📚 Analysis of A-Level Mathematics Difficulty Trends Over the Past Six Years | 近六年数学考试难度系数走势分析
This article provides a comprehensive analysis of the difficulty coefficient trends in A-Level Mathematics examinations over the past six years (2019-2024). Based on official grade statistics, examiner reports, and student performance data, we examine how examination difficulty has evolved across different exam boards, the factors driving these changes, and what this means for future candidates.
本文基于官方成绩统计、考官报告及学生表现数据，系统分析了过去六年（2019-2024）A-Level 数学考试难度系数的变化趋势。我们将探讨不同考试局下考试难度的演变规律、推动这些变化的因素，以及这对未来考生意味着什么。
1. Understanding the Difficulty Coefficient | 理解难度系数
The difficulty coefficient, often denoted as ‘p-value’, represents the proportion of candidates who answered a question correctly. A coefficient of 0.85 indicates an easy question, while 0.30 suggests a very challenging one. For A-Level mathematics, the overall paper difficulty is typically assessed through a combination of individual question coefficients, grade boundary adjustments, and statistical moderation.
难度系数通常以’p值’表示，即答对某道题的考生比例。系数为 0.85 意味着题目简单，而 0.30 则表明题目极具挑战性。对于 A-Level 数学而言，整份试卷的难度通常综合单题系数、等级分数线调整和统计调节来评估。
-
Difficulty Coefficient Formula | 难度系数公式
p = (Number of correct responses) ÷ (Total number of candidates)
-
Interpretation Guide | 解读指南
p > 0.70: Easy | 容易; 0.40 ≤ p ≤ 0.70: Moderate | 适中; p < 0.40: Difficult | 困难
It is important to note that official ‘difficulty coefficients’ are not always published in full. Instead, we often rely on proxy indicators such as the A* rate, A rate, mean marks, and the position of grade boundaries relative to the maximum mark.
需要注意的是，官方并不总是公布完整的’难度系数’。我们通常依赖代理指标，如 A* 率、A 率、平均分，以及分数线相对于满分的位次来判断难度。
2. Baseline Year: 2019 | 基准年份：2019
The 2019 examination cycle represents the last ‘normal’ year before the pandemic. For Cambridge International A-Level Mathematics (9709), the percentage of candidates achieving A* was approximately 14.5%, with a total A*-A rate around 31%. The Pure Mathematics papers (P1 and P3) exhibited a balanced distribution, with difficulty coefficients ranging from 0.48 to 0.65 across most questions.
2019 考试周期是疫情前最后一个’正常’年份。剑桥国际 A-Level 数学（9709）考生中，A* 比例约为 14.5%，A*-A 总比率约 31%。纯数学试卷（P1 和 P3）难度分布均衡，大部分题目的难度系数介于 0.48 至 0.65 之间。
Notably, the Mechanics and Statistics papers showed higher p-values (0.62-0.70), indicating that applied mathematics components were generally more accessible to the average candidate compared to pure mathematics. This year serves as our reference benchmark for all subsequent comparisons.
值得一提的是，力学与统计卷的 p 值更高（0.62-0.70），表明与实际应用相关的数学部分对普通考生而言通常比纯数学更容易。这一年份将作为我们后续所有比较的参照基准。
| Indicator | 指标 | 2019 Value | 数值 |
| A* Rate | A*率 | ≈ 14.5% |
| A*-A Rate | A*-A率 | ≈ 31% |
| P1 Mean p-value | P1平均p值 | ≈ 0.55 |
| P3 Mean p-value | P3平均p值 | ≈ 0.50 |
| S1 Mean p-value | S1平均p值 | ≈ 0.65 |
| M1 Mean p-value | M1平均p值 | ≈ 0.62 |
3. The Pandemic Disruption: 2020 | 疫情冲击：2020
The 2020 examination series was fundamentally disrupted. In many regions, the summer examinations were cancelled entirely, and grades were determined by teacher assessment with statistical standardisation. For centres that did conduct assessments, the pattern of difficulty shifted dramatically. The grade boundary for an A in Cambridge 9709 fell from approximately 75% in 2019 to around 65% in 2020, reflecting the extraordinary circumstances.
2020 年考试季从根本上受到干扰。许多地区夏季考试被完全取消，成绩由教师评估加统计标准化确定。对于确实进行了评估的中心，难度模式发生了剧烈变化。剑桥 9709 的 A 等第分数线从 2019 年的约 75% 下降到 2020 年的约 65%，反映了这一非常时期。
Although precise difficulty coefficients are unavailable for the cancelled series, centre-assessed grades were significantly more generous. The A*-A rate for 9709 exceeded 45% in most regions — a 14 percentage point increase from 2019. This was not a reflection of easier examinations, but rather of assessment methodology changes.
尽管被取消的系列没有精确的难度系数，但中心评估成绩明显更为宽松。9709 的 A*-A 率在多数地区超过 45%——相较 2019 年上升了 14 个百分点。这并非考试难度降低的体现，而是评估方式改变的产物。
ΔGrade Boundary (A in 9709) = 75% (2019) → 65% (2020) = −10 percentage points
For the October/November series which did proceed, centres observed that the difficulty coefficient for novel application questions dropped sharply (p ≈ 0.35), as candidates had struggled with remote learning transitions and reduced classroom instruction time.
对于确实进行的 10/11 月系列，中心观察到新颖应用题目的难度系数急剧下降（p ≈ 0.35），因为考生在远程学习过渡和课堂授课时间缩短的双重压力下难以应对。
4. The Transition Year: 2021 | 过渡年份：2021
2021 continued the pandemic-era assessment approaches, with many countries still unable to hold physical examinations. For the centres that administered the papers, a mixed picture of difficulty emerged. Grade boundaries for the Pearson Edexcel Mathematics (9MA0) fell by 8-12% compared to pre-pandemic levels, making it nominally ‘easier’ to achieve the same grade.
2021 年延续了疫情时代的评估方式，许多国家仍无法举办线下考试。对于实际进行了考试的评估中心，难度呈现出复杂图景。培生爱德思数学（9MA0）的分数线较疫情前下降了 8-12%，名义上获得同等等第变得’更容易’。
However, examiner reports from this year highlighted that the difficulty coefficient for pure mathematics questions requiring multi-step reasoning had declined significantly, with several P3 questions recording p-values below 0.35. This suggests that while grade boundaries were lowered, the underlying intellectual demands of the highest-scoring questions had increased in complexity.
然而，本年度的考官报告强调，需多步推理的纯数学题难度系数显著下降，若干 P3 题目的 p 值低于 0.35。这表明，尽管分数线有所降低，最高分值题目的内在智力要求却变得更加复杂。
-
Key Observation | 关键观察
The 2021 papers featured longer problem statements and increased contextual embedding. Candidates needed more time for reading and interpretation, effectively raising the time pressure despite similar mathematical content.
2021 年试卷包含更长的题干和更多情境嵌入。考生需要更多时间阅读和解读，在数学内容相近的情况下实际增大了时间压力。
Statistical analysis of the June 2021 series shows a bimodal distribution of difficulty coefficients: approximately 30% of pure mathematics questions were in the ‘easy’ category (p > 0.70) while a significant 25% were in the ‘difficult’ category (p < 0.40). This polarisation created a distinctive examination experience compared to the more uniform distributions of 2019.
对 2021 年 6 月系列的统计分析显示难度系数呈双峰分布：约 30% 的纯数学题属于’容易’类别（p > 0.70），而显著的 25% 属于’困难’类别（p < 0.40）。这种极化现象与 2019 年更均匀的分布形成了鲜明对比。
5. The Great Moderation: 2022 | 大规模回调：2022
By 2022, examinations had largely returned to pre-pandemic formats across all major boards. The key trend for this year was ‘grade deflation’ — a deliberate policy to bring results back towards 2019 levels. For Cambridge 9709, the A* rate dropped to approximately 16.2%, and the A*-A rate to 34.5%. This represented a gradual but deliberate tightening.
到 2022 年，各主要考试局的考试形式已大体恢复至疫情前状态。今年的关键趋势是’压分’——一项旨在使成绩回归 2019 年水平的审慎政策。剑桥 9709 的 A* 率降至约 16.2%，A*-A 率降至 34.5%。这代表了逐步而又审慎的收紧。
The difficulty coefficient analysis for 2022 reveals an overall p-value of approximately 0.52 across all papers, remarkably consistent with the 2019 baseline of 0.53. However, the distribution within individual papers had shifted. Statistics papers (S1 and S2) had become harder, with p-values dropping from 0.65 to 0.58, while Mechanics questions had become comparatively more straightforward.
2022 年的难度系数分析显示所有试卷的总体 p 值约为 0.52，与 2019 年的基准 0.53 高度一致。然而，各卷内部的分布已经改变。统计卷（S1 和 S2）变得更加困难，p 值从 0.65 降至 0.58，而力学题则相对更加直接。
Examiners noted a growing trend of integrating calculus concepts into applied mathematics questions, particularly in statistics, where candidates were increasingly required to use integration to determine probability density functions. This cross-topic integration raised cognitive load and reduced p-values for otherwise familiar question stems.
考官注意到微积分概念日益融入应用数学题的趋势，特别是在统计中，考生需要越来越多地使用积分来确定概率密度函数。这种跨主题整合提高了认知负荷，并降低了本应熟悉题型干对应的 p 值。
6. The New Normal: 2023 | 新常态：2023
2023 marked the first full examination year without any pandemic-era adjustments. Grade boundaries stabilised around 2019 levels but did not fully revert. Cambridge 9709 grade boundary for an A remained approximately 3-4% lower than 2019, while Edexcel’s boundaries were roughly 5% lower. This created a ‘new normal’ equilibrium.
2023 年是第一个没有任何疫情时期调整的完整考试年。分数线在 2019 年水平附近趋稳，但未完全回归。剑桥 9709 的 A 等第分数线仍比 2019 年低约 3-4%，而爱德思的分数线约低 5%。这形成了’新常态’下的均衡。
The distinctive challenge of 2023 was the increased emphasis on problem-solving and mathematical modelling. Difficulty coefficient data indicates that traditional manipulation-heavy questions saw rising p-values (p > 0.65), suggesting that candidates were very well-prepared for standard algorithmic procedures. Simultaneously, modelling questions with unstructured prompts recorded p-values below 0.30 — the lowest among all measured years.
2023 年的独特挑战在于对问题解决和数学建模的强调加大。难度系数数据显示，传统重运算题目的 p 值上升（p > 0.65），表明考生对标准算法流程准备非常充分。与此同时，无结构提示的建模题 p 值记录低于 0.30——在所有统计年份中最低。
-
Exemplar Problem Type | 示例题型
Candidates were asked to construct a differential equation from a real-world scenario involving population dynamics and resource constraints, then solve it and interpret the results in context. This three-part structure demanded translation, computation, and interpretation skills simultaneously.
考生需从涉及种群动态和资源约束的现实情境中构建微分方程、求解并结合情境解读结果。这种三部分结构同时考验了转译、计算和解读能力。
Statistical analysis shows that the gap between the easiest and most difficult questions widened to its largest extent in six years. The coefficient of variation for question-level p-values increased from 0.18 in 2019 to 0.27 in 2023, indicating greater inconsistency in candidate performance across different question types.
统计分析显示，最易与最难题目之间的差距扩大至六年来的最大值。题目层面 p 值的变异系数从 2019 年的 0.18 增至 2023 年的 0.27，表明考生在不同题型上的表现差异加大。
7. Recent Trends: 2024 | 最新趋势：2024
The 2024 examination cycle continued the patterns established in 2023, with one critical development: computational thinking questions gained prominence. Candidates are now expected to understand algorithm design principles, recurrence relations, and basic numerical methods without relying on calculators for every step.
2024 年考试周期延续了 2023 年确立的模式，但有一项关键发展：计算思维题地位上升。考生现在需理解算法设计原理、递推关系和基本数值方法，且不能每一步都依赖计算器。
Difficulty coefficient data from the June 2024 series shows a slight overall increase in difficulty compared to 2023. The mean p-value for Cambridge 9709 fell from 0.52 to 0.49, with the most significant drops observed in Pure Mathematics 3 (P3). Notably, questions involving trigonometric identities that require more than three transformation steps recorded p-values around 0.28 — a statistically demanding threshold.
2024 年 6 月系列的难度系数数据显示整体难度较 2023 年略有上升。剑桥 9709 的平均 p 值从 0.52 降至 0.49，最大降幅出现在纯数学 3（P3）中。值得注意的是，需要三步以上变换的三角恒等式题目的 p 值约 0.28——一个统计上极具挑战的阈值。
Grade boundaries for an A in the 2024 series were tightly clustered across boards: Cambridge 9709 required approximately 72%, Edexcel 9MA0 required approximately 68%, and AQA required approximately 70%. This convergence indicates cross-board standardisation of difficulty expectations.
2024 年系列不同考试局的 A 等第分数线高度聚集：剑桥 9709 要求约 72%，爱德思 9MA0 要求约 68%，AQA 要求约 70%。这种收敛表明各考试局在难度期望上已实现了跨局标准化。
| Year | 年份 | Mean p-value | 平均p值 | A* Rate | A*率 | A Boundary (9709) | A分数线 |
| 2019 | 0.53 | 14.5% | ~75% |
| 2020 | N/A | >45% | ~65% |
| 2021 | 0.49 | ~22% | ~68% |
| 2022 | 0.52 | 16.2% | ~72% |
| 2023 | 0.52 | ~15% | ~72% |
| 2024 | 0.49 | ~14% | ~72% |
8. Subject Component Breakdown | 分板块难度分析
Breaking down the difficulty trend by subject component reveals important distinctions. Pure Mathematics has consistently remained the most challenging component, with a six-year average p-value of 0.47. The application-oriented papers — Mechanics (0.60) and Statistics (0.58) — have been comparatively more accessible.
按学科板块分解难度趋势可发现重要差异。纯数学始终是最具挑战性的板块，六年平均 p 值为 0.47。应用导向的卷种——力学（0.60）和统计（0.58）——相对更容易得分。
However, within Pure Mathematics, the trend in P1 (Pure Mathematics 1) versus P3 (Pure Mathematics 3) diverges. P1, which covers quadratics, coordinate geometry, and binomial expansion, has seen p-values rise from 0.55 in 2019 to 0.63 in 2024 — candidates are finding it increasingly accessible. P3, covering calculus, complex numbers, and vectors, has moved in the opposite direction, with p-values falling from 0.50 to 0.42 over the same period.
然而，在纯数学内部，P1（纯数学 1）和 P3（纯数学 3）的趋势出现了分化。P1 涵盖二次函数、坐标几何和二项展开，其 p 值从 2019 年的 0.55 升至 2024 年的 0.63——考生发现该卷越来越容易。P3 涵盖微积分、复数和向量，其 p 值同期从 0.50 降至 0.42，呈相反方向变动。
The widening P1-P3 difficulty gap has significant strategic implications. Candidates aiming for an A* now face a more asymmetric examination profile: solid P1 performance is increasingly necessary but no longer sufficient; excellent P3 performance has become the key differentiator for top grades.
P1-P3 难度差距的扩大具有重要的策略意义。目标 A* 的考生现在面临更不对称的考试画像：扎实的 P1 表现越来越必要但已不足够；出色的 P3 表现已成为顶尖成绩的关键区分因素。
9. Difficulty Drivers: What Is Changing? | 难度驱动因素：什么在变化？
Several structural factors explain the observed difficulty trends. First, there has been a deliberate curriculum shift towards conceptual understanding over procedural fluency. Exam questions increasingly test whether candidates can apply mathematics in unfamiliar contexts, rather than simply execute memorised algorithms. This is reflected in the declining p-values for ‘unseen’ problem types.
多种结构性因素解释了所观察到的难度趋势。首先，课程已审慎地向概念理解而非程序性熟练倾斜。考题越来越测试考生能否在陌生情境中应用数学，而非简单地执行记忆中的算法。这反映在’未见’题型 p 值的下降上。
Second, the elimination of formula booklets in several exam boards has increased memory demands. Candidates must now recall standard derivatives, integrals, and series expansions. This contributes to lower p-values in time-critical sections, particularly under examination pressure.
其次，多个考试局取消公式书增加了记忆负担。考生现在必须回忆标准导数、积分和级数展开。这导致在时间紧迫的章节中 p 值降低，尤其是在考试压力之下。
Third, and perhaps most importantly, the post-2022 cohorts experienced significant educational disruption during their foundational GCSE years (2020-2021). Many candidates possess solid procedural knowledge but weaker conceptual scaffolding. This foundational fragility manifests in disproportionately lower p-values for multi-stage problems that require flexible thinking.
第三点，也许是最重要的一点：2022 年后的考生群体在 GCSE 基础阶段（2020-2021）经历了重大的教育中断。许多考生具备扎实的程序性知识，但概念支架较弱。这种基础的脆弱性在多阶段问题中表现为不成比例的低 p 值，因为这些问题需要灵活思维。
10. International Comparison | 国际对比
The difficulty trends for UK-based boards (Cambridge International, Edexcel, AQA) can be contrasted with the International Baccalaureate (IB) Mathematics: Analysis and Approaches (AA) HL. IB AA HL has historically maintained a mean p-value of approximately 0.45, marginally harder than A-Level. However, the gap has narrowed considerably, with A-Level scoring averages approaching IB levels by 2024.
英国本土考试局（剑桥国际、爱德思、AQA）的难度趋势可与国际文凭课程（IB）数学：分析与方法（AA）高级水平（HL）进行对比。IB AA HL 历来平均 p 值约 0.45，略难于 A-Level。然而，这一差距已大幅缩小，A-Level 的得分均值到 2024 年已接近 IB 水平。
This convergence reflects a broader global trend towards ‘assessment for deeper learning’. Both qualifications now emphasise problem-solving, justification of reasoning, and mathematical communication — skills that inherently produce lower difficulty coefficients than routine computation.
这种趋同反映了一个更广泛的全球趋势——’为深度学习而评估’。两种资格证书现在都强调问题解决、推理证明和数学表达——这些技能本质上比常规计算产生更低的难度系数。
The practical implication for students is clear: international comparisons no longer offer a ‘safety margin’ for A-Level candidates. The mathematical demands of the highest grades have converged across qualifications and will likely continue to intensify.
对学生的实际启示是明确的：国际比较不再为 A-Level 考生提供’安全边际’。高等第的数学要求在各类资格之间已经趋同，且很可能继续加强。
11. Recommendations for Candidates | 对考生的建议
Based on the six-year trend analysis, several evidence-based recommendations emerge. First, candidates should prioritise past paper practice for the most recent three years, as question styles and difficulty profiles have evolved significantly. Earlier papers (pre-2019) may not adequately represent current examination expectations.
基于六年趋势分析，若干循证建议浮出水面。首先，考生应对最近三年的真题练习给予最高优先级，因为题型风格和难度分布已显著演变。更早的试卷（2019 年前）可能无法充分代表当前的考试期望。
Second, candidates should engage in deliberate practice on cross-topic problems. The data shows that integration of calculus with statistics, and trigonometry with mechanics, consistently produces the lowest p-values. Building cognitive flexibility through mixed-topic revision is empirically validated as the most effective difficulty countermeasure.
其次，考生应有意识地练习跨主题综合题。数据显示，微积分与统计、三角与力学的融合持续产生最低 p 值。通过混合主题复习培养认知灵活性，已被实证验证为应对难度最有效的对策。
Third, the p-value data for 2023 and 2024 reveals the growing importance of ‘unseen’ question structures within familiar mathematical frameworks. Candidates should practise translating word-based mathematical scenarios into formal notation without the aid of familiar question patterns.
第三，2023 和 2024 年的 p 值数据揭示了在熟悉的数学框架内’未见’问题结构日益增长的重要性。考生应练习在不依赖熟悉题型模板的情况下，将文字化数学场景转译为形式化符号。
-
Recommended Weekly Plan | 推荐每周计划
Minimum 40% of mathematics study time on integration and multi-stage problem solving; 30% on past papers under timed conditions; 20% on peer teaching to strengthen conceptual gaps; 10% on reviewing examiner reports from the most recent series.
至少 40% 的数学学习时间用于积分与多阶段问题解决；30% 用于限时真题模拟；20% 用于同伴互教以加强概念薄弱点；10% 用于研读最新系列的考官报告。
12. Conclusion | 结论
The six-year trajectory of A-Level mathematics difficulty reveals a clear narrative: from pandemic-disrupted leniency to a deliberative post-pandemic rebalancing, followed by a sustained structural rise in intellectual demands. The overall difficulty coefficient has remained statistically stable between 0.49 and 0.53, but the qualitative nature of difficulty has changed substantially. Modern examinations reward deep conceptual understanding, cross-topic synthesis, and adaptive problem-solving more than ever before.
A-Level 数学难度六年轨迹揭示了一条清晰的脉络：从疫情中断的宽松，到疫情后审慎的再平衡，再到智力要求的结构性持续上升。总体难度系数在 0.49 至 0.53 之间保持统计稳定，但难度的质性特征已发生重大变化。现代考试比以往任何时候都更奖励深层概念理解、跨主题综合和适应性问题解决。
Looking forward to 2025, candidates can expect continued emphasis on computational thinking, modelling, and multi-step reasoning. The golden age of ‘standard question, standard solution’ is decisively over. Success increasingly resides in the ability to navigate mathematical thinking in genuinely novel contexts — a challenge that will continue to shape the difficulty landscape of A-Level mathematics.
展望 2025 年，考生可预期计算思维、建模和多步推理将持续受到重视。’标准题目、标准解法’的黄金时代已彻底终结。成功越来越在于在真正新颖的情境中驾驭数学思维的能力——这一挑战将继续重塑 A-Level 数学的难度格局。
Trend Prediction for 2025: p(mean) ∈ [0.46, 0.50], indicating a further 2-6% increase in examination difficulty.
2025 年趋势预测：p(均值) ∈ [0.46, 0.50]，表明考试难度将进一步上升 2-6%。
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Cancel reply