Inclusive, Ethical and Legally Compliant Recruitment

Introduction

Recruitment decisions affect access to employment, organizational performance, employee trust and public confidence. A recruitment process may appear neutral while disadvantaging people because of disability, sex, race, ethnicity, age, religion, pregnancy, family responsibilities, sexual orientation, gender identity, socioeconomic background or another protected or socially relevant characteristic. It may also create risks through inappropriate questions, undisclosed conflicts of interest, careless handling of personal information or the use of artificial-intelligence tools that reproduce historical bias.

Inclusive recruitment is therefore more than increasing the diversity of applicants. It involves designing and administering recruitment so that qualified people have a fair opportunity to learn about a vacancy, apply, participate in assessment and receive a decision based on relevant evidence. Ethical recruitment adds duties of honesty, respect, confidentiality, accountability and avoidance of harm. Legal compliance establishes minimum requirements; responsible HR practice should normally aim higher than minimum compliance.

Employment law differs across jurisdictions. This chapter explains widely applicable principles, but HR practitioners must consult current national and local legislation, regulatory guidance, collective agreements and organizational policies before making decisions.

Learning Objectives

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

1. Explain equality, equity, inclusion, discrimination, bias and adverse impact.
2. Distinguish direct discrimination, indirect discrimination, harassment and retaliation.
3. Identify bias and accessibility risks throughout the recruitment process.
4. Apply principles of consent, privacy, confidentiality and secure documentation.
5. Recognize and manage conflicts of interest and favoritism.
6. Use structured, evidence-based and accessible selection practices.
7. Evaluate the ethical and legal risks of recruitment technology and AI.
8. Recommend practical controls for a fair and auditable recruitment process.

1. Foundations of Inclusive Recruitment

Equality, equity and inclusion

Equality means that people are not unlawfully treated differently because of a protected characteristic. Equity recognizes that identical treatment does not always produce a fair opportunity. For example, providing the same online form to every applicant may be formally equal, but it may exclude a qualified applicant who requires an accessible format. An adjustment, such as a screen-reader-compatible form or additional time for an assessment, supports equitable participation.

Inclusion concerns whether people can participate meaningfully and with dignity. It includes the design of job advertisements, communication methods, interview practices, assessment criteria and decision-making culture. Diversity describes characteristics or perspectives represented in a workforce; inclusion describes the conditions that allow those people to contribute and progress.

Fairness and merit

A fair process does not mean selecting people without regard to qualifications. It means defining job-related requirements accurately, communicating them honestly, assessing them consistently and considering evidence relevant to successful performance. Criteria should be necessary rather than customary. Requiring a particular qualification, physical ability, work schedule or style of communication should be defensible in relation to the job.

A useful distinction is between essential criteria and desirable criteria. Essential criteria are requirements without which the principal duties cannot reasonably be performed, even with appropriate adjustments. Desirable criteria may improve performance but should not automatically exclude an otherwise suitable candidate. Inflated criteria can reduce applicant diversity and create indirect discrimination.

Important terminology

• Protected characteristic: A characteristic protected by applicable law, such as disability, sex, age, race or religion.
• Direct discrimination: Less favorable treatment because of a protected characteristic.
• Indirect discrimination: A seemingly neutral rule or practice that disadvantages a protected group and is not legally justified or proportionate.
• Harassment: Unwanted conduct connected with a protected characteristic that violates dignity or creates an intimidating, hostile, degrading or offensive environment.
• Retaliation or victimization: Penalizing a person for raising a concern, supporting a complaint or exercising a legal right.
• Bias: A systematic tendency that influences judgment, often without conscious intention.
• Stereotype: A generalized belief about members of a group.
• Adverse impact: A substantially poorer selection outcome for one group compared with another.
• Reasonable accommodation or adjustment: A change that enables a person with a disability or other legally recognized need to participate, unless it creates disproportionate hardship under applicable law.
• Conflict of interest: A personal, financial or relational interest that could improperly influence professional judgment.

2. How Discrimination Can Enter Recruitment

Recruitment is a chain of decisions. A lawful and inclusive outcome requires attention at every stage.

Workforce planning and job analysis

Discrimination can begin when a role is designed around an incumbent rather than actual work requirements. A manager may request a “young, energetic” person when the real need is the ability to meet deadlines and work collaboratively. HR should analyze tasks, outputs, knowledge, skills and working conditions, then separate genuine requirements from preferences.

Advertising and attraction

Advertisements should use neutral, precise language and identify essential duties, location, schedule, pay range where required, application method and available adjustments. Images and wording should not imply that only one type of person belongs in the organization. Recruitment channels should reach diverse communities without targeting or excluding people unlawfully.

Statements such as “recent graduate,” “native speaker,” “culturally fit,” or “must be physically strong” may be problematic unless objectively necessary and carefully defined. “Native speaker” may confuse language proficiency with nationality or ethnicity. A better requirement is a clearly described level of communication ability related to job tasks.

Application screening

Unstructured screening can rely on names, addresses, educational institutions, employment gaps, accents or assumptions about family commitments. These cues may function as proxies for protected characteristics. Where appropriate, organizations can use structured criteria, redact irrelevant personal information, and conduct periodic reviews of screening outcomes.

Employment gaps should not automatically be treated as evidence of poor performance. They may reflect caregiving, disability, illness, migration, redundancy or education. The relevant question is whether the applicant can perform the role and meet its requirements.

Testing and assessment

A test should measure a capability relevant to the job, not merely convenience or familiarity with a particular test format. Assessment instructions should be clear, accessible and consistent. Applicants should know what will be assessed and, where appropriate, how their information will be used. Tests should be validated or at least reviewed for job relevance, reliability and possible group disparities.

Interviews

Interviews are vulnerable to first impressions, similarity attraction, halo effects, confirmation bias and affinity bias. A structured interview uses the same core questions, defined rating scales and job-related evidence for all candidates. Follow-up questions may clarify evidence, but interviewers should avoid irrelevant questions about age, marital status, pregnancy, religion, health or family plans unless a lawful and genuinely necessary process requires a narrowly framed inquiry.

Selection and offer decisions

Decision-makers should record reasons linked to pre-established criteria. A candidate should not be rejected because they seem unlikely to “fit in” unless organizational fit has been translated into observable, job-related behaviors. References, background checks and medical inquiries must be relevant, proportionate, lawful and conducted with appropriate consent.

3. Bias and Adverse Impact

Types of bias

Explicit bias is a conscious preference or prejudice. Implicit bias refers to automatic associations that may affect judgment even when a person endorses equality. Confirmation bias causes an assessor to seek evidence supporting an initial view. Similarity bias favors candidates who share the assessor’s background or interests. Availability bias gives excessive weight to memorable examples. Attribution bias explains success or failure differently depending on the candidate’s group. Stereotype threat describes the risk that awareness of a negative stereotype may impair a person’s performance in an evaluative setting.

Bias cannot be eliminated by telling assessors to “be objective.” It is better managed through job analysis, structured criteria, diverse decision panels, assessor training, independent scoring, documentation and review of outcomes.

Adverse impact

Adverse impact is a statistical or practical indication that a selection procedure disadvantages a group. It is different from intentional discrimination: a neutral procedure can produce adverse impact without discriminatory intent. The organization should then examine whether the procedure is necessary, valid and reasonably replaceable by a less exclusionary method.

A commonly used screening indicator is the selection rate:

For two groups, a selection-rate ratio can be expressed as:

A ratio below 1 indicates a lower rate for one group. In some jurisdictions and professional guidance, the “four-fifths rule” treats a ratio below 0.80 as a warning sign requiring investigation. It is a screening rule, not a universal legal test. It should not be used mechanically when sample sizes are small, selection stages differ, or legal standards require other analyses.

Worked example. Suppose 50 of 100 applicants in Group A are selected and 30 of 100 applicants in Group B are selected. Group A’s selection rate is 50 ÷ 100 = 0.50, or 50%. Group B’s rate is 30 ÷ 100 = 0.30, or 30%. The ratio is 0.30 ÷ 0.50 = 0.60, or 60%. This is a strong signal for review. It does not by itself prove unlawful discrimination. HR should examine the criteria, job relevance, sample size, applicant composition, possible explanations and available alternatives.

The calculation is meaningful only when the groups, stage of selection, time period and denominator are defined consistently. Small numbers can produce unstable percentages, and applicant self-identification may be incomplete. Data should therefore be interpreted carefully and confidentially.

4. Accessibility and Reasonable Adjustments

Accessibility should be built into the process rather than treated as an exception. Job advertisements should be available in accessible digital formats. Application systems should work with assistive technologies, provide keyboard navigation, use adequate color contrast and avoid unnecessary time limits. Interview locations should be physically accessible, and remote interviews should account for captioning, connectivity and communication needs.

An applicant requesting an adjustment should not be required to disclose more medical information than is necessary. HR should ask what support is needed for participation, not seek a diagnosis unless lawful and necessary. Examples of adjustments include:

• extra time or rest breaks in a test;
• captioning, an interpreter or written questions;
• an accessible building or remote interview;
• alternative formats for documents;
• permission to use assistive technology;
• a change in scheduling or communication method.

Adjustments should be handled confidentially and shared only with people who need the information to implement them. Assessors should evaluate the required competence, not the applicant’s use of support.

5. Privacy, Consent and Confidentiality

Recruitment involves personal data, including identity details, contact information, employment history, assessment results, references and sometimes sensitive information. Privacy principles generally require a lawful basis for collection, transparency about purpose, data minimization, accuracy, security, retention limits and respect for individual rights. The exact requirements depend on applicable data-protection law.

Consent is not a substitute for transparency

Consent should be informed, specific, freely given and recorded where consent is the legal basis. However, organizations should not assume that consent makes every practice lawful or ethical. A job applicant may feel unable to refuse a request because of the power imbalance. HR should identify the correct legal basis and provide a clear privacy notice explaining what is collected, why, who receives it, how long it is retained and how applicants can exercise their rights.

Confidentiality and access control

Recruitment information should be accessible only to authorized personnel. CVs should not be left visible, interview notes should not be stored in personal devices, and candidate data should not be discussed casually. References and background reports should be obtained and used according to law and policy. Information about disability, health, religion or other sensitive matters requires heightened care.

Documentation must be sufficient to demonstrate a fair process without recording unnecessary personal commentary. “Poor fit” is weak documentation; “scored 2 of 5 on the published customer complaint scenario because the response omitted escalation and documentation steps” is more useful and defensible.

6. Conflicts of Interest and Ethical Conduct

A conflict of interest exists when personal relationships or interests could influence, or reasonably appear to influence, a recruitment decision. Examples include selecting a relative, close friend, former partner, business associate or person from whom the decision-maker expects a favor. A conflict does not necessarily prove misconduct, but failing to disclose it creates serious risk.

Good controls include:

1. requiring disclosure before assessment begins;
2. removing or recusing the conflicted person where appropriate;
3. appointing an independent decision-maker;
4. using documented criteria and multiple assessors;
5. prohibiting gifts, payments and promises of future benefit;
6. recording the management action taken.

Nepotism and favoritism damage fairness even when the preferred candidate is qualified, because other applicants cannot trust the process. Ethical recruitment also requires honest communication. Organizations should not promise a job, disguise unpaid work, misrepresent salary or working conditions, or collect applications for a role that does not genuinely exist.

7. A Compliant Recruitment Framework

A practical framework is the FAIR process:

• F — Define fair requirements: Conduct job analysis and distinguish essential from desirable criteria.
• A — Access for all: Use inclusive advertising, accessible systems and reasonable adjustments.
• I — Investigate evidence: Apply structured assessment, monitor outcomes and investigate adverse impact.
• R — Respect information and rights: Obtain appropriate consent, protect confidentiality, document reasons and provide respectful communication.

This framework can be implemented through the following process:

Step 1: Establish governance

Identify the responsible HR officer, hiring manager, assessors, data controller or equivalent role, escalation route and applicable legal requirements. Confirm who may access applicant data and how records will be retained.

Step 2: Conduct job analysis

Describe duties, outputs, competencies, working conditions and genuine requirements. Review whether each requirement is necessary and whether an adjustment could enable performance.

Step 3: Design inclusive attraction

Use neutral language, explain the process, publish adjustment routes and advertise through channels likely to reach qualified applicants from varied backgrounds.

Step 4: Build structured assessment

Prepare standardized questions, scoring anchors and evidence requirements. Train assessors to avoid irrelevant questions and to record job-related evidence.

Step 5: Manage data lawfully

Provide a privacy notice, collect only necessary information, secure records and establish retention and deletion rules. Separate equality-monitoring data from selection decisions wherever possible.

Step 6: Assess and decide consistently

Use at least two assessors for important decisions where practicable. Score independently before discussion to reduce group influence. Resolve disagreements by returning to evidence and criteria.

Step 7: Audit outcomes

Review progression and selection rates by relevant groups where lawful and ethically appropriate. Investigate unexpected disparities and document corrective action. Do not publish small-cell data that could identify individuals.

Step 8: Communicate respectfully

Notify candidates of outcomes within a reasonable time. Provide useful feedback where policy and law permit. Handle complaints promptly and without retaliation.

8. Recruitment Technology and Artificial Intelligence

Applicant-tracking systems and AI may improve speed, consistency and accessibility, but they can also reproduce historical inequalities or create opaque decisions. A model trained on previous hiring data may learn that successful past employees came from a narrow group and use irrelevant proxies such as postcode, institution, language patterns or employment history.

Before using an automated tool, HR should ask:

• What decision does the tool make or support?
• What data does it use, and are those data accurate and necessary?
• Has the tool been tested for disparate outcomes and accessibility?
• Can applicants obtain meaningful information about the process?
• Is there meaningful human review and an appeal route?
• Who is responsible when the tool is wrong?
• Are vendors contractually required to protect data and support audits?

Human review is not automatically a safeguard. Reviewers may simply accept an algorithmic recommendation, a phenomenon sometimes called automation bias. A responsible organization validates tools before deployment, monitors them continuously, suspends them when serious disparities appear and avoids using sensitive or proxy variables without a strong lawful and ethical basis.

AI-generated interview summaries also require checking for hallucinations, accent bias, transcription errors and inappropriate inferences. Tools should not infer personality, honesty, health or protected characteristics from facial expression, voice or writing without compelling evidence and legal authorization. In many contexts, such inferences are scientifically weak and ethically unacceptable.

9. Documentation, Audit and Accountability

Documentation serves three purposes: it supports consistent decisions, enables organizational learning and demonstrates compliance. A recruitment file may include the approved job description, advertisement, assessment plan, adjustment arrangements, scores, decision rationale, conflict disclosures and relevant communications. It should not contain gossip, stereotypes or unnecessary sensitive information.

A basic audit can examine:

• applicant numbers and progression at each stage;
• selection rates across relevant groups;
• completion and failure rates for assessments;
• use and timeliness of adjustments;
• complaints and resolution times;
• consistency of interview scores;
• data-access incidents;
• exceptions to standard procedures.

Audits should lead to action, not merely reporting. Possible responses include revising criteria, replacing a biased assessment, improving accessibility, training assessors, widening attraction channels or strengthening oversight. Monitoring must itself respect privacy and anti-discrimination law.

Illustrative Case: The Accessible Analyst Recruitment Exercise

Background

Northbridge Services, a fictional organization, is recruiting a junior data analyst. The manager requests “a young graduate with excellent presentation skills who can work long hours.” The online application includes a timed video interview and a numerical test that cannot be completed using keyboard navigation. The shortlist is selected by one manager, who later reveals that one applicant is a former classmate.

Problem

The language may exclude older applicants and may confuse a preference for a particular career stage with a genuine requirement. “Work long hours” may disadvantage people with caring responsibilities and may not reflect the actual schedule. The inaccessible test may exclude applicants with disabilities. A single decision-maker creates consistency and conflict-of-interest risks. The video interview may disadvantage applicants with limited bandwidth, speech differences or anxiety unrelated to job performance.

Analysis

HR conducts a job analysis and finds that the role requires data accuracy, communication of findings, collaboration and meeting scheduled deadlines. It does not require youth, unrestricted long hours or a specific degree. The numerical test measures relevant reasoning but its format creates an avoidable barrier. The former classmate relationship is a conflict that must be disclosed and managed.

Approach and solution

HR rewrites the advertisement using job-related criteria and states the normal working schedule. It offers an accessible numerical assessment, extra time where appropriate and an alternative live or written communication exercise. The organization publishes a privacy notice and explains how assessment data will be used. Two trained assessors score structured interview answers using anchored criteria. The manager discloses the relationship and does not participate in the final decision. HR records scores and reasons, separates equality-monitoring data from selection records and reviews group outcomes after the campaign.

Outcome

The revised process produces a broader, more relevant shortlist. One applicant uses an adjustment and demonstrates the required analytical competence. The final decision is based on documented evidence rather than age, personal familiarity or performance on an inaccessible format. Northbridge also identifies the need to review accessibility in all online assessments.

Learning points

1. Job-related criteria are more defensible than informal preferences.
2. Accessibility must be designed into assessments.
3. A conflict should be disclosed and managed before decisions are made.
4. Structured scoring improves consistency but does not replace professional judgment.
5. Monitoring and documentation turn individual good intentions into organizational accountability.

Challenges and Limitations

Inclusive recruitment involves difficult judgments. Not every disparity is caused by the selection process; applicant-pool differences, labor-market inequality and occupational segregation may contribute. Conversely, a lack of statistical evidence does not prove that a process is fair, especially when data are incomplete or groups are small.

Organizations may also face tension between privacy and monitoring. Collecting demographic data can help identify disparities but can create risks if poorly secured or used for decisions. The solution is careful purpose limitation, voluntary and lawful collection where appropriate, restricted access and transparent governance.

Legal compliance also varies. Definitions of disability, protected characteristics, positive action, background checks, medical inquiries, retention and automated decision-making differ between jurisdictions. HR professionals should avoid assuming that a practice lawful in one country is lawful elsewhere. Specialist advice is appropriate when the risk is significant.

Finally, training alone has limited effect. A one-hour unconscious-bias session cannot correct poorly designed criteria, inaccessible technology or unchecked managerial discretion. Structural controls are more reliable than relying on individual goodwill.

Best-Practice Checklist

Before opening a vacancy, ask:

• Are the requirements necessary, measurable and job-related?
• Could any requirement create indirect exclusion?
• Is the advertisement clear, truthful and accessible?
• Are adjustment arrangements easy to request?
• Are assessment questions and scoring rules standardized?
• Have assessors disclosed conflicts of interest?
• Is personal data collected transparently and securely?
• Are references, checks and technology lawful, proportionate and relevant?
• Are decisions documented with evidence rather than impressions?
• Will outcomes be monitored and corrective action taken?

Chapter Summary

Inclusive, ethical and legally compliant recruitment gives qualified applicants a fair opportunity to participate and ensures that decisions are based on relevant evidence. Equality concerns protection from unlawful disadvantage; equity may require adjustments or removal of barriers; inclusion concerns meaningful participation. Discrimination can occur in job design, advertising, screening, testing, interviewing and final decisions. Direct and indirect discrimination, harassment, retaliation, bias and conflicts of interest require active controls.

Structured job analysis, accessible processes, standardized assessment, documented criteria, appropriate adjustments, confidentiality and secure data handling provide the foundation of fair recruitment. Selection-rate analysis can identify adverse impact, but numerical indicators are warning tools rather than automatic proof of unlawful conduct. Privacy requires transparency, purpose limitation, data minimization and appropriate security. AI and recruitment technologies must be validated, monitored and subject to meaningful human accountability. Ultimately, ethical recruitment combines lawful governance with respect, honesty, evidence-based judgment and continuous review.

Key Terms

Equality: Protection from unlawful unequal treatment.

Equity: Fair opportunity supported by removal of relevant barriers or provision of adjustments.

Inclusion: Conditions that enable people to participate meaningfully and with dignity.

Protected characteristic: A characteristic protected under applicable equality law.

Direct discrimination: Less favorable treatment because of a protected characteristic.

Indirect discrimination: A neutral rule that disproportionately disadvantages a group without adequate justification.

Harassment: Unwanted conduct that violates dignity or creates a hostile or offensive environment.

Retaliation: Penalizing a person for raising a concern or exercising a legal right.

Bias: A systematic tendency that influences judgment.

Stereotype: A generalized belief about a group.

Adverse impact: A substantially poorer selection outcome for one group.

Reasonable adjustment: A change enabling participation or performance for a person with a relevant need.

Accessibility: Design that allows people with different abilities to use a process or service.

Consent: Informed and voluntary agreement to a specified use of personal information.

Confidentiality: Protection of information from unauthorized disclosure.

Data minimization: Collecting only information necessary for a defined purpose.

Conflict of interest: A personal interest that could improperly influence professional judgment.

Structured interview: An interview using consistent core questions and defined scoring criteria.

Automation bias: Excessive reliance on an automated recommendation.

Job-relatedness: The extent to which a criterion measures a requirement relevant to successful performance.

Review Questions

Short Answer Questions

1. Distinguish equality, equity and inclusion in recruitment.
2. What is the difference between direct and indirect discrimination?
3. State three practical ways to reduce interview bias.
4. Why should recruitment data be collected and retained carefully?
5. What is a conflict of interest, and how should it be managed?

Long Answer Questions

1. Explain how discrimination can enter each stage of the recruitment process and recommend controls.
2. Discuss the role of accessibility and reasonable adjustments in ethical selection.
3. Explain adverse impact and the uses and limitations of the four-fifths rule.
4. Describe a structured recruitment process that protects fairness, privacy and accountability.
5. Evaluate the benefits and risks of using AI in recruitment.

Analytical / Application Questions

1. An assessment is completed successfully by 72 of 120 applicants in Group A and 36 of 90 applicants in Group B. Calculate both selection rates and the selection-rate ratio, then explain what further investigation is required.
2. A hiring manager wants to ask every applicant about childcare arrangements because the team sometimes works late. Analyze the legal, ethical and practical risks and propose a better approach.
3. An applicant requests captioning and additional time for an online assessment. Design a response that protects fairness, privacy, assessment validity and confidentiality.

Multiple Choice Questions with Answers

Question 1

Which practice best supports structured and fair interviewing?

A. Asking each candidate different questions based on first impressions

B. Using common core questions and defined scoring criteria

C. Selecting the candidate who seems most culturally similar to the panel

D. Allowing the interviewer to rely entirely on memory

Correct Answer: B. Using common core questions and defined scoring criteria

Question 2

What is indirect discrimination?

A. A deliberate insult directed at an applicant

B. A neutral rule that disproportionately disadvantages a group without adequate justification

C. A disagreement between two interviewers

D. A candidate receiving an adjustment

Correct Answer: B. A neutral rule that disproportionately disadvantages a group without adequate justification

Question 3

Which is the best example of a reasonable adjustment?

A. Removing all job requirements for one applicant

B. Giving an applicant confidential access to another candidate’s scores

C. Providing captioning or an accessible assessment format

D. Guaranteeing selection before the interview

Correct Answer: C. Providing captioning or an accessible assessment format

Question 4

If 50 of 100 applicants in Group A are selected and 30 of 100 applicants in Group B are selected, what is Group B’s selection-rate ratio relative to Group A?

A. 0.30

B. 0.50

C. 0.60

D. 0.80

Correct Answer: C. 0.60

Question 5

Which statement about the four-fifths rule is most accurate?

A. It automatically proves unlawful discrimination whenever the ratio is below 0.80

B. It is a screening indicator that may signal the need for investigation

C. It applies only to interview performance

D. It replaces job analysis and legal advice

Correct Answer: B. It is a screening indicator that may signal the need for investigation

Question 6

What is the most appropriate response when a hiring manager is assessing a close friend?

A. Keep the relationship private if the friend is qualified

B. Allow the manager to decide but require a short explanation

C. Disclose and manage the conflict, potentially through recusal

D. Automatically reject the friend

Correct Answer: C. Disclose and manage the conflict, potentially through recusal

Question 7

Which principle requires an organization to collect only information necessary for a defined recruitment purpose?

A. Data minimization

B. Similarity attraction

C. Stereotype threat

D. Positive favoritism

Correct Answer: A. Data minimization

Question 8

Why can historical hiring data create risk in an AI recruitment tool?

A. Historical data are always illegal to use

B. The tool may learn and reproduce previous patterns of inequality

C. AI tools cannot process numerical information

D. Human review is prohibited when AI is used

Correct Answer: B. The tool may learn and reproduce previous patterns of inequality

Question 9

Which interview note is most useful and defensible?

A. The candidate was not impressive

B. The candidate did not fit the team

C. The candidate scored 2 of 5 because the response omitted escalation and documentation steps

D. The candidate seemed unusual

Correct Answer: C. The candidate scored 2 of 5 because the response omitted escalation and documentation steps

Question 10

What should an organization do after finding an unexpected disparity in selection outcomes?

A. Ignore it unless a complaint is filed

B. Delete the data to protect privacy

C. Investigate the criteria, validity, sample and possible alternatives

D. Automatically appoint candidates from the lower-selection group

Correct Answer: C. Investigate the criteria, validity, sample and possible alternatives