Ethics and AI: The 2020 International Baccalaureate Grading Scandal Custom Case Solution & Analysis

1. Evidence Brief: Case Data Extraction

Financial Metrics

• Annual Revenue: The International Baccalaureate (IB) operates as a non-profit foundation with annual revenues exceeding 200 million dollars, primarily derived from examination fees and school authorizations.
• Examination Fees: Schools pay approximately 119 dollars per student per subject, plus an additional base fee per candidate.
• Refund Liability: Following the cancellation of May 2020 exams, the IB faced pressure to refund fees for 174,410 registered students.
• Operational Reserves: The organization maintains significant cash reserves to manage cyclical fluctuations in exam cycles.

Operational Facts

• Student Population: 174,410 students across 146 countries were affected by the 2020 exam cancellations.
• The Algorithm Components: Final grades were calculated using three data streams: 1. Student coursework (Internal Assessments), 2. Teacher-predicted grades, and 3. Historical school performance data from the previous three years.
• Historical Weighting: A school-specific factor was applied to adjust predicted grades based on how accurately that school predicted results in the past.
• Result Distribution: Upon release on July 6, 2020, approximately 25 percent of students received grades significantly lower than their predicted marks.

Stakeholder Positions

• Siva Kumari (Director General): Positioned the algorithm as a necessary tool to maintain the global standard and comparability of the IB diploma during the pandemic.
• Students and Parents: Initiated petitions with over 20,000 signatures within 48 hours of results release, alleging the algorithm was discriminatory and penalized high-achieving students in historically low-performing schools.
• University Admissions Officers: Required final marks by mid-August for enrollment; expressed concern over the volatility of results compared to previous years.
• Ofqual (UK Regulator): Initially supported algorithmic approaches but later shifted stance following the similar A-Level grading crisis in the United Kingdom.

Information Gaps

• Weighting Ratios: The exact percentage weight assigned to historical school performance versus individual student work remains undisclosed.
• Algorithmic Audit: No evidence is provided regarding third-party ethical audits of the code prior to deployment.
• Appeal Cost: The specific financial cost to the IB for processing the surge in appeals is not quantified.

2. Strategic Analysis

Core Strategic Question

• How can the IB preserve the long-term institutional value of its certification while resolving an immediate crisis of fairness and transparency?
• Can a standardized assessment body maintain global comparability when the primary data source (standardized exams) is removed?

Structural Analysis

The IB operates in a high-stakes credentialing market where its primary value proposition is rigor and global consistency. The 2020 crisis disrupted the Value Chain at the assessment stage. By substituting individual performance with a statistical proxy (historical school data), the IB shifted from an achievement-based model to a probability-based model. This created a structural misalignment between the organization and its primary customers—students—who expect to be judged on personal merit, not the performance of their predecessors.

Strategic Options

Preliminary Recommendation

The IB must adopt Full Reversion to the higher of the predicted grade or the algorithmic grade. While this results in grade inflation, institutional survival depends on student and university trust. Statistical purity is a secondary concern when the primary data collection mechanism (exams) has failed. The IB cannot win a fight against the perception of systemic bias.

3. Implementation Roadmap

Critical Path

• Immediate (Day 1-5): Announce a policy shift to accept the higher of the teacher-predicted grade or the calculated mark. This stops the immediate PR hemorrhage.
• Short-term (Day 6-15): Establish a direct data link with major university clearinghouses (UCAS, Common App) to update student transcripts electronically.
• Mid-term (Day 16-45): Process revised diplomas for all 174,410 students to ensure physical documentation matches updated digital records.
• Long-term (Day 46-90): Commission an independent ethics review of algorithmic decision-making to prevent future systemic bias.

Key Constraints

• University Deadlines: The hard stop is the late-August enrollment window. Any implementation that exceeds this timeframe is a failure.
• Data Discrepancy: Reconciling differences between teacher predictions and Internal Assessments requires a simplified, automated rule-set to avoid manual review of 170,000 files.
• Legal Compliance: GDPR and other data privacy regulations in multiple jurisdictions limit how student data can be re-processed and shared.

Risk-Adjusted Implementation Strategy

Execution must prioritize speed over precision. The organization should deploy a dedicated task force for school-level communication to manage the volume of inquiries. Contingency plans must include a financial reserve for potential legal settlements if the revised grades still fail to secure university spots for students in specific regions.

4. Executive Review and BLUF

BLUF (Bottom Line Up Front)

The IB must immediately abandon its statistical model and award students the higher of their predicted grade or their algorithmic grade. The 2020 grading scandal was not a technical failure but a failure of institutional judgment. By prioritizing statistical distribution over individual student achievement, the IB violated its core mandate of equity. Institutional credibility is currently at a breaking point. Only a total reversion to teacher-led assessments can prevent a mass exodus of schools and long-term devaluation of the IB brand. Speed is the only metric that matters; university deadlines do not wait for administrative precision.

Dangerous Assumption

The single most dangerous assumption was that historical school performance is a valid predictor of individual student potential in a crisis year. This premise assumes that student cohorts are static and that school environments do not change, effectively punishing high-performers in improving schools and codifying socioeconomic disparity into a mathematical formula.

Unaddressed Risks

• Regulatory Retaliation: National education ministries may move to decertify the IB as a recognized equivalent to local diplomas if the perceived bias is not corrected, leading to a permanent reduction in market share. (Probability: Medium; Consequence: Critical)
• Teacher Alienation: By discounting teacher-predicted grades via a historical multiplier, the IB has signaled a lack of trust in its primary operational partners—the educators. This may lead to reduced school engagement and lower quality of future internal assessments. (Probability: High; Consequence: High)

Unconsidered Alternative

The team failed to consider a Deferred Examination Window. The IB could have offered a free, accelerated exam session in August or September for students unhappy with their algorithmic results. While logistically difficult, this would have provided an objective, merit-based alternative to the controversial statistical model while maintaining the rigor the IB brand claims to represent.

Verdict

APPROVED FOR LEADERSHIP REVIEW