📚 Oxbridge Mathematics Interview Questions Explained | 牛津剑桥数学专业面试真题解析
Oxford and Cambridge mathematics interviews are not about knowing the right answer immediately; they are designed to reveal how you think, reason, and respond to hints. Interviewers look for genuine curiosity, the ability to break a problem into manageable parts, and a willingness to explore ideas even when you are stuck. In this article we unpack a selection of classic interview-style problems, showing the step‑by‑step reasoning that can turn a daunting question into an exciting conversation.
牛津和剑桥的数学面试并非要求你立刻给出正确答案，而是旨在揭示你是如何思考、推理以及回应提示的。面试官看重的是真诚的好奇心、将问题拆解为可操作部分的能力，以及在卡壳时依然愿意探索的态度。本文拆解了一系列经典的面试风格问题，展示那种能将棘手的提问转化为精彩对话的逐步推理过程。
1. Prove that √2 is irrational | 证明根号2是无理数
This is perhaps the most famous “warm‑up” question. The interviewer observes whether you can construct a proof by contradiction and how clearly you handle the logic. They may begin by asking, “What does it mean for a number to be irrational?” Once you set up the standard proof, they might challenge steps like “Why must both p and q be even if p² is even?”
这可能是最著名的“热身”问题。面试官观察你能否构建反证法以及是否能清晰处理逻辑。他们可能会首先问：“一个数是无理数是什么意思？”当你建立起标准证明后，他们可能会对某个步骤提出质疑，例如“如果 p² 是偶数，为什么 p 和 q 一定都是偶数？”
Suppose √2 = p/q where p and q are coprime positive integers. Squaring gives 2 = p²/q² ⇒ p² = 2q². Thus p² is even, so p must be even (because the square of an odd number is odd). Write p = 2k; then (2k)² = 2q² ⇒ 4k² = 2q² ⇒ q² = 2k². Hence q² is even, so q is even. But then p and q are both even, contradicting the assumption that they are coprime. Therefore no such p/q exists, and √2 is irrational.
假设 √2 = p/q，其中 p 和 q 是互质的正整数。两边平方得 2 = p²/q²，所以 p² = 2q²。因此 p² 是偶数，故 p 必为偶数（因为奇数的平方是奇数）。记 p = 2k；则 (2k)² = 2q² ⇒ 4k² = 2q² ⇒ q² = 2k²。于是 q² 是偶数，所以 q 也是偶数。但这意味着 p 和 q 都是偶数，与互质的假设矛盾。因此这样的 p/q 不存在，√2 是无理数。
2. How many trailing zeros are in 100!? | 100的阶乘末尾有多少个零？
Interviewers like this question because it tests whether you can translate a real‑world question into simple prime factor counting. A trailing zero comes from a factor of 10 = 2×5. Since there are plenty of factors of 2 in 100!, the number of trailing zeros is limited by the number of factors of 5.
面试官喜欢此题，因为它考验你能否将实际问题转化为简单的质因数计数。末尾零来自因子 10 = 2×5。由于 100! 中有充足的因子 2，末尾零的个数取决于因子 5 的个数。
Count multiples of 5 up to 100: 100/5 = 20. Each contributes at least one 5. But numbers like 25, 50, 75, 100 are multiples of 25 and contribute an extra 5. So floor(100/25) = 4 extra fives. Multiples of 125 (none under 100) would give more. Total powers of 5 = 20 + 4 = 24. Thus 100! ends with 24 zeros.
统计 100 以内 5 的倍数：100/5 = 20，每个至少贡献一个 5。但像 25, 50, 75, 100 这样的数是 25 的倍数，额外贡献一个 5。所以 floor(100/25) = 4 个额外的 5。125 的倍数（100 以内没有）会提供更多。5 的总幂次 = 20 + 4 = 24。因此 100! 末尾有 24 个零。
3. Climbing a staircase: 1 or 2 steps at a time | 爬楼梯：每次走1级或2级
A candidate is asked: “You can climb a staircase by taking either one step or two steps at a time. How many distinct ways can you reach the n‑th step?” The interviewer watches for pattern spotting and recursive thinking. After a few small cases you realise that the number of ways to step n, denoted an, satisfies an = an‑1 + an‑2 with a1=1, a2=2 — the Fibonacci sequence.
面试者被问到：“你可以一次走 1 级或 2 级台阶爬上楼梯。到达第 n 级台阶共有多少种不同的方法？”面试官观察模式识别和递归思维。通过几个小的情形，你会发现到达第 n 级的方法数 an 满足 an = an‑1 + an‑2，且 a1=1，a2=2 —— 这就是斐波那契数列。
Reasoning: To reach step n, your last move was either a single step from step n‑1, or a double step from step n‑2. Hence an = an‑1 + an‑2. You can then compute a few terms (1,2,3,5,8,13,…) and discuss closed forms or generating functions if the interview progresses, but the core insight is the recurrence.
推理：要到达第 n 级，最后一步要么是从第 n‑1 级跨 1 级，要么是从第 n‑2 级跨 2 级。因此 an = an‑1 + an‑2。然后你可以计算若干项（1,2,3,5,8,13,…），如果面试深入，可讨论通项公式或生成函数，但核心洞见在于这个递推关系。
4. Evaluate ∫ ex sin x dx | 计算 ∫ eˣ sin x dx
Integration by parts is routine, but this question checks whether you notice the cyclic pattern and can solve for the integral algebraically. Let I = ∫ ex sin x dx. The interviewer may ask you to choose u = sin x, dv = ex dx, or the other way round, and then guide you to apply the method twice.
分部积分是常规操作，但此题考察你是否能发现循环模式并代数求解积分。设 I = ∫ ex sin x dx。面试官可能会让你选择 u = sin x, dv = ex dx，或者反过来，然后引导你应用两次该方法。
First integration: let u = sin x, dv = ex dx ⇒ du = cos x dx, v = ex. Then I = ex sin x − ∫ ex cos x dx. For the new integral, set U = cos x, dV = ex dx ⇒ dU = −sin x dx, V = ex. So ∫ ex cos x dx = ex cos x + ∫ ex sin x dx = ex cos x + I. Substituting back: I = ex sin x − (ex cos x + I) = ex sin x − ex cos x − I ⇒ 2I = ex(sin x − cos x) ⇒ I = ½ ex(sin x − cos x) + C.
第一次分部：令 u = sin x, dv = ex dx ⇒ du = cos x dx, v = ex。则 I = ex sin x − ∫ ex cos x dx。对于新积分，设 U = cos x, dV = ex dx ⇒ dU = −sin x dx, V = ex。所以 ∫ ex cos x dx = ex cos x + ∫ ex sin x dx = ex cos x + I。代回：I = ex sin x − (ex cos x + I) = ex sin x − ex cos x − I ⇒ 2I = ex(sin x − cos x) ⇒ I = ½ ex(sin x − cos x) + C。
5. Maximising a rectangle inscribed in a semicircle | 最大化内接半圆的矩形面积
A favourite optimisation problem: “A rectangle is inscribed in a semicircle of radius R with its base on the diameter. Find the dimensions that maximise the area.” The interviewer expects you to set up a variable, express the area in terms of it, differentiate, and justify the maximum.
一道备受青睐的优化题：“一个矩形内接于半径为 R 的半圆，其底边落在直径上。求使面积最大的尺寸。”面试官希望你设定一个变量，用该变量表示面积，求导并说明最大值。
Place the semicircle as the top half of x² + y² = R², y ≥ 0. Let the rectangle have half‑width x (from –x to x) and height y. Then area A = 2x·y = 2x√(R² − x²). Differentiate with respect to x: dA/dx = 2√(R² − x²) − 2x²/√(R² − x²). Set to zero: 2(R² − x²) − 2x² = 0 ⇒ R² − 2x² = 0 ⇒ x = R/√2. Then y = √(R² − R²/2) = R/√2. So the rectangle is a square of side R/√2 when stood on the base, giving max area = 2x·y = 2(R/√2)(R/√2) = R².
将半圆置于 x² + y² = R², y ≥ 0 的上半平面。设矩形半宽为 x（从 –x 到 x），高为 y。面积 A = 2x·y = 2x√(R² − x²)。对 x 求导：dA/dx = 2√(R² − x²) − 2x²/√(R² − x²)。令导数为零：2(R² − x²) − 2x² = 0 ⇒ R² − 2x² = 0 ⇒ x = R/√2。则 y = √(R² − R²/2) = R/√2。因此该矩形是底边上边长为 R/√2 的正方形，最大面积 = 2x·y = 2(R/√2)(R/√2) = R²。
6. Handshakes at a party: introducing graph theory | 聚会握手问题：图论入门
“At a party, some people shake hands. Prove that the number of people who have shaken hands an odd number of times is even.” The interviewer uses this to see if you can model the situation with vertices and edges, and whether you can give a clean degree‑sum argument.
“在一次聚会上，有些人互相握手。证明握过奇数次手的人数为偶数。”面试官借此观察你能否用顶点和边来建模，以及能否给出简洁的度和论证。
Model each person as a vertex, each handshake as an edge. The degree of a vertex is the number of handshakes that person had. The sum of all degrees equals twice the number of edges (each handshake contributes 2 to the sum), so the sum is even. Let O be the set of vertices with odd degree; the rest have even degree. Sum of degrees = (sum over O) + (sum over evens). The evens contribute an even number, so the sum over O must be even. A sum of odd numbers is even only if the number of terms is even. Hence |O| is even.
将每个人看作一个顶点，每次握手看作一条边。顶点的度即此人握手的次数。所有顶点的度之和等于边数的两倍（每次握手对和贡献 2），因此总和为偶数。令 O 为度数为奇数的顶点集合，其余顶点度数为偶数。度和 = (O 中度和) + (偶度数之和)。偶度数之和为偶数，故 O 中度和必须为偶数。若干个奇数之和为偶数仅当项数为偶数，因此 |O| 是偶数。
7. Expected tosses for two consecutive heads | 连续两次正面朝上的期望抛掷次数
A typical probability states: “A fair coin is tossed repeatedly. What is the expected number of tosses to get two heads in a row?” This tests whether you can set up a Markov chain or a system of equations based on recent history.
一道典型的概率题：“反复抛掷一枚公平硬币。首次出现连续两次正面朝上所需的期望抛掷次数是多少？”这考察你能否根据最近的历史建立马尔可夫链或方程组。
Let E be the expected number from the start, and EH be the expected number given that the last toss was a Head (but we haven’t yet seen two in a row). From start: after first toss, if Tails (prob ½) we are still at start, having used one toss. If Heads (prob ½) we move to state H. So E = 1 + ½ E + ½ EH. From state H: next toss Tails (½) returns to start, having used one toss; Heads (½) finishes in one toss. So EH = 1 + ½ E + ½ * 0. Solve: E = 1 + ½ E + ½ (1 + ½ E) ⇒ E = 1 + ½ E + ½ + ¼ E = 1.5 + 0.75E ⇒ 0.25E = 1.5 ⇒ E = 6. So the expected number is 6 tosses.
设 E 为从初始状态开始的期望值，EH 为已知上一次抛掷为正面（但尚未出现连续两个正面）时的期望值。从初始：第一次抛掷后，若反面（概率 ½），我们仍处于初始状态，已用一次抛掷；若正面（概率 ½），进入状态 H。所以 E = 1 + ½ E + ½ EH。从状态 H：下一次若反面（½）回到初始，用掉一次；若正面（½）则在一次后结束。因此 EH = 1 + ½ E + ½ * 0。求解：E = 1 + ½ E + ½ (1 + ½ E) ⇒ E = 1 + ½ E + ½ + ¼ E = 1.5 + 0.75E ⇒ 0.25E = 1.5 ⇒ E = 6。所以期望抛掷次数为 6 次。
8. Incircle radius of a right triangle | 直角三角形的内切圆半径
Geometry appears frequently. “A right triangle has legs a and b, and hypotenuse c. Find the radius r of the incircle (the circle tangent to all three sides).” Interviewers enjoy seeing whether you use area dissection or standard formulas, and whether you can relate r to the semiperimeter.
几何题经常出现。“一直角三角形两直角边为 a 和 b，斜边为 c。求内切圆（与三边相切的圆）的半径 r。”面试官乐于看到你是用面积分割法还是标准公式，以及能否将 r 与半周长联系起来。
Approach: The area of the triangle is ½ab. The incircle centre is at distance r from each side; drawing radii to the points of tangency divides the triangle into three smaller triangles whose total area is ½ar + ½br + ½cr = r(a+b+c)/2. Equate: r(a+b+c)/2 = ½ab ⇒ r = ab/(a+b+c). This elegantly simplifies using c² = a² + b² but is usually left as such. An alternative using the fact that the two tangents from each vertex are equal gives r = (a+b−c)/2; both are equivalent.
方法：三角形面积为 ½ab。内切圆圆心到各边的距离均为 r；画出到切点的半径可将三角形分割为三个小三角形，总面积 = ½ar + ½br + ½cr = r(a+b+c)/2。令其等于原面积：r(a+b+c)/2 = ½ab ⇒ r = ab/(a+b+c)。此式可借助 c² = a² + b² 优美化简，但通常保留此形式。另一种利用从各顶点出发的两条切线长相等的性质可得到 r = (a+b−c)/2；两者等价。
9. Convergence of a recursive sequence | 递推数列的收敛性
A common analysis question: “Define a sequence by a1 = 1, an+1 = √(2 + an). Prove it converges and find the limit.” The interviewer explores boundedness and monotonicity, sometimes hinting at induction.
一道常见的分析题：“定义数列 a1 = 1，an+1 = √(2 + an)。证明其收敛并求极限。”面试官会探讨有界性和单调性，有时会提示使用归纳法。
First show boundedness: all terms are positive. Claim an < 2 for all n. Base: a1=1<2; assume an <2, then an+1 = √(2 + an) < √(2+2) = 2. So the sequence is bounded above by 2. For monotonicity: a2 = √3 > 1 = a1. Suppose an > an‑1. Then an+1 = √(2+an) > √(2+an‑1) = an. By induction it is strictly increasing. A bounded increasing sequence converges. Let limit L satisfy L = √(2 + L). Square: L² = 2 + L ⇒ L² − L − 2 = 0 ⇒ (L−2)(L+1)=0 ⇒ L=2 (reject −1). The limit is 2.
首先证明有界性：所有项为正。断言对所有 n 有 an < 2。奠基：a1=1<2；假设 an <2，则 an+1 = √(2 + an) < √(2+2) = 2。因此数列以 2 为上界。单调性：a2 = √3 > 1 = a1。假设 an > an‑1，则 an+1 = √(2+an) > √(2+an‑1) = an。由归纳法知数列严格递增。有界递增数列必收敛。设极限 L 满足 L = √(2 + L)。平方：L² = 2 + L ⇒ L² − L − 2 = 0 ⇒ (L−2)(L+1)=0 ⇒ L=2（舍去 −1）。极限为 2。
10. Knights and Knaves logic puzzle | 骑士与无赖的逻辑趣题
To test precise logical thinking, an interviewer might present a version of the classic puzzle: “On an island, Knights always tell the truth and Knaves always lie. You meet two inhabitants, A and B. A says ‘Exactly one of us is a Knight.’ What can you determine?” The challenge is to navigate the cases without contradiction.
为了测试严谨的逻辑思维，面试官可能会给出经典谜题的变体：“在一个岛上，骑士永远说真话，无赖永远说假话。你遇到两位居民 A 和 B。A 说：‘我们当中恰好有一位是骑士。’你能推断出什么？”挑战在于无矛盾地遍历各种情形。
Case 1: A is a Knight. Then his statement is true; exactly one is a Knight. Since A is a Knight, B must be a Knave. Check: A Knight tells truth, B Knave lies — consistent. Case 2: A is a Knave. Then his statement is false. The truth is that it is not true that exactly one is a Knight. This means either both are Knights or both are Knaves. But A is a Knave, so both being Knights is impossible. Thus both must be Knaves. In this case B is also a Knave. Check: A Knave says “exactly one Knight”, which is false since there are zero Knights — consistent. So we cannot uniquely determine B; both cases (A Knight/B Knave or A Knave/B Knave) are possible unless the puzzle asks a more specific question. The lesson: clear case analysis reveals which information is missing.
情形 1：A 是骑士。则其陈述为真；恰好有一位骑士。既然 A 是骑士，B 必为无赖。验证：骑士说真话，无赖说假话——自洽。情形 2：A 是无赖。其陈述为假。实际情况并非“恰好有一位骑士”，这意味着要么两人都是骑士，要么两人都是无赖。但 A 是无赖，两人都是骑士不可能，因此两人必须都是无赖。此时 B 也是无赖。验证：无赖 A 说“恰好有一位骑士”，因实际有零个骑士故为假——自洽。因此我们无法唯一确定 B；两种情形（A 骑 B 无，或 A 无 B 无）均可能，除非问题进一步限制。启示：清晰的情形分析能揭示所缺失的信息。
Published by TutorHao | Mathematics 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