📚 How the TMUA Single Aggregate Score Works: Building a 1 to 9 Mark | TMUA 单一总分制：1 到 9 分是如何构成的
Since the 2024 admissions cycle, the Test of Mathematics for University Admission no longer hands you two separate numbers. You receive one aggregate score between 1.0 and 9.0, reported to one decimal place. That single figure is what admissions tutors at Cambridge, Imperial, LSE, Warwick and elsewhere place beside your A Level predictions, so understanding exactly how it is constructed is not academic curiosity - it is strategy.
自 2024 申请季起，TMUA（大学入学数学考试）不再给出两个独立分数，而只给出一个 1.0 至 9.0 之间的总分，保留一位小数。剑桥、帝国理工、LSE、华威等院校的招生官，就是把这个数字和你的 A Level 预估成绩并排来看。因此，搞清楚这个分数到底怎么构成，不是学术好奇心，而是实打实的备考策略。
1. What the TMUA Is and Why One Number Now Matters | TMUA 是什么，为什么现在只剩一个数字
The TMUA is a two-and-a-half-hour admissions test used by a group of UK universities for mathematically demanding degrees: Computer Science, Economics, Mathematics, and joint honours combinations of these. It is administered by UAT-UK on behalf of the participating institutions, and it is sat in October (and, in some cycles, January) of the application year.
TMUA 是一场时长两个半小时的入学笔试，被一批英国大学用于筛选数学要求较高的专业：计算机科学、经济学、数学，以及这些方向的联合荣誉学位。考试由 UAT-UK 代表参与院校组织，通常在申请当年的 10 月举行（部分考季还有 1 月场次）。
Before 2024, candidates received two scores - one for Paper 1 and one for Paper 2, each on its own 1.0 to 9.0 scale. Universities then had to interpret two numbers, often weighting them differently. From 2024 the reporting was simplified: the two papers are combined into a single aggregate score on the same 1.0 to 9.0 scale.
2024 年以前，考生会拿到两个分数：Paper 1 一个、Paper 2 一个，各自都是 1.0 到 9.0 的量表分。大学因此需要解读两个数字，而且不同院系的加权方式还不一样。从 2024 年起，成绩报告方式被简化：两卷合并为同一个 1.0 到 9.0 量表上的单一总分。
2. From Two Separate Scores to One Aggregate | 从两卷各自计分到单一总分
The change was driven by three practical pressures. First, a single number is easier for admissions tutors to compare across a large applicant pool. Second, it removes the temptation for candidates to treat one paper as optional. Third, it aligns the TMUA with how other admissions tests report a headline result.
这一改动背后有三个现实压力。第一，单一数字更便于招生官在庞大的申请池里横向比较。第二，它打消了考生把某一卷当作"可放弃项"的念头。第三，它让 TMUA 与其他入学笔试的成绩呈现方式保持一致。
What did not change is the underlying assessment. Both papers are still compulsory, still equally timed, and still contribute to the final outcome. The aggregate is not an average of two scale scores; it is built from the combined raw performance across both papers.
没有改变的是考试本身。两卷仍然都是必考，时长相同，也都计入最终结果。这个总分不是两个量表分的简单平均，而是由两卷合起来的原始表现构建出来的。
3. The Raw Material: Two Papers, 40 Raw Marks | 原始材料：两卷共 40 分原始分
Everything starts with raw marks. Each paper contains 20 multiple-choice questions, each worth one mark, with five options per question. There is no negative marking, so an unanswered question and a wrong answer are treated identically - both score zero.
一切从原始分开始。每卷 20 道选择题，每题 1 分，每题五个选项。考试不设倒扣分，所以空着不答和答错在计分上完全一样，都是 0 分。
- Paper 1 - Applications of Mathematical Knowledge: 20 questions, 75 minutes. Tests whether you can deploy standard A Level mathematics fluently under time pressure.
- Paper 2 - Mathematical Reasoning: 20 questions, 75 minutes. Tests logical deduction, proof, necessary and sufficient conditions, and elementary mathematical argument.
- Paper 1 数学知识的应用：20 题，75 分钟。考查你能否在时间压力下熟练调用标准 A Level 数学工具。
- Paper 2 数学推理：20 题，75 分钟。考查逻辑推演、证明、充分必要条件以及初等数学论证。
So the raw space is symmetric:
因此，原始分的空间是对称的：
Paper 1 raw (max 20) + Paper 2 raw (max 20) = Total raw (max 40)
Paper 1 原始分（满分 20）+ Paper 2 原始分（满分 20）= 总原始分（满分 40）
Each paper therefore carries exactly 50% of the available raw marks. This equal weighting is the single most important structural fact about the TMUA, and it drives almost every sensible revision decision.
也就是说，每卷恰好各占总原始分的 50%。这个等权结构是 TMUA 最关键的一条结构性事实，几乎每一个合理的复习决策都由它推导而来。
4. How Raw Marks Become a 1.0 to 9.0 Scale | 原始分如何转换为 1.0 到 9.0 量表分
The conversion from 40 raw marks to a 1.0 to 9.0 score is not a linear percentage. If it were, 40 out of 40 would be 9.0 and 20 out of 40 would be 4.5, but that is not how it works. The scale is constructed statistically so that the reported score reflects a candidate's position relative to a reference cohort, not simply a proportion of questions answered correctly.
从 40 分原始分到 1.0 到 9.0 的转换，并不是线性百分比。如果真是线性，那 40 分就是 9.0、20 分就是 4.5，但事实并非如此。这个量表是用统计方法构建的，报告出来的分数反映的是考生相对于参照考生群体的位置，而不只是答对题目的比例。
The construction can be summarised as a function of several inputs:
这个构建过程可以概括为若干个输入变量的函数：
Scale Score = f(Total Raw, Paper Difficulty, Cohort Ability)
- The total raw mark supplies the core signal.
- Paper difficulty adjusts for the fact that no two sittings contain identical questions.
- Cohort ability calibrates the scale so that a given score means the same thing in different years.
- 总原始分提供核心信号。
- 试卷难度用于修正"没有两次考试题目一模一样"这一事实。
- 考生群体能力用于校准量表，使同一个分数在不同年份含义一致。
Crucially, the reported range starts at 1.0, not 0. A candidate who answers almost nothing correctly still receives 1.0 rather than 0.0, and 9.0 is reserved for an extremely small fraction of the cohort rather than being a literal "perfect paper" marker.
关键在于，这个量表的下限是 1.0 而不是 0。几乎全部答错的考生拿到的也是 1.0 而不是 0.0；而 9.0 只留给极少数考生，并不是"满分卷"的字面标志。
5. Equating and Comparability Across Sessions | 等值调整与跨考次的分数可比性
Because different candidates sit different papers in different sessions, the scores must be made comparable. This process is called equating. Statistically, items and forms are linked through shared questions or through anchor candidates, allowing the awarding body to say that a 6.5 in October means the same standard as a 6.5 in January.
由于不同考次、不同考生做的是不同卷子，分数就必须变得可比。这个过程叫等值调整（equating）。在统计上，题目和卷本通过共同题或锚定考生建立联系，使考试机构能够宣称：10 月的 6.5 和 1 月的 6.5 代表同一水平。
Two consequences follow. First, there is no permanent, published raw-mark-to-scale-score conversion table you can memorise, because the mapping is recalculated for each cycle. Second, a hard paper and an easy paper can both produce the same score for equally able candidates - which is why comparing raw marks between years is meaningless.
由此产生两个结果。第一，不存在一张你可以背下来的、永久公开的原始分与量表分换算表，因为换算关系每个考季都会重新计算。第二，难题卷和简单卷对于能力相同的考生可以给出同样的分数——所以拿不同年份的原始分互相比，是没有意义的。
6. What the Distribution Looks Like | 分数分布大致长什么样
TMUA scores cluster in the middle and thin out sharply at the top. The distribution is roughly bell-shaped, with the bulk of candidates landing between about 3.0 and 7.0, and a long, sparse tail reaching up to 9.0. Very few candidates score above 7.5.
TMUA 分数集中在中段，越往顶部越急剧收窄。整体大致呈钟形分布，大多数考生落在约 3.0 到 7.0 之间，向上拖出一条稀疏的长尾直到 9.0。超过 7.5 的考生非常少。
| Score Band | Approximate Position in Cohort | 分数段 | 在考生群体中的大致位置 |
| 8.0 - 9.0 | Top 1-3% | 8.0 - 9.0 | 前 1-3% |
| 7.0 - 7.9 | Roughly top 10% | 7.0 - 7.9 | 大致前 10% |
| 6.0 - 6.9 | Roughly top 25% | 6.0 - 6.9 | 大致前 25% |
| 5.0 - 5.9 | Above average | 5.0 - 5.9 | 高于平均 |
| 4.0 - 4.9 | Around the median | 4.0 - 4.9 | 接近中位数 |
| 1.0 - 3.9 | Below average | 1.0 - 3.9 | 低于平均 |
The important lesson is that the median sits near 4.5, so a score of 5.5 or 6.0 is genuinely good, and 7.0 is a strong result that puts you in a small minority of all test-takers worldwide.
这里最重要的一点是：中位数大约在 4.5 附近，所以 5.5 或 6.0 已经是相当不错的成绩，而 7.0 已经是能让全球考生中很小一部分人才拿到的好成绩。
7. Reading the Scale: What Each Band Tells You | 读懂量表：各分数段意味着什么
An aggregate score is not just a number; it is a compressed description of how you performed across two very different types of mathematical thinking. Reading it correctly requires knowing what the scale deliberately does and does not measure.
总分不只是一个数字，它是对你在两种截然不同的数学思维上表现的压缩描述。要正确解读它，就必须知道这个量表有意测量什么、又刻意不测量什么。
- It measures problem-solving under time pressure, not breadth of syllabus knowledge.
- It ranks you against a competitive international cohort, not against your school class.
- It rewards consistency across both papers, because a collapse on one paper cannot be rescued by brilliance on the other.
- It does not forgive careless errors on easy questions, because those marks are the cheapest available.
- 它测的是时间压力下的解题能力，不是大纲知识的广度。
- 它把你放在竞争激烈的国际考生群体中排序，而不是和你的同班同学比。
- 它奖励两卷的稳定性，因为一卷崩盘无法靠另一卷的超常发挥来挽救。
- 它不会原谅简单题上的粗心失误，因为那是最便宜的分数。
8. Estimating Raw Marks Behind a Score | 估算分数背后的原始分
Candidates naturally want to translate a target score into "how many questions must I get right". Because the official conversion is recalculated each cycle, any such table is an estimate, but rough estimates are still useful for planning.
考生自然会想把这个目标分数翻译成"我得做对多少题"。由于官方换算每个考季都会重算，任何这类表格都只是估算，但粗略的估算对规划仍然有用。
| Aggregate Score | Indicative Total Raw (out of 40) | 总分 | 参考总原始分（满分 40） |
| 8.0+ | 33 - 40 | 8.0+ | 33 - 40 | Published by TutorHao | Exam Prep Revision Series | aleveler.com
