You are viewing a free preview of this lesson.
Subscribe to unlock all 4 lessons in this course and every other course on LearningBro.
Component 01 is the theory paper of OCR A-Level Computer Science. It rewards precise technical knowledge expressed in examiner-friendly language, the ability to perform data-representation calculations under time pressure, and — crucially — the discipline to give an answer that matches the command word and the marks available. This lesson takes you inside the paper: what each section assesses, how the assessment objectives are tested across the question styles, the calculation methods that recur, and a fully worked Specimen Question with banded model answers that show exactly how to climb from a Mid-band to a Top-band response. Knowing the paper this well is itself a revision technique: every mark you understand is a mark you stop giving away.
A theme runs through everything below. The content of H446/01 is taught in the sibling courses listed in the Spec Mapping; this lesson assumes you have learned that content and concentrates on the layer above it — the conversion of knowledge into marks. Two candidates can know the same facts and score very differently, and the gap between them is almost entirely the technique covered here: reading the command word, sizing the answer to the tariff, applying knowledge to the scenario, and structuring a levels-marked response so its reasoning is visible to the examiner.
This lesson covers the assessment of Component 01, "Computer Systems" (H446/01), which draws on specification sections 1.1 to 1.5:
The paper is synoptic: a single extended-response question can pull a hardware idea (1.1), a networking trade-off (1.3) and an ethical consequence (1.5) into one argument. The point of this lesson is the assessment technique layered on top of that content — how marks are won and lost, paraphrased throughout rather than quoted from the specification.
Every mark on H446/01 is tied to one of three assessment objectives (AOs). Knowing which AO a question targets tells you what kind of writing earns the marks.
| AO | What it credits | Where it dominates on Paper 1 | What a strong answer looks like |
|---|---|---|---|
| AO1 — knowledge and understanding | Recalling and explaining facts, terms and processes | Short-answer "state/define/describe" questions | Correct technical vocabulary; the right number of distinct points for the marks |
| AO2 — application | Applying knowledge to a given context or scenario | Calculations and "explain how X applies here" questions | Knowledge bent to the specific scenario in the stem, not a generic textbook answer |
| AO3 — analysis and evaluation | Making and justifying reasoned judgements | Levels-marked "discuss/evaluate" extended responses | Two or more weighed viewpoints leading to a justified conclusion |
Key Point: Roughly half of the paper credits AO1, with the remainder split between AO2 application (notably the calculations) and AO3 evaluation (the extended responses). A candidate who only revises facts can cap out in the middle of the paper; the AO2 and AO3 marks are won by technique, which is exactly what is trainable in the final weeks.
Paper 1 is a written exam (pen and paper), 2 hours 30 minutes, 140 marks, 40% of the A-Level. There is no choice — you answer every question. Questions are structured (multi-part), generally rising in demand from part (a) to the final part, and the paper builds from short recall toward extended evaluation.
| Section | Topic area assessed | Sibling course |
|---|---|---|
| 1.1 | Processor structure/function, processor types, I/O and storage | Processors and Hardware |
| 1.2 | Operating systems, software development, language types and translation | Software and Systems |
| 1.3 | Compression/encryption/hashing, databases, networks, web technologies | Networks; Databases and Ethics |
| 1.4 | Data types, number/text/image/sound representation, structures and algorithms | Data Representation; Data Structures; Boolean Algebra |
| 1.5 | Legislation and the moral/ethical/social/cultural impact of computing | Databases and Ethics |
These open most questions: state, define, identify, give an example. The mark tariff tells you exactly how many distinct points to make.
Worked micro-example (2 marks): State two characteristics of a RISC processor.
The trap here is repetition — "it has fewer instructions" and "it has a reduced instruction set" are the same point and score once. For a 2-mark question you need two different characteristics.
Describe asks what happens; explain additionally asks why. The single biggest cause of dropped marks on Paper 1 is answering describe when the question said explain, or vice versa.
Worked micro-example (4 marks): Describe how the fetch stage of the fetch-decode-execute cycle uses the processor's registers.
The Program Counter holds the address of the next instruction (1). That address is copied to the Memory Address Register and placed on the address bus (1). The instruction at that address is fetched into the Memory Data Register via the data bus (1), then copied into the Current Instruction Register ready for decoding, and the PC is incremented (1). Notice the use of register names — this content comes straight from the Processors and Hardware course and the names are what convert a vague sentence into a credited point.
Data-representation and networking calculations recur in every sitting. Method marks are available even when the final answer is wrong, but only if your working is visible. Always write the rule, the substitution, the intermediate values and the final answer with a unit.
You must address both items and use comparative connectives (whereas, in contrast, by comparison). A list of facts about A followed by a separate list about B is not a comparison and is marked down. Pair every point: "RISC uses many simple instructions, whereas CISC uses fewer, more complex ones."
These are marked by level (a band), not by counting points. To reach the top band you need balanced analysis — at least two genuinely different viewpoints — supported by accurate technical detail, leading to a justified conclusion. Section 1.5 (legal/ethical) and technology trade-offs (e.g. SSD vs HDD, TCP vs UDP) are common sources. The banded Specimen Question below shows what each level actually looks like on paper.
It helps to picture the three bands before you write, because each one describes a kind of answer, not a quantity of facts:
| Level | Character of the response | What is usually missing if you are stuck below it |
|---|---|---|
| Level 1 (lower) | A few relevant but undeveloped points; little or no application; one-sided | No "why"; no second viewpoint |
| Level 2 (middle) | Mostly accurate and applied, but the evaluation is asserted rather than reasoned | The trade-off is listed but not weighed |
| Level 3 (upper) | Technically grounded, two-sided, applied to the scenario, with a justified conclusion | (Top band) |
The practical consequence is that you cannot "grind" your way up the bands by adding more facts — a wall of accurate AO1 content with no balance and no judgement stays in Level 1 or 2. You move up by adding a second viewpoint and a weighed verdict. Plan those two things in the margin first.
Explain questions sit between recall and evaluation: they demand causation. Consider a 4-mark question — Explain why a solid-state drive generally has a faster access time than a magnetic hard disk drive.
A weak answer says "because it is faster and has no moving parts" — which restates the claim rather than explaining it. A credit-worthy answer chains cause to effect: an HDD must physically move a read/write head to the correct track (seek time) and wait for the platter to rotate the data under the head (rotational latency) (1), both of which are mechanical delays (1); an SSD addresses flash cells electronically with no mechanical movement (1), so there is no seek or rotational delay and the access time is far lower (1). Each "(1)" is a linked step, which is what explain rewards over describe.
The exact split varies between series, but the broad shape is stable. Use this to triage revision time, not as a guarantee.
| Topic | Indicative share of the paper | Revision priority |
|---|---|---|
| 1.1 Processors, I/O and storage | Around a fifth to a quarter | High |
| 1.4 Data types, structures and algorithms | Around a fifth to a quarter (includes calculations) | High |
| 1.3 Exchanging data (compression, networks, databases) | Around a fifth | High |
| 1.2 Software and software development | Around a fifth | Medium |
| 1.5 Legal, moral and ethical issues | Smaller share, but levels-marked | Medium (high impact per mark) |
Key Point: Section 1.5 carries fewer raw marks but a disproportionate share of the levels-marked marks. A confident, balanced ethics answer is one of the most reliable ways to bank an extra two or three marks, because the content is less technically fragile than a floating-point calculation.
These should be muscle memory by exam day — fast and accurate, with working shown.
Convert 10110101 to denary.
| Place value | 128 | 64 | 32 | 16 | 8 | 4 | 2 | 1 |
|---|---|---|---|---|---|---|---|---|
| Bit | 1 | 0 | 1 | 1 | 0 | 1 | 0 | 1 |
128 + 32 + 16 + 4 + 1 = 181.
Represent -45 in 8-bit two's complement: write +45 = 00101101, invert to 11010010, add 1 to get 11010011. Check: -128 + 64 + 16 + 2 + 1 = -45.
With a 6-bit mantissa 011010 and 4-bit exponent 0011 (both two's complement): the mantissa is 0.110102=0.8125; the exponent is +3, so shift the point three places right to give 0110.102=6.5.
The KaTeX rendering of the normalised form is:
value=mantissa×2exponent
A 1920 x 1080 image at 24 bits/pixel:
1920×1080×24=49,766,400 bits=6,220,800 bytes≈5.93 MB
(dividing by 8, then by 1024 twice). State your kilobyte convention if the question is ambiguous.
For mask 255.255.255.192, the final octet 11000000 leaves 6 host bits, so 26−2=62 usable hosts per subnet (subtracting the network and broadcast addresses).
Each command word is an instruction about what to do, and matching it is the cheapest mark on the paper.
| Command word | What it asks | Typical tariff | Discriminator |
|---|---|---|---|
| State / Give / Identify | A brief factual point | 1-2 | No explanation expected; one point per mark |
| Define | The precise meaning of a term | 1-2 | Technical accuracy, not an everyday paraphrase |
| Describe | The features or what happens | 2-4 | Sequence/detail, but no causation required |
| Explain | What happens and why | 3-6 | Causal connectives: because, therefore, this means |
| Compare | Similarities and/or differences | 3-6 | Both items paired with comparative connectives |
| Discuss | Different aspects toward a conclusion | 8-12 | Multiple viewpoints + evidence + conclusion |
| Evaluate | A justified judgement of worth | 8-12 | Weighed strengths/weaknesses + a reasoned verdict |
| Calculate | A numerical result | 2-6 | Show working; give the unit |
Exam Tip: Underline the command word and circle the mark tariff before you write. The pair tells you the shape of the answer: a 6-mark explain needs roughly three developed because-statements, not six disconnected facts.
A school is replacing the mechanical hard disk drives (HDDs) in its student laptops with solid-state drives (SSDs).
Evaluate the use of solid-state storage rather than magnetic hard disk storage in this context. [12]
| AO | Demand in this question | Approx. emphasis |
|---|---|---|
| AO1 | Accurate knowledge of how SSDs and HDDs store data and their characteristics | Knowledge floor |
| AO2 | Applying those characteristics to student laptops specifically (portability, drops, cost per machine) | Application |
| AO3 | Weighing the trade-offs and reaching a justified verdict for this context | Where Level 3 is decided |
This is a levels-marked question. Markers place your response in a band by its quality of reasoning and balance, then fine-tune within the band — they do not tally individual ticks.
SSDs are faster than HDDs because they have no moving parts and use flash memory, so the laptops will boot and load programs more quickly. SSDs are also more durable because there is no read/write head to be damaged if a student drops the laptop, which happens often with school equipment. However, SSDs cost more per gigabyte than HDDs, so the school may get less storage for the same money. Overall SSDs are a good choice because speed and durability matter for laptops that are carried around.
This is competent and applied — it ties durability to students dropping laptops. But the evaluation is thin: the conclusion asserts "a good choice" without genuinely weighing cost against benefit, and there is no technical depth on how the storage works or on secondary factors (power, lifespan, capacity needs).
An HDD stores data magnetically on spinning platters read by a moving head, whereas an SSD stores data as charge in NAND flash cells with no mechanical parts. For student laptops this matters in two applied ways. First, SSDs have far lower access times because there is no seek/rotational latency, so boot and application-load times fall noticeably — useful when a class of students all log in at once. Second, with no moving head an SSD tolerates being knocked or dropped, which is common for portable school devices, reducing breakage and data loss. Against this, SSDs cost more per gigabyte, so for the same budget the school buys less capacity; and flash cells have a finite number of write cycles. For typical student work (documents, light coding) capacity demands are modest and write volumes are low, so the lifespan concern is minor. On balance the durability and speed benefits outweigh the higher cost per gigabyte in this portable, multi-user context.
Now both technologies are explained at the mechanism level, the application is specific, and the trade-off is actually weighed (cost and write-lifespan are raised and then judged against the use-case). The conclusion follows from the argument rather than being tacked on.
Magnetic HDDs record data as polarised regions on rotating platters accessed by an actuated read/write head, so access time is dominated by seek and rotational latency; SSDs store data as trapped charge in NAND flash and are addressed electronically, giving access times an order of magnitude lower. In the school-laptop context this produces three applied benefits. (1) Performance: simultaneous logins and application loads complete faster because there is no mechanical contention, improving lesson time-on-task. (2) Physical robustness: with no moving head, an SSD survives the drops and knocks routine for carried devices, lowering both hardware failure and the data-loss safeguarding risk that accompanies a failed student drive. (3) Power and heat: lower idle/active power slightly extends battery life and reduces fan noise in a classroom. The countervailing factors are cost per gigabyte, which is higher, and finite program/erase endurance of flash. Both, however, must be weighed against this workload: student documents and coursework are small, write volumes are low, and modern wear-levelling spreads writes so that endurance is unlikely to be reached within a device's service life; meanwhile capacities sufficient for student use are now affordable. The remaining genuine cost premium buys measurable lesson-time, safeguarding and reliability gains. I therefore judge SSDs the better choice here, with the caveat that if the school instead needed bulk media or archival storage on a fixed budget, the HDD cost-per-gigabyte advantage could reverse the decision — which underlines that the verdict is context-dependent rather than absolute.
The top-band answer reaches Level 3 because every claim is technically grounded and applied to the specific scenario, the analysis is genuinely two-sided (it raises the strongest counter-arguments, cost and endurance, and then resolves them with reasoning rather than dismissing them), and the conclusion is justified and qualified — it states the condition under which the decision would flip. The Mid-band answer loses marks not on accuracy but on evaluation: it lists pros and a con but never weighs them, so it cannot demonstrate the AO3 judgement the command word evaluate demands. The single most valuable upgrade between bands is replacing "overall it is good" with "I judge X because the weighed benefits Y exceed the costs Z in this context."
A streaming service stores a 3-minute audio clip sampled at 44{,}100 Hz, using a sample resolution of 16 bits, recorded in stereo (2 channels).
(a) Calculate the uncompressed file size in megabytes. Show your working. [4]
(b) Explain one reason the service would apply lossy compression before streaming this clip to mobile users. [2]
| Part | AO | Demand |
|---|---|---|
| (a) | AO2 | Apply the sound-file-size method to these sampling parameters with correct units |
| (b) | AO1/AO2 | Knowledge of lossy compression applied to the mobile-streaming context |
State the rule, then substitute:
size=sample rate×resolution×channels×time
=44,100×16×2×180=254,016,000 bits
Convert to bytes and then megabytes:
8254,016,000=31,752,000 bytes≈1024×102431,752,000≈30.3 MB
Mark allocation: the rule (1), correct substitution including the 180-second conversion (1), conversion bits→bytes (1), final answer in MB with unit (1). A candidate who writes only "30.3 MB" with no working forfeits up to three method marks if the arithmetic slips.
Mid-band (1 mark): "Lossy compression makes the file smaller so it downloads faster." — correct but generic; it does not engage mobile users or explain how the saving arises.
Top-band (2 marks): "Lossy compression discards audio data the human ear is least likely to perceive, substantially reducing the bit rate; for mobile users on limited or metered bandwidth this means the clip streams without buffering and uses less of their data allowance, at a quality loss most listeners will not notice on a phone." — this earns both marks because it states the mechanism (perceptual data removal) and applies it to the named context (mobile bandwidth and data caps).
Part (a) is a pure AO2 mark-for-method question: the discriminator is visible working with units, not memorised answers. The most common error is forgetting to convert minutes to seconds (using 3 instead of 180) or stopping at bits/bytes without reaching megabytes. Part (b) shows how even a 2-mark explain rewards a mechanism-plus-context sentence over a generic one; "smaller file" alone tops out at 1 mark because it neither names the perceptual basis of lossy compression nor ties it to the mobile scenario. The transferable lesson: on applied questions, name the context inside the answer. This sound-representation method is developed fully in the Data Representation course; the compression theory sits in Networks.
Paper 1 spans five sections, and its highest-tariff questions deliberately cross between them. Treating each section as a sealed silo is the most common revision mistake at this level. The table below shows the kinds of crossing the paper exploits, so you can revise the connections, not just the topics.
| Crossing | Example synoptic question shape | Where it is taught |
|---|---|---|
| 1.1 ↔ 1.4 | "Explain how the choice of data structure affects processor cache performance" | Processors and Hardware ↔ Data Structures |
| 1.3 ↔ 1.5 | "Evaluate the privacy implications of storing user data in a cloud database" | Networks / Databases and Ethics |
| 1.2 ↔ 1.4 | "Explain why an interpreted language may be preferred during the development of a data-processing algorithm" | Software and Systems ↔ Data Representation |
| 1.3 ↔ 1.4 | "Calculate the transmission time for a compressed image over a given link" | Data Representation ↔ Networks |
| 1.1 ↔ 1.2 | "Explain how the operating system schedules processes on a multi-core processor" | Software and Systems ↔ Processors and Hardware |
Key Point: When you revise a topic, finish by asking "which other section does this touch?" A floating-point calculation (1.4) becomes a networking bandwidth question (1.3); an encryption method (1.3) becomes an ethics evaluation (1.5). The marks at the top of the paper live on these bridges, and the sibling courses listed above are where each bridge is built.
Automating arithmetic frees thinking time for the levels-marked questions. Use these working targets when you practise data-representation questions under the clock:
| Calculation type | Target time | Most common slip |
|---|---|---|
| Binary ↔ denary (8-bit) | Under 45 seconds | Mis-aligning place values |
| Two's complement of a negative | Under 60 seconds | Forgetting the final "+1" |
| Floating-point to denary | Under 90 seconds | Wrong direction of point shift |
| Image/sound file size | Under 2 minutes | Skipping a unit conversion (bits→bytes→MB) |
| Subnet host count | Under 60 seconds | Forgetting the "−2" for network/broadcast |
These are exam-technique failures rather than gaps in content knowledge — and they are recoverable in the final weeks, often without learning a single new fact. The distinction matters when you analyse a marked paper: if you lost a mark because you genuinely did not know the content, the fix is revision; but if you knew the content and still lost the mark, the fix is technique, and technique errors are far cheaper to eliminate. When you review a practice paper, label every lost mark as either "didn't know" or "knew but mis-answered" — most students are startled by how many fall into the second category, and those are the marks this lesson is designed to recover.
| Pitfall | Why it costs marks | The fix |
|---|---|---|
| Answering describe when asked to explain | The "why" marks are simply never attempted | Underline the command word; for explain, force a because/therefore into every point |
| Final answer only on a calculation | Method marks cannot be awarded without visible working | Write the rule, the substitution and intermediate values every time |
| Repeating one point in different words | A 2-mark state question with one idea scores 1 | Check each bullet is a genuinely different fact |
| One-sided extended responses | Without a second viewpoint the answer is capped below Level 3 | Plan two opposing points before writing the first sentence |
| Generic textbook answer to an applied question | AO2 marks require the scenario, not the textbook | Name the context (e.g. "student laptops") inside the answer |
| Ignoring the mark tariff | Over-writing a 1-mark question wastes time; under-writing a 6-mark one caps the score | Match developed points to marks: roughly one developed point per 1-2 marks |
| Vague comparatives ("it's better/faster") | Unquantified, unexplained claims earn nothing | Say what is better, why, and where possible by how much |