Overview
A 7 in IB Computer Science comes from two things done consistently: writing theory answers in precise computer-science terminology, and being able to design, trace, and reason about algorithms rather than just describe them. It is not a memorisation subject and it is not a pure programming subject — it is logic-and-problem-solving hard *plus* written-theory precise. This guide shows how the papers and the internal assessment are marked, and where the marks that separate a 5 from a 7 actually sit. (Weightings and paper details shift between syllabus versions, so confirm the specifics with your current subject guide and teacher.)
What a 7 actually takes
Grade boundaries move slightly each session, so chasing a fixed percentage misses the point. What matters is thinking in markbands and mark points. Structured theory questions are marked point-per-mark — a three-mark answer usually needs three distinct, correct ideas expressed in the right technical language — while extended responses are placed in a level descriptor band. Two students can "know" the same topic and score very differently: one writes the exact mechanism and correct term, the other writes a fuzzy paraphrase that earns nothing.
So the real question is never "did I write enough?" — it is "does this answer contain the specific points, in the examiner's language, that the mark scheme rewards?" Computer Science sits in Group 4 (the sciences), and its core spans system fundamentals, computer organisation, networks, and computational thinking and programming. Top candidates move fluently between explaining a concept in words, reasoning about a trade-off, and reading or writing pseudocode — because all three appear in the written papers.
The assessment and exactly how each part is marked
Computer Science is assessed by external papers plus the internal assessment. SL sits the theory papers plus options content; HL adds HL-only theory and an extra paper built on a pre-released case study. Both levels complete the same programming IA.
| Component | Level | What it tests | Where 7s are won |
|---|---|---|---|
| Paper 1 | SL & HL | Core theory: systems, organisation, networks, computational thinking | Precise terminology; correct pseudocode/algorithm logic |
| Paper 2 | SL & HL | Options / applied content | Applying concepts, not restating definitions |
| Paper 3 | HL only | Annual pre-released case study | Justified recommendations with genuine trade-offs |
| IA (computational solution) | SL & HL | A working program for a real client | Meeting each criterion, especially design and evaluation |
Paper 1 is where precision matters most. The core topics reward exact vocabulary — a "cache" is not "some fast memory", and an examiner marking against the scheme wants the specific term and its specific role. Algorithm questions frequently ask you to trace, complete, or reason about pseudocode, so the ability to run code in your head with a trace table is a core scoring skill.
Paper 2 turns the concepts into applied questions. The candidates who score highest here apply ideas to the scenario rather than re-explaining the textbook definition.
Paper 3 (HL only) is built on an annual pre-released case study. You get the material before the exam, and the marks go to candidates who have researched the underlying technology deeply and can justify recommendations by weighing trade-offs, not by listing features. See the full breakdown in the IB Computer Science HL past papers guide and, for the other level, the SL past papers guide.
Writing theory answers that actually score
For the IB Diploma Programme, this is the single biggest differentiator between a 5 and a 7 on the theory papers. Two habits are non-negotiable:
- Use precise technical terminology. Mark schemes reward the correct term used correctly. Learn the exact vocabulary for each core topic and use it deliberately — "the operating system schedules processes" beats "the computer decides what to run".
- Answer the command term, not the topic. Outline, describe, explain, evaluate, and discuss each demand a different depth. Writing a description when the question says evaluate leaves marks on the table however correct the content. Study IB command terms explained so each term triggers the right response.
A reliable approach on structured questions: identify the command term, count the marks, and make sure your answer contains at least that many distinct, correctly-termed points — including the mechanism where the question asks how or why something works.
Algorithms, pseudocode, and trace tables
For the IB Diploma Programme, pseudocode and programming logic appear directly in the written papers, so this is technique you must drill until it is automatic. Marks are lost when students can read code but cannot reliably *trace* it under time pressure.
Core skills to practise:
- Trace tables. Step through a loop line by line, tracking every variable's value at each iteration. This is the fastest way to answer "what does this output?" questions correctly.
- Reading and completing pseudocode. Recognise the standard constructs and fill gaps so the logic works — off-by-one errors and misread loop conditions are the usual mark-losers.
- Reasoning about efficiency and correctness. Explain why an algorithm works or where it fails, using the correct terminology.
Students who can already program tend to find both this and the IA smoother, because the mental model of "run the code in my head" is already there. If programming is newer to you, deliberate trace-table practice closes the gap quickly. Work through these topic by topic in the free Computer Science SL course or HL course.
The IA: a working computational solution
Both SL and HL complete the same IA — you design, build, and evaluate a working program for a real client with a genuine problem. It is criterion-marked, and because the criteria are published, it represents accessible marks too many students leave until the last minute. The biggest self-inflicted wounds are feature creep (building more than you can document and evaluate well) and a thin evaluation that does not test the product against the client's actual success criteria. Get a real client and clear criteria agreed early, keep the scope tight, and document design decisions as you go. Our [IB Computer Science IA guide](/blog/ib-computer-science-ia-guide) walks through each criterion.
Common mistakes that cap you at a 5
This section covers Common mistakes that cap you at a 5 — what IB examiners reward most often in past papers and coursework.
- Vague, non-technical answers. Fuzzy paraphrase scores nothing when the scheme wants the precise term.
- Misreading the command term — describing when the task says evaluate or discuss.
- Weak algorithm tracing — losing "what does this output?" marks that a trace table would secure.
- Paper 2/3: restating definitions instead of applying concepts to the scenario or case study.
- HL: under-preparing the case study — treating it as light reading rather than deep research into the trade-offs.
- IA feature creep and a shallow evaluation against the client's criteria.
A weekly LEARN → PRACTICE → GET-MARKED study system
For the IB Diploma Programme, turn revision into a repeatable loop rather than passive rereading:
- LEARN — take one core syllabus point, study the theory, and lock the exact terminology and any relevant algorithm. Use the free Computer Science SL course or HL course to work syllabus-by-syllabus.
- PRACTICE — do a timed past question on that point: a Paper 1 theory/algorithm question, a Paper 2 applied question, or (HL) case-study preparation. Trace any code by hand.
- GET-MARKED — mark it against the scheme, then get an answer marked for a second opinion aligned to the criteria. Keep a mistake log of your top three recurring errors and drill those first next week.
Run this cycle across all core topics, weighting the ones where your marking is weakest, and pull past questions from the Computer Science HL past papers or SL past papers.
How MarkScheme helps
Self-marking against the scheme is essential, but extended theory and IA drafts benefit from an outside read. After a past paper or IA section, [get criterion-based feedback](/mark) mapped to IB assessment objectives — the same habit that lifts exam scripts also sharpens coursework. Pair it with the free [Computer Science SL](/ib/courses/computer-science-sl) and [HL](/ib/courses/computer-science-hl) courses, and browse everything from the [IB guides hub](/guides/ib).
Frequently asked questions
This section covers Frequently asked questions — what IB examiners reward most often in past papers and coursework.
Is a 7 harder at HL than SL?
The core theory and IA skills are the same, but HL adds HL-only content (such as OOP, abstract data structures, resource management, and control) and the Paper 3 case study. An HL 7 needs that extra material mastered on top of the same precise theory technique. See IB Computer Science SL vs HL.
Do I need to be a strong programmer to get a 7?
It helps — students who can already program tend to find the algorithm questions and the IA smoother. But the theory papers reward precise written answers as much as coding, so disciplined trace-table practice and terminology drilling can close the gap if programming is newer to you.
How much does the HL case study matter?
A lot — Paper 3 is built entirely on the annual pre-released case study. Candidates who research the underlying technology deeply and can justify recommendations with trade-offs score highest. Check your current subject guide for exactly how it is weighted.
What is the most common reason students lose marks?
Vague, non-technical answers and misreading the command term. Mark schemes want distinct points in exact CS vocabulary; general statements score nothing. Point-per-mark practice fixes this fast.
When should I start the IA?
Early. It is the same computational-solution task at both levels, it is criterion-marked, and it is easy to underestimate. Agree a real client and clear success criteria as soon as you can, and keep the scope tight to leave time for a strong evaluation.