📚 IB OCR Computer Science: MCQ Hack Techniques | IB OCR 计算机：选择题秒杀技巧
Mastering multiple-choice questions in IB or OCR Computer Science requires more than just memorising facts — it demands strategic thinking, sharp pattern recognition and time-efficient calculation methods. This article distils a set of battle-tested techniques that will help you crack tricky questions in topics ranging from binary arithmetic and Boolean logic to algorithm tracing and networking. By internalising these hacks, you can boost accuracy and speed under exam pressure.
在 IB 或 OCR 计算机科学考试中攻克选择题，不仅需要背诵知识点，更需要策略性思维、敏锐的模式识别和高效的计算方法。本文提炼了一系列久经考验的技巧，帮助你攻克从二进制运算、布尔逻辑到算法追踪和网络等领域的难题。掌握这些技巧，你将在考试压力下提升正确率和速度。
1. Understanding the Question Stem and Avoiding Traps | 理解题干，避开陷阱
Read the stem twice and underline absolute words such as ‘always’, ‘never’, ‘only’ or ‘must’. These signal conditions that are rarely true in computer science — for instance, ‘all recursive algorithms are faster than iterative ones’.
仔细读题两遍，在 ‘always’、’never’、’only’ 或 ‘must’ 等绝对化词语下划线。这些词往往表示在计算机科学中极少成立的条件，例如“所有递归算法都比迭代算法快”。
Watch out for double negatives and words like ‘except’ or ‘not’. Rephrase the stem into a positive statement: “Which of the following does NOT improve cache hit rate?” becomes “Which factor leaves cache hit rate unchanged or makes it worse?”
注意双重否定以及 ‘except’、’not’ 等词。将题干改写为肯定句：“下列哪项不能提高缓存命中率？”可转化为“哪个因素会使缓存命中率不变或变差？”
In IB/OCR questions, distractors often use partially correct phrases that sound plausible but violate a small but crucial detail. Mentally add the missing condition before evaluating choices.
在 IB/OCR 题目中，干扰项常使用听起来合理但违反某个微小却关键细节的表述。评估选项前，在心里补上缺失的条件。
2. Process of Elimination: Narrowing Down Options | 排除法：缩小选项
Strike out choices that contradict a fundamental definition. For example, if a question asks about the properties of a stack, immediately eliminate any option that mentions FIFO — stacks are LIFO.
划掉与基本定义矛盾的选项。例如，若题目问栈的性质，立即淘汰任何提及 FIFO 的选项——栈是 LIFO。
When two options are exact opposites, the correct answer is often one of them. Examine the context: “A full adder produces…” with choices “SUM only”, “CARRY only”, “both SUM and CARRY”, “neither”. “Sum only” and “carry only” are extreme; “both” is the fact-based answer.
当两个选项完全相反时，正确答案往往是其中之一。审视上下文：“全加器产生……”，选项为“只有 SUM”、“只有 CARRY”、“SUM 和 CARRY 都有”、“都没有”。前两者极端，基于事实“两者都有”是正确答案。
Use partial knowledge: if you remember that a particular protocol operates at the transport layer, eliminate all options that place it at the network or application layer — even if you forget some wording.
利用局部知识：若记得某个协议工作在传输层，便可淘汰所有将其归入网络层或应用层的选项，即使遗忘了部分措辞。
3. Substitution and Verification | 代入验证法
If a question presents an expression or a piece of pseudocode, substitute a small concrete value and trace. For instance, to evaluate ‘2ⁿ mod 4’ for n>1, test n=2 → 4 mod 4 = 0, n=3 → 8 mod 4 = 0; the pattern is always 0.
若题目给出表达式或伪代码片段，代入一个小数值并追踪。例如，要计算 n>1 时 ‘2ⁿ mod 4’，测试 n=2 → 4 mod 4 = 0，n=3 → 8 mod 4 = 0，模式总是 0。
For Boolean identities like A + A’B = A + B, test all four combinations of A and B mentally or with a quick truth table. If one combination fails, discard the identity.
对于像 A + A’B = A + B 的布尔恒等式，在脑海里或快速列真值表测试 A 和 B 的四种组合。只要有一种组合不成立，就抛弃该恒等式。
When comparing growth rates (Big O), plug in large numbers: n² vs n log n. For n=1000, n² = 1,000,000 while n log n ≈ 1000×10 = 10,000. This instantly confirms that n² grows faster.
比较增长率（大 O）时，代入大数字：n² vs n log n。取 n=1000，n² = 1,000,000，而 n log n ≈ 1000×10 = 10,000，立刻确认 n² 增长更快。
4. Using Counterexamples to Eliminate Answers | 利用反例排除
When a statement claims “X is always true”, construct the simplest counterexample. Claim: “In a binary tree, number of leaf nodes is always one more than internal nodes.” Counterexample: a tree with only one node (root) has 1 leaf and 0 internal — 1 is not 0+1. So false.
当陈述声称“X 总为真”时，构造最简单的反例。声称：“二叉树中，叶节点数总比内部节点数多 1。”反例：只有一个节点的树（根）有 1 个叶子和 0 个内部节点，1 ≠ 0+1。故错误。
For algorithm complexity: “All sorting algorithms that compare elements require Ω(n²) time.” Provide insertion sort on nearly sorted data — it runs in O(n) comparisons. The claim collapses.
对于算法复杂度：“所有基于比较的排序算法都需要 Ω(n²) 时间。”举出插入排序在近乎有序数据上只需 O(n) 次比较，该说法不攻自破。
In networking, a distractor may state “HTTPS uses port 80”. Recall actual port 443 for HTTPS and port 80 for HTTP. The single memory of 443 eliminates that choice.
网络题中，干扰项可能说“HTTPS 使用端口 80”。想到 HTTPS 实际使用端口 443，HTTP 使用 80，仅靠对 443 的记忆即可排除该项。
5. Drawing Diagrams and Visualising | 画图与可视化
For memory allocation or pointer-based linked-list questions, sketch boxes with addresses. A quick pencil drawing in the margin of the exam paper instantly clarifies pointer dereferencing and avoids off-by-one errors.
对于内存分配或基于指针的链表题，画出带有地址的方框。在试卷边缘快速用铅笔绘图，能立刻厘清解引用指针，避免差一错误。
In logic circuits, redraw the given gate combination in a standardised form (AND-OR or NAND-only). Labelling intermediate outputs with expressions like X = (A NAND B) helps you spot simplification opportunities.
在逻辑电路题中，将给定门组合重绘为标准形式（与-或或全与非）。用表达式标记中间输出，如 X = (A NAND B)，有助于发现简化机会。
For tree traversals (preorder, inorder, postorder), draw the tree and then literally trace your finger. Write the sequence step by step; do not rely on mental visualisation alone.
对于树的遍历（前序、中序、后序），画出树，然后用手指实际追踪。逐步写下序列；不要仅靠脑内空间想象。
6. Pseudocode Tracing and Algorithm Walkthroughs | 伪代码追踪与算法遍历
Use a variable table with columns for each variable and update rows as you mentally execute the pseudocode. This structured approach prevents losing track of loop counters or accumulation variables.
使用变量表，为每个变量列一栏，随着脑内执行伪代码逐行更新。这种结构化方法可避免遗忘循环计数器或累加变量的变化。
When encountering loops with an unknown number of iterations, identify the loop invariant. Example: In a while loop that repeatedly divides i by 2 until i ≤ 1, the invariant is that i reduces logarithmically, giving O(log n) complexity.
遇到不确定迭代次数的循环时，识别循环不变量。例如，在一个不断将 i 除以 2 的 while 循环中，直到 i ≤ 1，不变量是 i 呈对数减少，复杂度为 O(log n)。
For recursive calls, build a recursion tree. Label each call with its parameter values and return values bottom-up. This makes it easy to spot whether the recursion is tail-recursive or can cause stack overflow.
对于递归调用，构建递归树。自底向上标记每次调用的参数值和返回值，这便于识别递归是否尾递归，或可能导致栈溢出。
7. Binary & Hexadecimal Quick Calculations | 二进制与十六进制速算
Convert binary to hex in groups of four bits from the right. 10111010₂ → 1011 1010 → B A → 0xBA. No need to convert to decimal first.
二进制转十六进制时，从右开始每四位一组：10111010₂ → 1011 1010 → B A → 0xBA，无需先转十进制。
To quickly find two’s complement of a negative number, flip all bits after the first ‘1’ from the right. Example: -6 in 8-bit: positive 6 = 00000110₂. Trailing 1 at position 1, flip everything after: 11111010₂. Verify: 256-6 = 250, correctly 11111010.
快速求负数的二进制补码：从右往左，在第一个 ‘1’ 之后的所有位取反。例如 -6 用 8 位表示：正 6 = 00000110₂。右侧第一个 1 在位置 1，其后所有位取反得 11111010₂。验证：256-6=250，即 11111010。
For binary addition, break into nibbles: A = 0110 1101, B = 0011 1010. Add nibble by nibble: 1101+1010 = 1 0111 (carry 1 to next nibble). Then 0110+0011+carry=1010. Result: 1010 0111.
二进制加法可用半字节分组：A = 0110 1101，B = 0011 1010。逐个半字节相加：1101+1010 = 1 0111（向下一半字节进位 1）。然后 0110+0011+进位=1010。结果 1010 0111。
8. Boolean Algebra Simplification Tricks | 布尔代数化简技巧
Memorise the core identities and apply them in a sequence: absorption (A + AB = A), distribution, De Morgan’s laws. Spot common patterns such as (A+B)(A+C) = A + BC without expanding fully.
熟记核心恒等式并按序应用：吸收律 (A + AB = A)、分配律、德摩根定律。识别常见模式如 (A+B)(A+C) = A + BC，无需完全展开。
Use the consensus theorem: XY + X’Z + YZ = XY + X’Z. Scan the expression for a term that contains a literal and its complement in two other terms; that third term is redundant.
利用一致律定理：XY + X’Z + YZ = XY + X’Z。在表达式中扫描，若一个项包含的字面量分别出现在另外两项中且互补，则该第三项是冗余的。
When options offer different simplified forms, plug in a test vector (e.g., A=0, B=1, C=1) into both the original and the candidate. A mismatch eliminates that option instantly.
当选项提供不同简化形式时，代入测试向量（如 A=0, B=1, C=1）到原式和候选式，不匹配的立刻排除。
9. Logic Circuits & Truth Table Hacks | 逻辑电路与真值表秒杀
For a small number of inputs (≤4), construct a quick truth table in your head by toggling bits like a binary counter. Fill the output column by evaluating the expression gate by gate.
对于少量输入（≤4），在脑中像二进制计数器一样翻转位来快速构建真值表。逐门求值来填写输出列。
Identify standard gate patterns: an AND gate followed by a NOT is a NAND; XOR can be built from basic gates as A’B + AB’. Recognising these blocks lets you jump to the final Boolean function without redrawing the whole circuit.
识别标准门模式：与门后接非门就是与非门；异或门可由基本门构建为 A’B + AB’。识别这些模块后可直接写出布尔函数，无需重绘整个电路。
When a question asks “which gate combination yields a given truth table?”, compare the output column to known functions. If the output is 1 only when inputs differ, it is XOR; if all 1s except when all inputs are 1, it is NAND.
若题目问“哪种门组合产生给定真值表？”，将输出列与已知函数对比。若输入相异时输出 1，则是异或门；若除了全 1 外输出都是 1，则是与非门。
10. Data Structure Operation Analysis | 数据结构操作分析
For time complexity of operations on a BST, remember the worst case is a degenerate (skewed) tree. Search and insertion both become O(n). If the stem mentions “balanced BST”, it is O(log n).
对于二叉搜索树操作的时间复杂度，记住最坏情况是退化（倾斜）树，查找和插入均为 O(n)。若题干提到“平衡 BST”，则为 O(log n)。
In a hash table question about collisions, separate chaining keeps load factor manageable, but worst-case search is O(n) when all keys collide. Open addressing suffers from clustering; linear probing can degrade to O(n) even with moderate load.
在哈希表冲突题中，分离链接法可使负载因子可控，但最坏情况查找在所有键冲突时为 O(n)。开放地址法受聚集影响，线性探测即使负载适中也可能降级为 O(n)。
When comparing heap operations, note that insertion and deletion of min in a binary min-heap are both O(log n). Building a heap from an unsorted array can be done in O(n) using Floyd’s method — a common distractor claims O(n log n).
比较堆操作时，注意二叉最小堆的插入和删除最小元素均为 O(log n)。从未排序数组建堆可用 Floyd 方法在 O(n) 完成——常见干扰项会错误宣称 O(n log n)。
11. Network & Protocol Common Sense Judgment | 网络与协议常识判断
Layer functions must be clear: transport layer (TCP/UDP) handles ports and reliability; network layer (IP) handles logical addressing; data link layer (Ethernet) handles MAC addresses. A question mixing these layers offers easy elimination.
各层功能必须清晰：传输层（TCP/UDP）处理端口和可靠性；网络层（IP）处理逻辑地址；数据链路层（Ethernet）处理 MAC 地址。把各层混淆的题目可轻松排除。
For protocol purposes, remember: DHCP assigns IP addresses dynamically, DNS resolves names to IPs, ARP resolves IP to MAC. Any option that swaps these roles is wrong.
协议用途要牢记：DHCP 动态分配 IP 地址，DNS 将域名解析为 IP，ARP 将 IP 解析为 MAC。任何互换这些角色的选项都是错误的。
In security questions, symmetric encryption is fast but requires secure key exchange; asymmetric uses public/private key pairs. A scenario requiring key distribution over an insecure channel points to asymmetric encryption.
在安全题中，对称加密快但需要安全的密钥交换；非对称加密使用公/私钥对。需要在非安全信道分发密钥的场景应指向非对称加密。
12. Object-Oriented Concept Differentiation | 面向对象概念辨析
Encapsulation binds data and methods, hiding internal state. Inheritance allows a subclass to reuse superclass code. Polymorphism enables a single interface to represent different underlying forms (method overriding). When a stem describes “different classes responding to the same method call in their own way”, it is polymorphism, not inheritance.
封装将数据和方法绑定，隐藏内部状态。继承允许子类复用超类代码。多态使一个接口能表现不同的底层形态（方法重写）。当题干描述“不同类以自己的方式响应同一方法调用”时，这是多态，而非继承。
UML class diagram questions: an empty diamond represents aggregation (“has-a” with independent lifecycle); a filled diamond is composition (“owns-a” with dependent lifecycle). Check the wording “consists of” vs “uses a”.
UML 类图题：空菱形表示聚合（“has-a”，生命周期独立）；实心菱形表示组合（“owns-a”，生命周期依赖）。注意题干措辞是“consists of”还是“uses a”。
Abstract class vs interface: abstract classes can have implemented methods and state; interfaces (in Java) only declare method signatures. A question asking “can provide partial method implementation” points to an abstract class.
抽象类与接口：抽象类可有已实现的方法和状态；接口（Java 中）只能声明方法签名。若题目问“可以提供部分方法实现”，则应指向抽象类。
Published by TutorHao | Computer Science Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
屏轩国际教育cambridge primary/secondary checkpoint, cat4, ukiset,ukcat,igcse,alevel,PAT,STEP,MAT, ibdp,ap,ssat,sat,sat2课程辅导，国外大学本科硕士研究生博士课程论文辅导