IB Predicted Grades Best Practices for Reducing Bias

The strongest IB predicted grades best practices replace informal impressions with a shared, evidence-based process. Coordinators should define acceptable evidence, align teachers with IB standards, moderate borderline decisions, document judgments, and audit accuracy after results are released.

The IB defines a predicted grade as the teacher's prediction of the grade a candidate is expected to achieve, based on available evidence and knowledge of IB standards. Reducing grading bias does not mean removing teacher judgment. It means structuring that judgment so irrelevant factors have less influence.

Why predicted-grade bias matters

Predicted grades can affect university applications, counselling decisions, student expectations, and confidence in the school's assessment system. Even conscientious teachers may be influenced by behaviour, participation, confidence, prior reputation, or assumptions about a student's background.

Research outside the IB context has found that prediction errors are not always distributed evenly. UK university-admissions research, for example, found that some high-attaining disadvantaged students were more likely to receive pessimistic predictions. This finding cannot automatically be transferred to every IB school, but it shows why coordinators should examine outcomes rather than assume good intentions guarantee fairness.

Bias must be distinguished from ordinary prediction error. Students may exceed reasonable predictions after rapid improvement or perform below them because of illness or examination pressure. Bias becomes a concern when irrelevant characteristics or recurring procedural differences systematically affect judgments.

Establish a school-wide predicted-grade policy

A written policy should state the purpose of each prediction, acceptable evidence, timelines, moderation expectations, documentation requirements, and the process for reviewing concerns. It should distinguish concepts that are frequently confused:

A prediction is not a reward for effort, a motivational target, or the highest grade a student might conceivably achieve. University-facing forecasts may also be produced at a different time from predictions submitted to the IB. Schools need a transparent policy explaining how these processes relate. RevisionDojo's guide to when and how IB predicted grades are determined can support communication with families.

Standardize the evidence used

Departments should use a common evidence framework while retaining subject-specific judgment. Identical percentages for every subject would be misleading because assessment structures differ, but allowing each teacher to select entirely different evidence creates inconsistency.

Teachers should normally consider:

• recent timed assessments covering representative syllabus content
• mock examinations under controlled conditions
• relevant internally or externally assessed coursework
• performance across applicable assessment objectives
• patterns of achievement rather than one unusual result
• sustained improvement that has already been demonstrated

Parallel classes should use comparable tasks, conditions, permitted resources, and marking processes where possible. RevisionDojo's guidance on timed IB mock examinations can help departments design representative practice, although third-party results should support rather than replace teacher judgment.

Schools should not automatically convert one mock percentage into a predicted grade. Final IB boundaries are determined using examiner judgment, candidate work, statistics, and grade descriptors. Historical boundaries provide context, but they are not guaranteed future thresholds.

Anchor judgments in IB standards

Teachers should use the current subject guide, assessment objectives, markschemes, and official grade descriptors. The IB's Diploma Programme grade descriptors describe characteristics of performance and are intended to support grade prediction.

Departments should maintain secure, anonymized reference materials where possible, including assessed work at different levels, annotated borderline examples, previous moderation adjustments, and relevant examiner feedback. These materials calibrate judgment rather than provide templates for students to reproduce. They must be updated when subject guides or assessment models change.

Moderate grades before final approval

Moderation works best when teachers judge evidence independently before discussion. If a senior teacher announces a grade first, colleagues may anchor on that opinion.

A practical moderation cycle is:

1. Remove names where feasible and lawful.
2. Give two teachers the same evidence and prediction template.
3. Record independent provisional grades and rationales.
4. Resolve differences using criteria and work samples.
5. Refer unresolved cases to the department head or coordinator.
6. Record the final decision and decisive evidence.

Anonymization can reduce the influence of reputation or personality, but it is not always possible for oral work, supervised coursework, or small classes. It should be one control within a broader process. Coordinators can also sample files across departments to confirm that each subject follows the school-wide policy.

Separate attainment from behaviour and potential

Participation, punctuality, diligence, and homework completion may support learning, but they are not automatically evidence of the achievement measured by IB assessments. A quiet student should not be underpredicted because they speak less, while a confident student should not receive a higher prediction merely because they appear academically assured.

Prediction templates should separate demonstrated assessment evidence, relevant objectives, recent trends, uncertainty, and final professional judgment. If improvement appears across several valid assessments, it belongs in the evidence. If it is only hoped for, it should inform support planning rather than inflate the prediction.

Review difficult cases consistently

Schools should define triggers for additional review, such as assessor disagreement, incomplete evidence after documented disruption, a prediction above the established pattern, or performance close to a likely boundary.

The same opportunity must be available to every student. Assertive parents or students should not gain access to reconsideration that others do not know exists. A published predicted-grade review process should clarify deadlines, eligible evidence, decision-makers, and the difference between procedural review and negotiation.

Reasonable adjustments and access arrangements enable students to demonstrate achievement fairly. They do not justify raising or lowering academic standards.

Audit predictions after final results

Post-results analysis turns prediction into an improvement cycle. For each candidate, calculate predicted grade minus final grade. A positive average indicates overprediction, while a negative average indicates underprediction.

Coordinators should also examine exact agreement, agreement within one grade, and mean absolute error because positive and negative differences can cancel out. Review patterns by subject, teacher, level, prediction point, and evidence type. Where lawful and ethical, aggregated comparisons across relevant student groups may reveal uneven errors, but schools should protect confidentiality and avoid conclusions from very small samples.

The IB assessment overview states that the organization works with schools that consistently underpredict or overpredict. Internal auditing helps coordinators identify tendencies early and respond through calibration, training, or revised assessment design.

Coordinator checklist

Before approving predictions, confirm that:

• departments used current IB standards
• evidence was recent, representative, and obtained under known conditions
• behaviour was not substituted for attainment
• common assessments were marked consistently
• borderline cases received documented moderation
• reviews were equally accessible
• university pressure did not determine grades
• rationales were securely recorded
• post-results analysis is scheduled

A predicted-grade calculator guide may support component weighting and indicative calculations. No calculator can know future boundaries or replace professional judgment.

Conclusion

Reducing bias in IB predicted grades requires a disciplined process rather than confidence in intuition. Coordinators should establish shared rules, use representative evidence, apply IB standards, moderate difficult decisions, provide equitable reviews, and compare predictions with final outcomes.

RevisionDojo can support this framework through Questionbank tasks, Mock Exams, assessment feedback, and Grade Predictor tools. These resources work best as sources of comparable evidence within a coordinator-led policy, not as substitutes for official standards or teacher judgment.

Sources and referenced URLs

• IB student assessment and predicted-grade guidance
• Official IB Diploma Programme grade descriptors
• Official IB assessment guide for teachers and coordinators
• UK government review of teacher-assessment bias
• UCL research on predicted grades and admissions
• Research on objective criteria and judgmental bias
• RevisionDojo guide to how predicted grades work
• RevisionDojo guide to IB mock examinations
• RevisionDojo predicted-grade calculator guide
• RevisionDojo predicted-grade review guide
• RevisionDojo guide to communicating with parents