Employee Selection, Hiring and Onboarding

Introduction

Employee selection converts workforce requirements into decisions about who should join an organization. Hiring then formalizes the employment relationship, while onboarding and organizational socialization help the new employee become a competent, connected and committed organizational member. Weakness at any stage can produce poor performance, avoidable turnover, legal exposure and reputational harm.

A sound selection system should satisfy five qualities: it should be job-related, reliable, valid, fair and practical. These qualities are connected. A highly consistent test has little value if it measures irrelevant attributes; a valid method may still be unacceptable if it creates unjustified barriers; and a sophisticated assessment may be impractical if its cost exceeds its value.

Selection is not merely an administrative screening exercise. It is an evidence-based decision process in which imperfect information about applicants is used to predict future work behavior. Because hiring decisions affect both organizational outcomes and applicants’ livelihoods, they require procedural discipline, ethical judgment and continuing evaluation.

Learning Objectives

After completing this chapter, students should be able to:

1. Explain the relationship among job analysis, selection, hiring and onboarding.
2. Evaluate selection methods using reliability, validity, fairness, practicality and utility.
3. Design structured interviews, employment tests, work samples, assessment centers and reference checks.
4. Combine predictor information using defensible decision rules.
5. Identify sources of bias and apply legal, ethical and privacy safeguards.
6. estimate the financial utility of a selection system using unit-consistent assumptions.
7. Design an onboarding and organizational socialization process.
8. Evaluate selection and onboarding outcomes through appropriate metrics.

1. Selection as an Evidence-Based Process

Selection begins with job analysis. A job description identifies duties, responsibilities and working conditions, while a person specification identifies the knowledge, skills, abilities and other characteristics—often abbreviated as KSAOs—required for effective performance. These requirements should determine what is assessed.

A coherent process links each predictor to a job requirement and each job requirement to an important outcome. For example, a customer-service simulation may assess conflict management because job analysis shows that resolving difficult customer interactions is essential. Asking applicants unrelated personal questions does not satisfy this logic.

Figure: Selection, hiring and onboarding as a sequence with performance and retention feedback to assessment design and job analysis.

The feedback loops are essential. If new hires repeatedly struggle with an unassessed task, the organization should reconsider its job analysis and assessment design rather than simply blaming individuals.

2. Standards for Selection Quality

2.1 Reliability

Reliability is the consistency of measurement. Classical test theory represents an observed score as a true-score component plus random error:

Here, X is the observed score, T is the stable attribute relevant to the measurement purpose, and E is random measurement error. Under the classical assumption that true scores and random errors are uncorrelated, reliability can be expressed as:

where r_xx is the reliability coefficient, σ_T² is true-score variance and σ_X² is observed-score variance. Reliability ranges conceptually from 0 to 1, although sampling error can produce unusual estimates.

Common forms include:

• Internal consistency: whether items intended to measure the same construct produce coherent scores.
• Test–retest reliability: score stability across time when the construct is expected to remain stable.
• Inter-rater reliability: agreement among interviewers, observers or assessors.
• Parallel-forms reliability: consistency across equivalent versions of an instrument.

The standard error of measurement estimates uncertainty around an observed score:

Here, SEM and SD_X use the test-score unit. The formula assumes an appropriate classical reliability estimate. For example, if a test has a standard deviation of 10 points and reliability of 0.84, then SEM = 10 × √(0.16) = 4 points. A score should therefore not be interpreted as perfectly precise.

Reliability is necessary but not sufficient for validity. A scale that consistently measures the wrong attribute is reliable but invalid for selection.

2.2 Validity

Validity concerns whether evidence supports the interpretation and use of scores for a particular purpose. It is not an inherent label permanently attached to a test.

A criterion-related validity coefficient is commonly a correlation:

Here, X is the predictor score, Y is the criterion such as job performance, cov(X,Y) is covariance, and σ_X and σ_Y are their standard deviations. Correlation does not by itself prove causation. Criterion contamination, unreliable performance ratings, range restriction and small samples can distort results.

In predictive validation, applicant scores are collected before hiring and compared with later outcomes. In concurrent validation, current employees complete the assessment and their scores are compared with current criteria. Predictive designs more closely reproduce actual selection conditions, whereas concurrent designs are faster but may suffer from survivor and experience effects.

2.3 Fairness, practicality and applicant reactions

Fairness concerns equitable opportunity, consistent procedures and the absence of irrelevant barriers. Practicality concerns cost, time, expertise, scalability, security and operational feasibility. Applicant reactions also matter because opaque, disrespectful or excessively burdensome processes may reduce offer acceptance and damage the employer’s reputation.

No method should be chosen solely because it is inexpensive or technologically novel. Selection design requires balancing evidence quality, fairness, candidate experience and utility.

3. Major Selection Methods

3.1 Application forms and résumés

Application forms collect standardized information about qualifications and employment history. They are more comparable than unstructured résumés, but organizations should request only job-relevant information. Résumés can support initial screening but are vulnerable to inconsistent interpretation, exaggeration and bias associated with names, addresses, institutions or employment gaps.

Screening criteria should be specified before applications are reviewed. Reviewers should distinguish genuinely essential requirements from preferences that unnecessarily narrow the talent pool.

3.2 Employment interviews

Interviews are widely used because they permit two-way communication and allow assessment of complex behavior. Their quality varies substantially.

An unstructured interview changes questions and evaluation standards across applicants. It is vulnerable to first-impression effects, similarity attraction, confirmation bias, halo effects and excessive reliance on intuition.

A structured interview uses:

1. questions derived from job analysis;
2. the same core questions in the same order;
3. behaviorally anchored scoring guides;
4. trained interviewers;
5. independent ratings before discussion; and
6. documented decision rules.

Two important formats are:

• Behavioral-description questions: ask what the applicant did in a relevant past situation.
• Situational questions: ask what the applicant would do in a hypothetical work situation.

A behaviorally anchored scale for conflict management might assign 1 point to avoidance without follow-up, 3 points to gathering facts and discussing the problem, and 5 points to a balanced response that investigates, communicates, resolves and documents appropriately.

Panel interviews can broaden perspectives, but only if members score independently and are not dominated by the most senior participant. Interviewer training should include probing, note-taking, legal boundaries and calibration exercises.

3.3 Cognitive, knowledge and skill tests

General mental ability tests can predict learning and performance across many jobs, especially complex roles. However, group differences and accessibility concerns require careful validation, monitoring and consideration of less adverse alternatives.

Job-knowledge tests assess what applicants know, while skill or proficiency tests assess what they can do. Cut scores should be linked to minimum job competence rather than selected merely to reduce the applicant pool.

3.4 Personality and integrity measures

Personality inventories may predict relevant tendencies such as conscientiousness, dependability or interpersonal orientation. They should use validated, occupationally appropriate instruments. Clinical instruments designed to diagnose mental disorders are generally unsuitable for ordinary employment selection and may raise serious disability and privacy concerns.

Integrity measures assess attitudes or behavioral tendencies associated with counterproductive conduct. They should not be treated as infallible lie detectors. Faking, cultural interpretation and false positives require cautious use.

3.5 Work samples and simulations

Work samples ask applicants to perform tasks resembling actual work. Examples include editing a report, troubleshooting equipment or responding to a simulated customer complaint. Because the connection to work is visible, applicants often perceive them as fair. They may nevertheless be costly to develop and can favor applicants with prior experience when the organization intends to hire for potential.

Simulation fidelity should match the decision need. High physical realism is not always necessary; what matters is whether the simulation elicits the critical behavior.

3.6 Assessment centers

An assessment center is not necessarily a physical place. It is a standardized process in which multiple assessors evaluate participants across several exercises, often including role plays, in-basket exercises, presentations and group discussions.

A well-designed assessment center uses multiple exercises, trained assessors, explicit competency definitions, systematic observation and integrated ratings. It can provide rich evidence for managerial or complex interpersonal roles, but it is resource-intensive. Assessor discussion must integrate evidence rather than substitute hierarchy or intuition for recorded observations.

3.7 Biodata and situational judgment tests

Biodata instruments use systematically scored information about past experiences and behavior. Items must be demonstrably relevant and legally permissible. Situational judgment tests present job-related scenarios and ask applicants to select or rate responses. Their effectiveness depends on realistic scenarios, defensible scoring and appropriate reading demands.

3.8 Reference, background and credential checks

Reference checks can verify dates, responsibilities, conduct or eligibility for rehire, but references may be reluctant, inconsistent or influenced by personal relationships. A structured reference process should obtain consent, ask the same job-related questions and document responses.

Background checks should be proportionate to the role. Criminal-history, credit, driving-record or identity checks may be regulated differently across jurisdictions. Decision makers should consider relevance, recency, seriousness, evidence accuracy and any required opportunity for the applicant to respond. Credential checks confirm claimed qualifications and should be distinguished from subjective references.

4. Combining Evidence and Making Decisions

Selection methods generate evidence; they do not make decisions independently. Organizations need explicit rules for combining scores.

4.1 Multiple-hurdle and compensatory models

A multiple-hurdle model requires applicants to pass each stage, often sequentially. It is suitable when a minimum level on every essential requirement is necessary. For example, a driver may need both a valid license and a minimum safety score. Early inexpensive hurdles can reduce costs, but each hurdle must be defensible.

A compensatory model permits strength on one predictor to offset weakness on another. A weighted composite can be written as:

Here, S is the composite selection score, X_i is an applicant’s score on predictor i, and w_i is its predetermined weight. Scores should be placed on compatible scales before combining them. Weights should reflect validation evidence and job importance, not post hoc preference for a favored applicant.

Mechanical combination of evidence generally produces more consistent decisions than informal holistic judgment. Human review remains important for data quality, accommodation and exceptional circumstances, but discretionary overrides should be rare, documented and audited.

4.2 Selection rate and base rate

The selection rate is:

A lower selection rate allows greater selectivity but may increase assessment cost and adverse-impact risk. The base rate is the proportion of applicants who would perform successfully under the existing method or without the new predictor. Selection utility is greatest when there is room to improve on the current success rate.

5. Selection Utility

Selection utility estimates whether improved hiring decisions create benefits exceeding assessment costs. A transparent introductory model is:

where:

• EU = expected net utility, in dollars;
• n = number of employees hired, measured as employees;
• ΔP = expected improvement in performance per employee attributable to the new method, measured in performance units per employee;
• V = dollar value of one performance unit for each employee, measured in dollars per performance unit; and
• C = total incremental cost of the method, in dollars.

The units are consistent:

This simplified model is appropriate only when the organization can credibly estimate average incremental performance and its monetary value over the chosen period. It should not be presented as a universal substitute for more detailed utility analysis.

Worked example. An organization expects to hire 10 employees. A structured assessment is expected to improve average performance by 0.25 performance units per employee. Each performance unit for each employee is valued at $4,000 over the analysis period. The total incremental assessment cost is $6,000.

The estimated net utility is $4,000. This result depends on the assumptions; it is not a guaranteed accounting profit.

A more specialized Brogden–Cronbach–Gleser-type model is sometimes expressed as:

where N is the number selected, T is the average tenure in years, r_XY is predictor validity, SD_Y is the dollar standard deviation of job performance per employee per year, Z_X is the average standardized predictor score of those selected, and C is total selection cost. This model requires stronger assumptions, including a defensible monetary performance distribution, appropriate standardization and stable validity. Utility estimates should therefore be accompanied by sensitivity analysis using conservative, expected and optimistic assumptions.

6. Bias, Fairness and Legal Governance

6.1 Sources of bias

Bias can enter through job descriptions, recruitment channels, predictor design, administration, scoring and final judgment. Common cognitive biases include:

• Halo effect: one favorable characteristic influences unrelated ratings.
• Horn effect: one unfavorable characteristic dominates evaluation.
• Similarity attraction: preference for people resembling the evaluator.
• Confirmation bias: searching for evidence supporting an early impression.
• Stereotyping: attributing group assumptions to an individual.
• Contrast effect: rating an applicant relative to another applicant rather than a standard.

Standardization, anchored scoring, assessor training, independent ratings and audits reduce—but do not eliminate—these risks.

6.2 Adverse impact monitoring

One screening indicator compares a group’s selection rate with that of the group having the highest selection rate:

Suppose 24 of 60 applicants in Group A are selected, giving 40%, while 18 of 60 applicants in Group B are selected, giving 30%. The impact ratio for Group B is 0.30 ÷ 0.40 = 0.75, or 75%.

In some jurisdictions, an impact ratio below 80% is used as an administrative screening guideline. It is not a universal legal definition of discrimination and is not a substitute for statistical analysis, job-relatedness evidence or local legal advice. Small samples can make ratios unstable, while large samples may reveal meaningful differences not captured adequately by a simple threshold.

6.3 Legal and ethical requirements

Specific laws vary, but defensible systems generally incorporate:

• job-related criteria supported by analysis and evidence;
• consistent administration and scoring;
• reasonable accommodation for qualified applicants where required;
• restrictions on disability-related or medical inquiries at specified stages;
• informed handling of personal information;
• consent and notice for regulated checks;
• secure records and defined retention periods;
• a process for correcting inaccurate information; and
• documentation sufficient to explain decisions.

Ethical practice can exceed minimum legal compliance. Applicants should be treated with dignity, told what to expect, given a reasonable opportunity to demonstrate competence and protected from unnecessary intrusion.

7. Technology and Artificial Intelligence in Selection

Applicant-tracking systems can organize applications, schedule assessments and preserve decision records. Video interviews, online simulations and algorithmic scoring can increase reach, but they also introduce risks involving accessibility, surveillance, data security and opacity.

AI systems may reproduce historical inequalities if trained on past hiring decisions or proxy variables. Removing protected characteristics does not guarantee fairness because other variables may encode similar information. Claims that facial expression, voice or emotion analysis reveals stable job-relevant traits require especially strong scientific scrutiny.

Responsible governance should include:

1. a clearly defined, job-related purpose;
2. validation under actual conditions of use;
3. accessibility and accommodation testing;
4. subgroup outcome monitoring;
5. vendor documentation and audit rights;
6. data minimization and security controls;
7. meaningful human accountability;
8. applicant notice where required; and
9. periodic revalidation after jobs, populations or algorithms change.

Automation should support consistent evidence-based judgment rather than conceal an indefensible decision behind technical complexity.

8. Hiring and the Employment Offer

Selection identifies a preferred candidate; hiring establishes the employment relationship. A conditional written offer commonly states the role, compensation, work location, start date, key benefits, reporting relationship and lawful conditions that remain to be completed.

Organizations should distinguish negotiable terms from standardized terms and authorize who may make commitments. Material promises should be documented. Unsuccessful candidates should be informed promptly and respectfully, subject to operational and legal constraints.

A realistic job preview communicates both attractive and challenging aspects of the role. Although candid information may cause some applicants to withdraw, it can improve expectation alignment and reduce avoidable early turnover. Employers should avoid making promises about promotion, flexibility or culture that actual work practices cannot support.

Pre-employment checks should occur in a lawful sequence. Medical or health-related examinations, where permitted and necessary, may be restricted until after a conditional offer and should be separated from ordinary selection information.

9. Onboarding and Organizational Socialization

Onboarding is the planned process through which a new employee acquires the knowledge, relationships, resources and confidence needed to function effectively. Orientation is usually a short event within the broader onboarding process.

Organizational socialization concerns how newcomers learn role expectations, workplace norms and organizational identity. A useful framework organizes onboarding around four elements:

• Compliance: policies, legal requirements and administrative procedures.
• Clarification: role expectations, priorities and performance standards.
• Culture: values, norms, history and how work is actually accomplished.
• Connection: relationships with managers, peers, clients and support networks.

Socialization can be institutionalized through common training, scheduled milestones and assigned mentors, or individualized through informal learning. Standardization promotes consistency, while personalization allows newcomers to contribute their distinct experience. Effective systems balance both.

9.1 A staged onboarding design

The manager has a central role. HR may design the system, but employees learn priorities and norms mainly through everyday interaction with supervisors and colleagues. A new employee should know what good performance looks like, where to obtain help and how progress will be reviewed.

9.2 Evaluating onboarding

Useful measures include time to proficiency, completion of essential training, role clarity, new-hire engagement, early error rates, manager evaluations and voluntary turnover at defined intervals. Measures should be segmented carefully to identify barriers without invading privacy.

Retention alone is insufficient: an employee may remain while poorly socialized or underperforming. Conversely, not all turnover is preventable or undesirable. Evaluation should combine performance, experience and retention indicators.

10. Illustrative Case: Meridian Health Services

Background

Meridian Health Services is a fictional regional provider hiring 30 patient-service coordinators annually. Its previous process consisted of résumé screening followed by an informal manager interview. Early turnover reached 32%, and patient complaints indicated inconsistent conflict handling.

Problem

Managers asked different questions and frequently selected candidates on the basis of “fit.” The organization had no scoring guides, reference protocol or systematic onboarding plan. Analysis showed that successful performance required accurate data entry, calm communication, privacy awareness and judgment under pressure.

Analysis

HR conducted interviews with job incumbents and supervisors, observed work and reviewed critical incidents. It identified four selection dimensions: data accuracy, interpersonal de-escalation, situational judgment and dependability. Résumé prestige was not strongly connected to these requirements.

The team also found a broken transition between hiring and work. Accounts and equipment were often unavailable on the first day, supervisors assumed HR had explained performance expectations, and new employees received difficult assignments before guided practice.

Approach

Meridian introduced a four-stage process:

1. an application screen limited to essential requirements;
2. a short data-accuracy work sample;
3. a structured panel interview with behaviorally anchored questions;
4. consent-based credential and structured reference checks for finalists.

Interviewers completed calibration training and rated candidates independently. HR monitored completion rates and selection rates across demographic groups, investigated material differences and documented any process changes.

The organization also created a 90-day onboarding plan. New hires received schedules and system guidance before arrival, met a trained peer mentor on day one, completed privacy simulations during the first week and received progressively more complex assignments. Supervisors conducted weekly first-month check-ins and formal reviews at 30, 60 and 90 days.

Outcome

After one year, early turnover fell from 32% to 18%, average time to independent case handling declined from nine to seven weeks, and patient complaints involving coordinator communication decreased. These outcomes did not prove that every change caused the improvement; labor-market and supervisory changes were possible alternative explanations. Meridian therefore continued collecting data and planned periodic revalidation.

Learning Points

The case demonstrates that selection and onboarding should be designed as one talent-entry system. Job analysis provided the foundation, structure reduced arbitrary judgment, realistic exercises increased job relevance, and staged socialization converted hiring potential into performance. The organization also avoided treating favorable first-year results as permanent proof of validity.

11. Implementation Challenges and Best Practices

Selection systems operate under uncertainty. Performance criteria may be deficient or contaminated, jobs may change, applicant pools may be small, and managers may resist structured procedures because they value autonomy. Validation evidence may not transfer perfectly across occupations, locations or populations.

Good practice therefore requires continuing governance:

• begin with current job analysis;
• assess only important requirements;
• use multiple sources of relevant evidence where justified;
• standardize administration and scoring;
• train and calibrate assessors;
• offer reasonable accommodations;
• monitor subgroup outcomes and candidate reactions;
• document decisions and overrides;
• protect applicant data;
• connect hiring information to onboarding without disclosing unnecessary sensitive data; and
• examine performance, retention and fairness outcomes over time.

A system is not evidence-based merely because it uses a test, numerical score or algorithm. Evidence-based selection requires valid interpretation, responsible implementation and willingness to revise the process when outcomes reveal weaknesses.

Chapter Summary

Employee selection predicts future work behavior from job-related evidence, hiring formalizes the employment relationship, and onboarding helps convert employee potential into sustained performance. Effective methods are job-related, reliable, valid, fair and practical. Structured interviews, tests, work samples, assessment centers and reference checks differ in cost, evidence quality and suitability, so methods should be matched to job requirements.

Explicit multiple-hurdle or compensatory rules improve consistency. Utility analysis compares expected performance value with selection costs, but its conclusions depend on transparent, unit-consistent assumptions. Bias, adverse impact, privacy and accommodation require continuing legal and ethical governance. Technology and AI do not remove organizational accountability. Finally, onboarding should integrate compliance, clarification, culture and connection through a planned process extending beyond first-day orientation.

Key Terms

Adverse impact: A substantially different selection outcome for members of different groups, evaluated under applicable statistical and legal standards.

Assessment center: A standardized process using multiple exercises and assessors to evaluate competencies.

Base rate: The proportion of applicants expected to succeed under an existing selection approach.

Behaviorally anchored rating scale: A scoring scale whose points are illustrated by observable examples of behavior.

Compensatory model: A decision model in which strength on one predictor may offset weakness on another.

Construct validity evidence: Evidence that a method measures the intended theoretical attribute.

Criterion-related validity: The relationship between predictor scores and relevant work outcomes.

Impact ratio: One group’s selection rate divided by the highest comparison-group selection rate.

Inter-rater reliability: The degree of consistency among assessors.

Job analysis: Systematic identification of job duties, conditions and worker requirements.

KSAOs: Knowledge, skills, abilities and other characteristics required for work.

Multiple-hurdle model: A process requiring applicants to meet a minimum standard at each stage.

Onboarding: The planned process that supports a newcomer’s entry, learning and integration.

Predictive validation: Evaluation of predictor scores against outcomes measured later.

Realistic job preview: Balanced information about favorable and challenging job conditions.

Reliability: The consistency or precision of measurement.

Selection utility: The estimated economic value generated by a selection process after costs.

Structured interview: An interview using standardized questions, administration and scoring.

Organizational socialization: The process through which newcomers learn roles, norms and relationships.

Work sample: An assessment requiring performance of tasks resembling actual work.

Review Questions

Short Answer Questions

1. Why is reliability necessary but insufficient for validity?
2. What features distinguish a structured interview from an unstructured interview?
3. What is the difference between orientation and onboarding?
4. What information does an impact ratio provide?
5. Why should reference checks use standardized questions?

Long Answer Questions

1. Explain how job analysis supports the legal and technical defensibility of a selection system.
2. Compare cognitive tests, work samples and assessment centers in terms of validity, fairness and practicality.
3. Discuss the causes of interviewer bias and the controls that can reduce it.
4. Explain how compliance, clarification, culture and connection contribute to organizational socialization.
5. Critically evaluate the opportunities and risks associated with AI-enabled selection.

Analytical / Application Questions

1. Design a selection process for an entry-level customer-service role, identifying the job requirements, assessment methods and decision rule.
2. An organization selects 45% of applicants from one group and 27% from another. Calculate and interpret the impact ratio, including the limitations of the result.
3. Evaluate Meridian Health Services’ redesigned process and recommend two additional measures that would strengthen causal conclusions about its outcomes.

Multiple Choice Questions with Answers

Question 1

Which statement best describes the relationship between reliability and validity?

A. A reliable measure is automatically valid.

B. Reliability is unnecessary when content validity is high.

C. Reliability supports validity, but a reliable method may measure the wrong attribute.

D. Validity concerns consistency, whereas reliability concerns job relevance.

Correct Answer: C. Reliability supports validity, but a reliable method may measure the wrong attribute.

Question 2

Which practice most clearly improves the structure of an employment interview?

A. Allowing each interviewer to choose different questions

B. Using job-related questions and behaviorally anchored scoring guides

C. Discussing applicants before interviewers record ratings

D. Relying on first impressions to reduce interview time

Correct Answer: B. Using job-related questions and behaviorally anchored scoring guides

Question 3

When is a multiple-hurdle model most appropriate?

A. When strength on any predictor should compensate for every weakness

B. When applicants must meet minimum standards on several essential requirements

C. When the organization has no defensible cut scores

D. When final decisions are based entirely on managerial intuition

Correct Answer: B. When applicants must meet minimum standards on several essential requirements

Question 4

An organization hires 10 employees. A new assessment improves expected performance by 0.25 performance units per employee, each unit is worth $4,000 for each employee, and total assessment cost is $6,000. Using EU = n × ΔP × V − C, what is expected net utility?

A. $3,500

B. $4,000

C. $6,000

D. $10,000

Correct Answer: B. $4,000

Question 5

Group A has a 40% selection rate and Group B has a 30% selection rate. What is Group B’s impact ratio relative to Group A?

A. 0.25

B. 0.40

C. 0.70

D. 0.75

Correct Answer: D. 0.75

Question 6

Which selection method requires multiple assessors to evaluate candidates across several standardized exercises?

A. Assessment center

B. Unstructured reference check

C. Résumé screen

D. Realistic job preview

Correct Answer: A. Assessment center

Question 7

Which onboarding activity primarily supports the connection element of socialization?

A. Completing payroll documentation

B. Reading the disciplinary policy

C. Meeting a peer mentor and key stakeholders

D. Taking an identity verification check

Correct Answer: C. Meeting a peer mentor and key stakeholders

Question 8

Which action is most appropriate when using AI in selection?

A. Assume automation removes human accountability.

B. Validate the system for its actual job-related use and monitor subgroup outcomes.

C. Keep the system’s purpose undefined to protect vendor confidentiality.

D. Retain all applicant data indefinitely in case it becomes useful.

Correct Answer: B. Validate the system for its actual job-related use and monitor subgroup outcomes.

Question 9

A predictor correlates with later job performance. This is primarily evidence of:

A. criterion-related validity

B. parallel-forms reliability

C. administrative practicality

D. organizational socialization

Correct Answer: A. criterion-related validity

Question 10

What is the main purpose of a realistic job preview?

A. To present only the most attractive features of employment

B. To replace all other selection assessments

C. To provide balanced information that improves expectation alignment

D. To guarantee that no new employee will resign

Correct Answer: C. To provide balanced information that improves expectation alignment