Assessment Criteria A–E

The CS Extended Essay is marked out of 30 across five criteria. Examiners use positive marking with a best-fit approach – they pick the descriptor band that most accurately captures your work as a whole. Awarding the top mark in a band does not require a faultless essay.

Criterion	Marks	Where it is assessed
A: Framework for the essay	6	The essay
B: Knowledge and understanding	6	The essay
C: Analysis and line of argument	6	The essay
D: Discussion and evaluation	8	The essay
E: Reflection	4	500-word reflective statement on the RPF

If your work does not reach the standard described by the lowest band of any criterion, 0 marks are awarded for that criterion.

Criterion A: Framework for the Essay (6 marks)

Guiding question: Do the research question, research methods and structural conventions provide an effective framework for the essay?

Strand	1–2 marks	3–4 marks	5–6 marks
Research question	Stated but lacks relevance to the topic, clarity or focus	Relevant and clear but only partially focused in relation to the scope	Relevant, clear and focused in relation to the scope
Research methods	Used but mostly unsuitable for the research question	Mostly suitable, explained and applied with partial effectiveness	Suitable, explained and applied effectively
Structure	Conventions present but do not support communication	Support some aspects of the communication	Effectively support communication of the research

What this means for CS

The RQ should explore a computational principle, not a general technology issue.
Methods should be grounded in CS practice: simulations, algorithm testing, dataset analysis, structured literature review.
Variables and parameters must be fine-tuned – you should be able to defend why each one was chosen.
Standard structure: introduction, background, methodology, results, discussion, conclusion.

Quick checks

Does the introduction explain why the topic matters and preview how the question will be answered?
In CS Sciences essays the methodology conventionally follows the background. If yours sits elsewhere, can you explain why the structure helps the reader?
Could a reader who only read your RQ predict roughly how the essay would be structured?

Criterion B: Knowledge and Understanding (6 marks)

Guiding question: Does the student demonstrate knowledge and understanding of the subject matter being used in their research?

Strand	1–2 marks	3–4 marks	5–6 marks
Knowledge	Research materials used but lack relevance	Relevant materials establish some knowledge	Comprehensive, relevant materials establish knowledge
Understanding – Terminology	Used but lacks accuracy or is very limited	Some relevant terminology used accurately	Relevant terminology used accurately and consistently
Understanding – Concepts	Identified but not relevant or explained superficially	Some relevant concepts explained with partial effectiveness	Relevant concepts explained and used effectively

What this means for CS

Go past basic definitions. Apply concepts – do not just define them.
Core CS knowledge that should appear: algorithm complexity, data structures, logic, plus specialised knowledge for your topic (cryptographic primitives, neural-network architectures, network protocols, etc.).
Mathematical expressions must be interpreted in computing terms – a Big-O analysis is not just a formula, it is a claim about how the algorithm behaves at scale.
Source quality is part of this criterion. Blogs, YouTube, and Wikipedia as primary sources will hurt your mark even if your writing is strong.

Quick checks

Can you point to a concrete piece of recent research (not a textbook) that you engaged with?
For every algorithm or concept you describe, can you also explain its trade-offs or failure modes?

Criterion C: Analysis and Line of Argument (6 marks)

Guiding question: Does the student analyse the information and produce a coherent line of argument?

Strand	1–2 marks	3–4 marks	5–6 marks
Analysis	Descriptive rather than analytical	Partially effective, produces some relevant findings	Effective, consistently produces relevant findings
Line of argument	Partial	Partially consistent, links the research question, findings and conclusions	Clear, sustained, links the research question, findings and conclusions

What this means for CS

Interpret results – explain what they reveal about the system or algorithm.
Examine patterns, trade-offs, anomalies, and limitations using sound technical reasoning.
Use defined evaluation criteria (accuracy, speed, scalability, resource efficiency) for meaningful comparison.
Listing performance data without interpretation is not analysis.
Compare your findings with existing literature.

Description vs. analysis: the single most common gap

This is the difference between a 4/6 and a 6/6 on this criterion.

Description (avoid): “Algorithm X completed in 0.3 seconds while Algorithm Y took 1.2 seconds, so X was faster.”

Analysis (aim for): “Algorithm X outperformed Y by a factor of 4 in this configuration, consistent with its theoretical O(n log n) advantage on partially sorted inputs. The gap narrows as input randomness increases (Figure 3), suggesting that X’s optimisation for ordered runs – not raw asymptotic efficiency – drives the performance difference.”

A useful drill when reviewing your draft: highlight every paragraph where you describe data without interpreting it, and write “so what?” in the margin. Every “so what?” is a missed analysis opportunity.

Criterion D: Discussion and Evaluation (8 marks)

Guiding question: Does the student discuss the findings and evaluate the essay?

This is the highest-weighted criterion, and the one where students most often fall short of the top band.

Strand	1–2 marks	3–4 marks	5–6 marks	7–8 marks
Discussion	Significance described, no supporting evidence	Partially balanced, sometimes supported by evidence	Balanced, often supported by evidence	Balanced, fully supported by appropriate evidence
Evaluation	Partial, strengths and limitations not stated	Present, some strengths and limitations stated	Present, relevant strengths and limitations described	Present, relevant strengths and limitations explained

What this means for CS

Discussion examines significance and implications, not just describes results.
Link outcomes back to the RQ and to a broader computing context.
Evaluation is integrated into the discussion – not a separate “Evaluation” section tacked on at the end.
Assess the reliability of your data; consider alternative interpretations.
Reflect on the suitability of your methods: dataset choice, parameters, evaluation metrics.
Compare your results with existing literature – do they agree, disagree, or extend prior findings?
Acknowledge limitations. Suggestions for further research should derive from observations, not be tacked on randomly.

Common features of 7–8 mark essays

The rubric only requires “balanced, fully supported by appropriate evidence” plus strengths and limitations explained. In practice, top-band CS essays tend to share most of these features:

Discussion is balanced and fully supported by appropriate evidence.
Strengths and limitations are explained – why each one matters and how it affects conclusions – not just stated.
Cross-referencing with existing literature. The student says explicitly whether their results agree or disagree with at least one published study, and why.
Broader implications are discussed. What do the findings mean beyond this specific experiment?
Methodological reflection. The student says what they would do differently and why it would produce better results.

Treat the list as a self-check, not a checklist – an essay can score in the 7–8 band without ticking every item, and ticking every item does not guarantee 7–8.

Criterion E: Reflection (4 marks)

Guiding question: Does the student evaluate the effect of the extended essay learning experience on them as a learner?

This criterion is assessed only on the 500-word reflective statement on the RPF.

1 mark	2 marks	3 marks	4 marks
Reflection is descriptive; limited evidence of growth	Reflection includes some evaluative elements; some evidence of growth	Reflection is often evaluative with specific examples; shows growth and some transfer of learning	Reflection is consistently evaluative with specific examples; shows consistent evidence of growth and transfer of learning

What examiners look for

Reflection is consistently evaluative, not chronological.
Specific examples, not vague generalisations.
Evidence of growth – how you developed as a researcher, thinker, or writer.
Evidence of transfer of learning – how you can apply these skills in another context.
CS-specific reflection – not just generic study skills like time management.

If the RPF is not submitted, is blank, or is in a different language from the essay, 0 marks are awarded.

See the Reflection page for how the three reflection sessions feed into this statement, and what to write (and not write).

Assessment objectives mapping

Criterion	Assessment objective(s)
A: Framework	Knowledge and understanding; Communication of research
B: Knowledge and understanding	Knowledge and understanding
C: Analysis and line of argument	Application and analysis; Synthesis and evaluation
D: Discussion and evaluation	Synthesis and evaluation
E: Reflection	Synthesis and evaluation

EE + TOK points matrix

The EE grade (A–E) combines with your TOK grade (A–E) to contribute up to 3 points toward your IB diploma total.

	TOK A	TOK B	TOK C	TOK D	TOK E or N
EE A	3	3	2	2	Failing condition
EE B	3	2	2	1	Failing condition
EE C	2	2	1	0
EE D	2	1	0	0
EE E or N	Failing condition

A grade of E in either EE or TOK is a failing condition for the diploma.

N in the matrix means no grade awarded – the work was not submitted, was withdrawn, or was deemed inadmissible (for example, on academic-integrity grounds). For the purposes of the matrix, treat it as worse than E.