📚 Common Misconceptions in IB & WJEC Computer Science | IB和WJEC计算机科学常见误区
In both IB and WJEC Computer Science courses, students often encounter conceptual pitfalls that can lead to lost marks in exams and flawed problem-solving. This article clarifies the most common misconceptions, from algorithm analysis to system design, helping you build a rock-solid understanding.
在IB和WJEC计算机科学课程中，学生常常会遇到一些概念误区，这些误区可能导致考试失分和解题错误。本文澄清了从算法分析到系统设计的最常见误解，帮助你建立扎实的理解。
1. Big O Notation and Efficiency | 大O表示法与效率混淆
One of the most prevalent misunderstandings in algorithm analysis involves the interpretation of Big O notation. Students frequently struggle with constant factors, logarithmic bases and the meaning of ‘worst case’.
算法分析中最普遍的误解之一涉及对大O表示法的解释。学生们经常在常数因子、对数底数和“最坏情况”的含义上挣扎。
A common mistake is to treat O(2n) as distinct from O(n). In Big O, constant multipliers are dropped because we care about the growth rate, not exact run times. Thus, O(2n), O(5n) and O(n/2) are all equivalent to O(n).
一个常见错误是把O(2n)与O(n)区别对待。在大O中，常数乘数被舍弃，因为我们关心的是增长率，而不是精确的运行时间。因此，O(2n)、O(5n)和O(n/2)都等价于O(n)。
Another error is believing that O(log n) requires specifying base 2 or base 10. In Big O, the base of a logarithm is irrelevant because logₐ n differs from log_b n only by a constant factor, which is ignored. So O(log₂ n) and O(log₁₀ n) are the same complexity class.
另一个错误是认为O(log n)必须指定是以2为底还是以10为底。在大O中，对数的底数无关紧要，因为logₐ n与log_b n只相差一个常数因子，而常数因子被忽略。所以O(log₂ n)和O(log₁₀ n)属于相同的复杂度类别。
Many learners also confuse ‘worst case’ with ‘always’. An algorithm that is O(n²) in the worst case might run in O(n) for most inputs. Big O upper bounds do not describe typical behaviour.
许多学习者还将“最坏情况”与“总是”混淆。一个在最坏情况下为O(n²)的算法可能在大多数输入下以O(n)运行。大O上界并不描述典型行为。
O(1) < O(log n) < O(n) < O(n log n) < O(n²) < O(2ⁿ) < O(n!)
2. Recursion vs Iteration | 递归与迭代
Recursion is often seen as inherently slower or more memory-intensive than iteration, but that is not always true. Tail-recursive functions can be optimised by compilers to run as efficiently as loops, while poorly designed recursion without a proper base case can cause stack overflow.
递归通常被认为天生比迭代更慢或更占内存，但这并非总是如此。尾递归函数可以被编译器优化为与循环一样高效，而缺乏正确基本情况的低质量递归则可能导致堆栈溢出。
A widespread misconception is that every recursive solution must have an explicit return statement for the base case. In some languages, a missing base case leads to infinite recursion and crash, but the base case might be expressed via a condition that does not explicitly use ‘return’ if the language allows implicit returns.
一个普遍的误解是每个递归解都必须为基本情况包含明确的返回语句。在某些语言中，缺少基本情况会导致无限递归和崩溃，但如果语言允许隐式返回，基本情况可能通过条件表达而不显式使用“return”。
Another error is confusing recursion depth with time complexity. For example, a binary search algorithm has O(log n) time complexity but its recursion depth is also log n. However, a Fibonacci recursive version has O(2ⁿ) time despite a depth of only n, because of repeated subproblems.
另一个错误是混淆递归深度与时间复杂度。例如，二分查找算法的时间复杂度是O(log n)，其递归深度也是log n。然而，斐波那契递归版本尽管深度仅为n，却具有O(2ⁿ)的时间，因为存在重复子问题。
Students also incorrectly assume that all iterative solutions are more efficient than recursive ones. In functional programming, recursion is natural and often compiled to efficient iterative bytecode, while in imperative exams, recursion may be penalised for clarity but not for asymptotic performance.
学生也错误地认为所有迭代解都比递归解更高效。在函数式编程中，递归是自然的，且常被编译为高效的迭代字节码，而在命令式考试中，递归可能因可读性而被扣分，但在渐进性能上却未必更差。
3. Data Types and Type Casting | 数据类型与类型转换
A very common mistake is to assume that integer division automatically yields a floating-point result. In many languages like Java or C, dividing two integers performs integer division, discarding the remainder, which can lead to off-by-one or zero-like errors.
一个非常常见的错误是假设整数除法会自动产生浮点结果。在像Java或C这样的许多语言中，两个整数相除执行的是整数除法，会丢弃余数，这可能导致差一错误或近似为零的错误。
Type casting is another area of confusion. Implicit casting (widening) from int to float is safe, but explicit narrowing from double to int truncates rather than rounds. Students often forget that casting a floating-point 3.99 to int gives 3, not 4.
类型转换是另一个容易混淆的领域。从int到float的隐式转换（拓宽）是安全的，但从double到int的显式窄化转换是截断而不是四舍五入。学生常常忘记将浮点3.99转换为int会得到3，而不是4。
In WJEC pseudocode and IB programming, character encoding like ASCII vs Unicode is sometimes misunderstood. ‘A’ + 1 may equal ‘B’ in ASCII, but Unicode code points can be larger, and operations on chars depend on the language specification.
在WJEC伪代码和IB编程中，像ASCII与Unicode的字符编码有时被误解。’A’ + 1在ASCII中可能等于’B’，但Unicode码点可能更大，而对字符的操作取决于语言规范。
The distinction between primitive types and object references is crucial for OOP tasks. Assigning one array variable to another copies the reference, not the elements. Modifying one affects the other, which surprises many learners.
基本类型与对象引用之间的区别对于OOP任务至关重要。将一个数组变量赋值给另一个只会复制引用，而不是元素。修改其中一个会影响另一个，这令许多学习者感到意外。
4. Object-Oriented Programming Pitfalls | 面向对象编程的陷阱
Many students believe that ‘extends’ always establishes an ‘is-a’ relationship. In reality, inheritance should model a genuine subtype relationship. Misusing inheritance for code reuse can violate the Liskov substitution principle, causing unexpected behaviour.
许多学生认为’extends’总是建立起’is-a’关系。实际上，继承应当模拟真正的子类型关系。为代码重用而滥用继承可能违反里氏替换原则，导致意想不到的行为。
The difference between overriding and overloading is frequently mixed up. Overriding provides a method with the same signature in a subclass, enabling polymorphism. Overloading uses the same method name but different parameters within the same class and is resolved at compile time.
重写和重载之间的区别经常被混淆。重写是在子类中提供具有相同签名的方法，从而实现多态性。重载则在同一个类内使用相同方法名但不同的参数，并在编译时确定调用。
Static methods cannot be overridden, only hidden. In both IB and WJEC syllabi, a static method called on a reference variable calls the method of the declared type, not the actual object type, which contradicts polymorphic expectations.
静态方法不能被重写，只能被隐藏。在IB和WJEC大纲中，通过引用变量调用的静态方法会调用声明类型的方法，而非实际对象类型的方法，这与多态期望相矛盾。
Encapsulation is not just about declaring fields private. True encapsulation means restricting direct access and providing controlled interfaces. Simply adding getters and setters for every field does not guarantee proper encapsulation.
封装并不仅仅是将字段声明为私有。真正的封装意味着限制直接访问并提供受控接口。仅仅为每个字段添加getter和setter并不能保证正确的封装。
5. Stack and Queue Operations | 栈与队列的操作误区
Students often confuse LIFO (Last In, First Out) with FIFO (First In, First Out). A stack supports push and pop at the same end, whereas a queue enqueues at the rear and dequeues from the front. Swapping these operations leads to invalid algorithms.
学生经常混淆LIFO（后进先出）与FIFO（先进先出）。栈在同一端支持压入和弹出，而队列在尾部入队、在头部出队。交换这些操作会导致算法无效。
In exam trace tables, a common error is to pop from an empty stack without checking. This can occur in reverse Polish notation evaluation or parenthesis matching. Always ensure the data structure is non-empty before popping.
在考试跟踪表中，一个常见错误是在没有检查的情况下从空栈弹出。这在逆波兰表示法求值或括号匹配中可能发生。务必在弹出前确保数据结构非空。
The concept of a priority queue is mistaken for a standard FIFO queue. A priority queue removes elements based on priority, not insertion order. It is often implemented using a heap, which reintroduces O(log n) behaviour.
优先级队列的概念常被误认为是标准的FIFO队列。优先级队列根据优先级而不是插入顺序移除元素。它通常使用堆实现，这会重新引入O(log n)的行为。
In linked-list implementations of stacks and queues, confusion about head and tail pointers can lead to O(n) operations instead of O(1). A properly designed linked queue should maintain both head and tail references for O(1) enqueue and dequeue.
在用链表实现栈和队列时，对头尾指针的混淆可能导致O(n)操作而非O(1)。设计良好的链式队列应同时维护头部和尾部引用，以实现O(1)的入队和出队。
6. Networking Protocols and Layers | 网络协议与层次模型
It is a misconception that TCP guarantees delivery or bandwidth. TCP provides reliable delivery through acknowledgements and retransmissions, but it does not guarantee a specific transmission speed. Network congestion can cause variable throughput.
认为TCP保证传输或带宽是一个误区。TCP通过确认和重传提供可靠传输，但它并不保证特定的传输速度。网络拥塞可能导致可变的吞吐量。
Another error is to think IP addresses are fixed per device. While static IPs exist, most devices use DHCP to obtain dynamic IPs. Moreover, with IPv4 exhaustion, Network Address Translation (NAT) allows multiple devices on a private network to share one public IP.
另一个错误是认为IP地址是每个设备的固定标识。虽然静态IP存在，但大多数设备使用DHCP获取动态IP。此外，随着IPv4地址耗尽，网络地址转换（NAT）允许私有网络上的多个设备共享一个公共IP。
The OSI model layers are frequently memorised but poorly understood. For instance, a common misconception places HTTP at the transport layer because it ‘transports web pages’. HTTP actually belongs to the application layer, relying on TCP at layer 4.
OSI模型层次常被死记硬背，但理解不足。例如，一个常见误解是将HTTP放在传输层，因为它“传输网页”。HTTP实际上属于应用层，依赖于第4层的TCP。
In both IB and WJEC papers, students mislabel diagrams showing packet switching. Each packet may take a different route and be reassembled at the destination. Circuit switching, on the other hand, reserves a dedicated path, which is less efficient for bursty data.
在IB和WJEC试卷中，学生在示意分组交换的图上常标错。每个分组可能走不同的路径，并在目的地重新组装。而电路交换则会预留一条专用路径，对于突发性数据传输效率较低。
7. Database Normalization | 数据库范式化
A widespread misunderstanding is that normalisation always improves performance. While it reduces data redundancy and update anomalies, highly normalised databases can require expensive joins that degrade query speed. Exam questions highlight the trade-off between normalisation and denormalisation for specific use cases.
一个普遍的误解是范式化总能提高性能。虽然它减少了数据冗余和更新异常，但高度范式化的数据库可能需要代价高昂的连接，从而降低查询速度。考题会强调针对特定用例在范式化与非范式化之间的权衡。
Students often confuse 2NF and 3NF. Second normal form eliminates partial dependencies on a composite primary key, while third normal form removes transitive dependencies on non-key attributes. A table in 2NF can still suffer from transitive anomalies.
学生经常混淆2NF和3NF。第二范式消除了复合主键上的部分依赖，而第三范式消除了非键属性上的传递依赖。一个满足2NF的表仍然可能遭受传递异常。
A further mistake is considering every table with a primary key as being in 1NF. 1NF requires that each column contains atomic values and there are no repeating groups. A column storing comma-separated strings violates 1NF.
另一个错误是认为任何有主键的表都满足1NF。1NF要求每列包含原子值且没有重复组。存储以逗号分隔的字符串的列就违反了1NF。
In relational design, surrogate keys are sometimes misused. A surrogate key is an artificial identifier, but natural keys formed from real data can also be used. Choosing the wrong primary key can lead to unnecessary join complexity.
在关系设计中，代理键有时被误用。代理键是人为的标识符，而由真实数据形成的自然键也可以使用。选择错误的主键可能导致不必要的连接复杂性。
8. Process vs Thread | 进程与线程
A classic misconception is that a process is the same as a program. A program is a passive collection of instructions stored on disk, while a process is an active instance of that program in execution, with its own memory space and resources.
一个典型的误解是认为进程等同于程序。程序是存储在磁盘上的被动指令集合，而进程是该程序在运行中的活动实例，拥有自己的内存空间和资源。
Another error is assuming that multi-threading always speeds up execution. On a single-core processor, threads are interleaved but cannot run truly in parallel. Moreover, thread creation and context switching incur overhead, which might even slow down the task.
另一个错误是认为多线程总能加速执行。在单核处理器上，线程是交错执行的，但无法真正并行运行。此外，线程创建和上下文切换会带来开销，甚至可能使任务变慢。
In shared memory environments, race conditions occur when multiple threads access shared data concurrently without synchronisation. Students often overlook the need for mutexes or semaphores, leading to data corruption, deadlocks or livelocks.
在共享内存环境中，当多个线程同时访问共享数据而没有同步时就会发生竞态条件。学生经常忽略对互斥锁或信号量的需要，从而导致数据损坏、死锁或活锁。
When comparing processes and threads, a common IB question points out that processes are heavyweight and isolated, whereas threads share the same address space within a process. This makes threads lighter to create but riskier regarding memory corruption.
在比较进程和线程时，一个常见的IB考题指出进程是重量级的且相互隔离，而线程在同一个进程内共享地址空间。这使得线程创建更轻量但内存损坏风险更大。
9. Boolean Logic and Logic Gates | 布尔逻辑与逻辑门
Students frequently misapply De Morgan’s laws, believing NOT (A AND B) equals NOT A AND NOT B. The correct transformation is NOT (A AND B) = NOT A OR NOT B. This error persists until they internalise the rule with truth tables.
学生经常误用德摩根定律，认为NOT (A AND B)等于NOT A AND NOT B。正确的转换是NOT (A AND B) = NOT A OR NOT B。这个错误会一直存在，直到他们通过真值表内化规则。
A common misconception in logic gate circuits is that the XOR gate can be built by simply OR-ing and AND-ing inputs. XOR outputs true when inputs are different, and its Boolean expression is A ⊕ B = (A ∧ ¬B) ∨ (¬A ∧ B). Using just OR and AND does not achieve this.
在逻辑门电路中的常见误解是，异或门可以通过简单的或操作和与操作来构建。当输入不同时XOR输出真，其布尔表达式为A ⊕ B = (A ∧ ¬B) ∨ (¬A ∧ B)。仅仅使用OR和AND是无法实现的。
In both WJEC and IB, truth tables require careful ordering. A common slip is to write a minterm for a row where the function is 0. Minterms are formed only for rows with output 1, and then OR-ed together to get the sum-of-products.
在WJEC和IB中，真值表要求仔细排序。一个常见的疏漏是将函数为0的行写成最小项。最小项仅针对输出为1的行构造，然后通过或运算组合得到积之和表达式。
Simplifying Boolean expressions using identities like A + AB = A (absorption) is often overlooked. Students try to expand everything rather than spotting patterns, leading to unnecessarily complex circuits.
使用恒等式如A + AB = A（吸收律）简化布尔表达式的做法常常被忽视。学生倾向于展开一切而不是识别模式，导致不必要的复杂电路。
10. Memory Hierarchy and Cache | 内存层次结构与缓存
A key misconception is that a bigger cache always leads to better performance. While a larger cache can store more data, it may have higher latency and hit time. Cache design involves trade-offs in size, associativity, and block size.
一个关键的误区是更大的缓存总能带来更好的性能。虽然更大的缓存能存储更多数据，但它可能具有更高的延迟和命中时间。缓存设计涉及大小、关联度和块大小之间的权衡。
The principle of locality is often misunderstood. Temporal locality means recently accessed data is likely to be accessed again soon; spatial locality means nearby addresses will be accessed. Basing cache design on one without the other leads to inefficiency.
局部性原理经常被误解。时间局部性意味着最近访问的数据可能会再次被访问；空间局部性意味着附近的地址将被访问。如果缓存设计只考虑其中一种而忽略另一种，就会导致效率低下。
When drawing memory hierarchy diagrams in exams, students sometimes place cache below RAM. The correct order from top (fastest, smallest) to bottom (slowest, largest) is: registers, cache (L1, L2, L3), main memory (RAM), and then secondary storage (disk).
在考试中绘制内存层次结构图时，学生有时会把缓存放在RAM下方。从最顶层（最快、最小）到最底层（最慢、最大）的正确顺序是：寄存器、缓存（L1、L2、L3）、主存（RAM），然后是辅存（磁盘）。
Virtual memory is not an extension of cache. Cache is hardware-managed, transparent and faster. Virtual memory uses disk space to extend RAM, managed by the OS, and swapping occurs at page level with a significant performance hit if overused.
虚拟内存不是缓存的扩展。缓存由硬件管理，透明且更快。虚拟内存使用磁盘空间扩展RAM，由操作系统管理，并按页面级别进行交换，如果过度使用会对性能造成显著影响。
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