisitions, candidates, applications and selection stages.

Learning management system: A platform that administers learning content, enrolments, completions and certifications.

Operational data: Data used to conduct routine organisational transactions and processes.

Analytical data: Data structured and combined for analysis, reporting and decision support.

Metadata: Data describing the structure, meaning, origin, quality, sensitivity or use of other data.

Data lineage: A record of data origin, movement, transformation and consumption.

Data quality: The fitness of data for a defined purpose.

Master data: Shared core entities such as person, worker, job, organisation and location.

Reference data: Controlled classification values such as status, country or absence codes.

Master data management: Processes and controls for maintaining authoritative shared entities and identifiers.

Data owner: The person or function accountable for decisions, risk and policy concerning a data domain.

Data steward: The role responsible for definitions, quality coordination and issue resolution.

Data custodian: A technical role responsible for storage, security, backup and operational controls.

Least privilege: Providing users only the access required for an authorised purpose.

Purpose limitation: Using personal data for specified and legitimate purposes rather than unrestricted secondary use.

Persistent identifier: A stable identifier used to link records for an entity over time.

Data contract: An explicit agreement about data meaning, schema, quality, ownership and change management.

Review Questions

Short Answer Questions

1. What is the difference between operational data and analytical data in HR?
2. Why should a candidate record not be treated as identical to an employee record without a controlled conversion process?
3. Define data lineage and state two reasons it is important.
4. Name four dimensions of data quality.
5. Distinguish a data owner from a data custodian.

Long Answer Questions

1. Explain how HRIS, ATS, LMS, payroll and survey systems can be integrated across the employee lifecycle. Discuss the principal risks of integration.
2. Discuss the role of shared definitions and metadata in producing trustworthy workforce metrics. Use headcount or turnover as an example.
3. Explain master data management and identity resolution in a multinational organisation. Why are persistent identifiers important?
4. Evaluate the relationship between data governance, privacy, security and ethical HR analytics.
5. Design a practical workforce-data governance framework for an organisation with fragmented HR systems.

Analytical / Application Questions

1. An organisation reports 8,000 employees at the beginning of a year, 8,400 at the end, and 720 leavers during the year. Using the beginning-and-ending average headcount convention, calculate the period turnover rate and explain one limitation of the result.
2. An engagement dashboard shows results for a team of four employees, including demographic breakdowns and verbatim comments. Identify the governance risks and propose reporting controls.
3. HRIS records 10,200 active workers, payroll records 9,950 paid workers and the workforce plan records 10,500 total workers including contractors. Develop a reconciliation approach and explain why the figures may all be valid for different decisions.

Multiple Choice Questions with Answers

Question 1

Which system is typically the primary source for employee employment status and organisational assignment?

A. Learning management system

B. Human resource information system

C. Employee survey platform

D. External labour-market database

Correct Answer: B. Human resource information system

Question 2

What does data lineage primarily document?

A. Employee satisfaction with HR services

B. The hierarchy of organisational positions

C. The origin, movement and transformation of data

D. The number of users in a payroll system

Correct Answer: C. The origin, movement and transformation of data

Question 3

Which data-quality dimension concerns conformity to permitted formats, ranges or categories?

A. Validity

B. Timeliness

C. Uniqueness

D. Completeness

Correct Answer: A. Validity

Question 4

What is the main purpose of a persistent worker identifier?

A. To replace all access controls

B. To link records for the same worker over time

C. To reveal an employee’s salary to managers

D. To eliminate the need for data definitions

Correct Answer: B. To link records for the same worker over time

Question 5

Which role is primarily accountable for data-domain policy and risk decisions?

A. Data owner

B. Data consumer

C. Database operator

D. Survey respondent

Correct Answer: A. Data owner

Question 6

A report counts active employees on the last day of the month and excludes contractors. This statement is an example of a:

A. Data breach notification

B. Metric definition

C. Network protocol

D. Retention exception

Correct Answer: B. Metric definition

Question 7

Which principle means that users receive only the access necessary for an authorised role?

A. Data duplication

B. Least privilege

C. Open aggregation

D. Historical overwriting

Correct Answer: B. Least privilege

Question 8

An organisation has 1,000 employees at the beginning of a period and 1,100 at the end. Using the beginning-and-ending average, what is the average headcount?

A. 950

B. 1,000

C. 1,050

D. 1,100

Correct Answer: C. 1,050

Question 9

Which is the strongest reason to suppress very small survey groups?

A. To increase the number of survey questions

B. To reduce the risk of identifying individuals

C. To make all scores equal

D. To remove the need for encryption

Correct Answer: B. To reduce the risk of identifying individuals

Question 10

Why should completion of a learning course not automatically be interpreted as competence?

A. Completion data are always illegal to collect

B. Learning completion does not necessarily demonstrate behavioural transfer or capability

C. LMS records cannot contain dates

D. Competence can only be measured through payroll

Correct Answer: B. Learning completion does not necessarily demonstrate behavioural transfer or capability