How to Answer "Explain" Questions in Cambridge International AS & A Level Computer Science

"Explain" questions ask for more than a correct fact. You need to show why something happens, how a process works, what purpose a feature serves, or how one idea leads to another. In Computer Science, that often means connecting technical knowledge into reasoning rather than listing isolated definitions. An answer can therefore contain accurate terminology and still feel incomplete. A useful approach is to identify the core concept first, then make the "how" or "why" explicit so the relationship between your points is visible.

Start by identifying exactly what needs explaining

Before writing, isolate the part of the question that actually needs an explanation. It may be asking why a hardware feature improves performance, how a security method protects data, how a process is carried out, or why one approach has an advantage over another. The surrounding topic tells you what knowledge is relevant, but the wording of the question tells you what relationship you need to develop. This matters because Computer Science topics can contain many related facts. If a question asks how an interrupt from an input device is detected and handled during the fetch-execute cycle, for example, simply defining an interrupt is not enough. The response needs to follow the mechanism: detection, the point at which the processor responds, what happens to the current state, and how execution continues. The answer has to explain the sequence rather than just name its parts.

Click to enlarge

Move from the fact to the reason

A common weak answer stops one step too early. It gives a relevant fact, but it never says why that fact produces the outcome in the question. If you are explaining why increasing cache memory can improve CPU performance, saying that cache is fast memory close to the CPU is useful knowledge, but it does not yet complete the explanation. The reasoning needs to continue. A larger cache can hold more frequently used instructions and data, which can reduce the number of times the CPU has to access slower main memory. That reduces waiting for data and can improve execution speed. The important difference is not simply that the answer is longer; it makes the cause-and-effect relationship visible.

Make the links between stages explicit

Many Explain questions are really testing whether you understand how several parts of a system connect. This is especially clear in questions about processes such as encryption, databases, networking, processor operation or artificial intelligence. Listing the stages may show that you remember them, but an explanation should make clear what each stage does and why the next stage follows. A 2025 Paper 1 question shown in NeuraGeek asks how AI is used in a program that photographs words on an item, translates them and outputs the result as audio. The mark scheme accepts several possible stages, including image recognition or optical character recognition to locate characters in the image, natural-language processing to form or interpret words and match them to a chosen language, and text-to-speech to generate the audio output. What makes this a useful Explain example is that the answer is not one definition of AI. The response has to connect the technology to the job the program is performing.

Click to enlarge

Use precise Computer Science terminology

Technical vocabulary matters because it allows you to express the mechanism accurately. Terms such as cache, RAM, public key, private key, ciphertext, OCR, natural-language processing, foreign key and referential integrity have specific meanings. Replacing them with vague phrases can weaken a response even when the general idea is sensible. Precision does not mean stuffing the answer with jargon. Use the terms that help explain what is happening. In a relational-database question, for example, saying that data is "organised better" is vague. Explaining that related data can be stored in separate tables and linked through relationships gives a clearer reason why duplication can be reduced. The terminology should support the explanation, not substitute for it.

Develop every requested point

Some Explain questions ask for two reasons, two benefits or several stages. Each requested point needs development rather than a label. If you are asked for two benefits of a relational database over a file-based approach, for example, do not stop at "reduced redundancy" and "improved integrity". Explain how linked tables can reduce repeated data, then separately explain how database rules and relationships can help keep references valid and data consistent.

Explain is different from describe

A description tells the reader what something is like or what happens. An explanation gives the reason, mechanism or purpose behind it. The two can overlap, but stopping at description is a common way to leave an Explain answer underdeveloped. Imagine a question asking why a modern television can be considered an embedded system. Describing the television as containing a processor, memory and software gives characteristics. Explaining the classification requires connecting those characteristics to the concept: the computer system is built into a larger device and performs a defined set of functions rather than operating as a general-purpose computer. That relationship is what turns the response into an explanation.

Match the depth to the question

The amount of development you need depends on the question and its mark allocation. A short Explain question may require one clearly developed relationship, while a four-mark question may give credit for several connected technical points. Treat the marks as a guide to the scale of the response, not as a rigid sentence-count formula. Check whether each important point answers "how?" or "why?". If the connections are still left for the examiner to infer, add the missing reasoning rather than more unrelated facts.

Use a simple explanation chain when you get stuck

When you are unsure how to build the answer, ask yourself three questions: What is the relevant concept? How or why does it operate in this situation? What consequence or purpose completes the explanation? That produces a simple chain of concept, mechanism and outcome. For asymmetric encryption, the concept is the public/private key pair. The sender uses the recipient's public key to encrypt the document, while the corresponding private key is kept by the recipient. The consequence is that an intercepted ciphertext cannot simply be read without the private key, which is why the process supports confidentiality. The exact structure changes from question to question, but the reasoning remains visible.

Check the explanation before moving on

Before leaving an Explain question, read the response once and check whether it goes beyond the initial fact. Have you actually answered how or why? Are the stages connected? Is the terminology precise? If the question asked for more than one explanation, did you develop each one rather than just list it? Inside NeuraGeek, you can attempt the past-paper question first, compare your answer with the mark scheme and then use Paper Tutor when a marking point is still unclear. The useful part is not getting a model answer before you try. It is seeing which stage or relationship your own response failed to make explicit, then using that feedback on the next question.

Click to enlarge

The goal with Explain questions is not to make every answer longer. Make the reasoning visible: state the relevant idea, show how or why it matters, connect it to the consequence, then move on.