📚 The judicial review process | 司法审查程序
At first glance, judicial review might seem purely a matter of law and politics, but hidden beneath every decision is a framework of logic, probability, and decision theory – the very mathematics that A-Level students study. From assessing the weight of evidence to quantifying the standard of proof, mathematical reasoning provides the backbone for fair judicial outcomes. In this article, we explore the judicial review process through the lens of Edexcel A-Level Mathematics, connecting hypothesis testing, Bayesian inference, and statistical errors to real-world courtrooms.
乍看之下，司法审查似乎纯粹是法律与政治问题，但隐藏在每个裁决背后的，是一套逻辑、概率和决策理论的框架——这正是 A-Level 学生所学的数学。从评估证据分量到量化证明标准，数学推理为公正的司法结果提供了支撑。本文将从爱德思 A-Level 数学的视角审视司法审查过程，将假设检验、贝叶斯推断和统计错误与现实法庭联系起来。
1. Understanding Judicial Review | 理解司法审查
Judicial review is the process by which courts examine the lawfulness of decisions made by public bodies. It is not about whether a decision was right or wrong on the facts, but whether it was made in accordance with the law. At its core, this involves a logical appraisal of whether the decision-maker acted within their powers and followed proper procedures.
司法审查是法院审查公共机构所做决定是否合法的过程。其重点不在于根据事实判断决定是否正确，而在于决定是否依法作出。从核心来看，这涉及对决策者是否在权限范围内行事并遵循正当程序进行逻辑评估。
Mathematically, this mirrors the concept of checking assumptions and validity in a proof. We ask: given the evidence presented, is the conclusion logically sound? Just as in hypothesis testing, where we examine whether sample data provide sufficient grounds to reject a null hypothesis, judicial review asks whether the reasoning process supports the outcome.
在数学上，这类似于在证明中检查假设和有效性的概念。我们会问：鉴于所呈证据，结论在逻辑上是否成立？正如在假设检验中，我们检查样本数据是否提供了拒绝零假设的充分理由，司法审查也会问推理过程是否支持该结果。
l terms, this is analogous to choosing a significance level (α). A stricter standard, like 'irrationality', demands a very high bar, much like setting α = 0.01 rather than 0.05.
一个基本概念是审查标准：根据情境不同，会适用不同程度的审查。用数学术语来说，这类似于选择显著性水平（α）。更严格的标准，例如“非理性”标准，要求极高的门槛，就像将 α 设为 0.01 而非 0.05。
2. The Burden of Proof and Likelihood Ratios | 举证责任与似然比
In any judicial review, the burden of proof determines who must convince the court. Usually, the claimant must show that the decision was unlawful. This concept aligns with the idea of likelihood ratio tests in statistics, where we compare the probability of observing the evidence under two competing hypotheses.
在任何司法审查中，举证责任决定了谁必须说服法院。通常，申请人必须证明该决定不合法。这一概念与统计学中的似然比检验相一致，即我们比较在两个对立假设下观察到证据的概率。
Let H₀ represent the null hypothesis: 'the decision is lawful'. H₁ is the alternative: 'the decision is unlawful'. The likelihood ratio (LR) is given by:
令 H₀ 表示零假设：“决定合法”。H₁ 为备择假设：“决定不合法”。似然比（LR）由下式给出：
LR = P( Evidence | H₁ ) / P( Evidence | H₀ )
A large LR suggests the evidence is much more likely under H₁, shifting the burden towards the public body. In judicial review, the intensity of scrutiny can be thought of as the threshold LR required to declare a decision unlawful.
较大的 LR 表明证据在 H₁ 下出现的可能性更大，从而将举证责任转向公共机构。在司法审查中，审查强度可被视为宣布决定不合法所需的 LR 阈值。
Courts often use phrases like 'sufficiently weighty' evidence, which mathematically translates into a likelihood ratio exceeding a certain critical value, analogous to a test statistic exceeding a critical value in a hypothesis test.
法院经常使用“足够分量”的证据等措辞，这在数学上可以转化为似然比超过某一临界值，类似于假设检验中检验统计量超过临界值。
3. Type I and Type II Errors in Justice | 司法中的第一类和第二类错误
A core concept in A-Level statistics is the possibility of errors in decision-making. In judicial review, a Type I error (false positive) occurs when a court wrongly declares a lawful decision unlawful. A Type II error (false negative) is when a truly unlawful decision is upheld.
A-Level 统计学中的一个核心概念是决策中可能出现的错误。在司法审查中，第一类错误（假阳性）发生在法院错误地宣布一个合法的决定为不合法时。第二类错误（假阴性）则是实际不合法的决定被维持。
The consequences of these errors shape legal principles. Minimising Type I errors protects public bodies from excessive interference, while minimising Type II errors safeguards citizens' rights. This trade-off is captured in the table below:
这些错误的后果塑造了法律原则。最小化第一类错误可以保护公共机构免受过度干预，而最小化第二类错误则保障公民权利。这种权衡体现在下表中：
| Truth / 真实情况 | Decision Lawful / 决定合法 | Decision Unlawful / 决定不合法 |
|---|---|---|
| Accepted lawful / 裁定合法 | Correct / 正确 | Type II Error / 第二类错误 (β) |
| Declared unlawful / 宣告不合法 | Type I Error / 第一类错误 (α) | Correct / 正确 |
In judicial review, the significance level α is often set low by conventions such as 'Wednesbury unreasonableness', which demands a decision so absurd that no reasonable authority would have made it – a very high critical value indeed.
在司法审查中，通常通过诸如“温斯伯里不合理性”等惯例将显著性水平 α 设定得很低，该标准要求决定荒谬到任何合理当局都不会做出的程度——这确实是一个非常高的临界值。
4. Bayesian Inference in Courtroom Decisions | 法庭决策中的贝叶斯推断
Bayesian statistics provides a natural framework for updating beliefs in the light of new evidence. In judicial review, judges start with a prior probability that a public body’s decision is lawful, and then update this based on the evidence presented.
贝叶斯统计为根据新证据更新信念提供了自然框架。在司法审查中，法官从公共机构决定合法的先验概率出发，然后根据呈堂证据更新这一概率。
Bayes' theorem states:
贝叶斯定理表明：
P( H₁ | E ) = P( E | H₁ ) · P( H₁ ) / ( P( E | H₁ )P( H₁ ) + P( E | H₀ )P( H₀ ) )
Here, P(H₁) is the prior probability of unlawfulness, and P(H₁|E) is the posterior probability after considering the evidence E. The judicial review process can thus be modelled as a sequence of updates, with each piece of evidence refining the probability of unlawfulness.
这里，P(H₁) 是不合法的先验概率，P(H₁|E) 是在考虑证据 E 后的后验概率。司法审查过程因此可以建模为一个更新序列，每一份证据都会修正不合法的概率。
This mathematical perspective explains why some cases require stronger evidence than others: if the prior probability of a decision being lawful is very high (say, due to a presumption of regularity), then the likelihood ratio must be extremely large to shift the posterior probability over the decision threshold.
这种数学视角解释了为何某些案件需要比其他案件更强的证据：如果决定合法的先验概率非常高（例如，由于常规性推定），那么似然比必须极其大才能将后验概率推移过决策阈值。
5. Evaluating Evidence with Conditional Probability | 用条件概率评估证据
Judicial review frequently involves assessing the relevance and weight of multiple pieces of evidence. Each piece contributes conditionally – its impact depends on what else is already known. This is precisely the domain of conditional probability, a key topic in Edexcel Mathematics.
司法审查常常涉及评估多份证据的关联性和分量。每一份证据都是有条件地贡献——其影响取决于已经知道的其他情况。这正是条件概率的领域，也是爱德思数学中的一个关键主题。
Suppose two pieces of evidence, A and B, are presented. The probability that the decision is unlawful given both is P(H₁|A ∩ B). Using the multiplication rule and independence assumptions, we can compute:
假设提出了两份证据 A 和 B。给定两者后决定不合法的概率为 P(H₁|A ∩ B)。利用乘法法则和独立性假设，我们可以计算：
P( H₁ | A ∩ B ) ∝ P( A | H₁ ) · P( B | H₁ ) · P( H₁ )
Courts intuitively assess whether evidence is 'probative', meaning it has a positive likelihood ratio. A piece of evidence with LR > 1 supports the claimant; LR < 1 supports the public body. Mathematically skilled judges would check that each piece of evidence shifts the overall likelihood in a consistent direction.
法院会直觉地评估证据是否具有“证明力”，即其似然比是否大于1。似然比大于1的证据支持申请人；小于1的支持公共机构。具备数学技能的法官会检查每一份证据是否将总体似然朝着一致的方向推移。
6. The Prosecutor's Fallacy | 检察官谬误
A notorious error in legal reasoning – the prosecutor's fallacy – also contaminates judicial review if not carefully guarded against. This fallacy confuses P(E|H₀) with P(H₀|E). For example, concluding that because the probability of this evidence arising if the decision were lawful is very low, the decision must be unlawful.
法律推理中的一个臭名昭著的错误——检察官谬误——若不加仔细防范，也会污染司法审查。该谬误混淆了 P(E|H₀) 和 P(H₀|E)。例如，因为如果决定合法则该证据出现的概率非常低，就推断决定一定不合法。
In mathematical terms, equating a small p-value to the probability that the null hypothesis is true is a classic mistake. In judicial review, a judge might see a decision that appears irrational in a narrow procedural sense and quickly conclude it is unlawful, ignoring the prior probability and other context.
在数学术语中，将一个小的 p 值等同于零假设成立的概率是一个典型错误。在司法审查中，法官可能会看到一个在狭义程序意义上显得不合理的决定，就迅速断定其不合法，而忽视了先验概率和其他背景信息。
Edexcel students learn to interpret p-values correctly: a small p-value indicates the data are unlikely under H₀, but it does not directly give P(H₀ is false). The same discipline must apply in judicial review to avoid miscarriages of justice.
爱德思的学生学习如何正确解释 p 值：小的 p 值表明在 H₀ 下数据不太可能出现，但它并不直接给出 H₀ 为假的概率。司法审查中也必须运用同样的纪律，以避免司法不公。
7. Statistical Significance and Reasonable Doubt | 统计显著性与合理怀疑
In criminal law, the standard of proof is 'beyond reasonable doubt'. In judicial review, the standard is often the balance of probabilities or a heightened irrationality test. Both can be mapped to statistical significance thresholds. A balance of probabilities corresponds to a posterior probability exceeding 50%, while 'beyond reasonable doubt' might require a posterior above 90% or even 99%.
在刑法中，证明标准是“排除合理怀疑”。在司法审查中，标准通常是盖然性权衡或加码的不合理性检验。二者都可以映射到统计显著性阈值上。盖然性权衡对应的后验概率超过50%，而“排除合理怀疑”可能需要后验在90%甚至99%以上。
Choosing the right mathematical threshold involves decision theory: the cost of making a wrong decision. In judicial review of complex economic regulations, for instance, courts often defer to the regulator unless a decision is manifestly wrong, reflecting an asymmetric loss function that penalises type I errors more heavily.
选择合适的数学阈值涉及决策理论：做出错误决定的代价。例如，在对复杂经济监管进行司法审查时，除非决定明显错误，法院通常会尊重监管机构，这反映了一种对第一类错误惩罚更重的非对称损失函数。
8. Decision Theory and Utility in Judicial Review | 司法审查中的决策理论和效用
Decision theory combines probability with utility to recommend the action that maximises expected utility. In judicial review, the judge must decide whether to uphold or quash a decision. The expected utility of quashing is:
决策理论将概率与效用结合起来，以推荐最大化期望效用的行动。在司法审查中，法官必须决定是维持还是撤销一个决定。撤销的期望效用是：
U( quash ) = P( unlawful ) · U( correct quash ) + P( lawful ) · U( wrongful quash )
Upholding has a similar expression. The judge chooses the action with the higher expected utility. This explains why courts demand very strong evidence when the consequences of overturning a public body's decision are severe – the utility loss from a wrongful quash is high, so the decision threshold shifts.
维持决定也有类似的表达式。法官会选择期望效用较高的行动。这就解释了为什么当推翻公共机构决定的后果很严重时，法院会要求非常强的证据——错误撤销的效用损失很大，因此决策阈值会偏移。
This mathematical underpinning also aligns with the principle of proportionality in judicial review, which evaluates whether a decision's impact on individual rights is proportionate to the public interest aim – essentially a balancing of utilities.
这种数学基础也与司法审查中的比例原则相一致，该原则评估一项决定对个人权利的影响是否与公共利益目标成比例——本质上是一种效用平衡。
9. The Role of Prior Probability | 先验概率的作用
One controversial element in Bayesian modelling of judicial review is the choice of prior probability P(H₁). Some argue it should be based on historical data about how often public bodies act unlawfully. Others suggest a neutral prior of 0.5, while legal presumptions of regularity push the prior very low.
在司法审查的贝叶斯建模中，一个有争议的要素是先验概率 P(H₁) 的选择。有人主张应基于关于公共机构行为不合法频率的历史数据。另一些人建议使用 0.5 的中立先验，而合法性推定则会将先验推得很低。
In practice, judges rarely state numerical priors, but their behaviour can be retro-fitted to an implied prior. A court that rarely finds against the government implicitly uses a small prior for unlawfulness. Transparency in this implicit prior could lead to more consistent judicial review outcomes.
实践中，法官很少陈述数值先验，但他们的行为可以回溯拟合出一个隐含的先验。一个很少裁定政府败诉的法院，隐含地对不合法性使用了较小的先验。这种隐含先验的透明化可以促成更一致的司法审查结果。
10. Combining Multiple Evidences | 组合多项证据
A judicial review case rarely rests on a single piece of evidence. Instead, many strands are woven together. The mathematical way to combine independent evidences is through the sequential application of Bayes' theorem or by multiplying likelihood ratios if the evidences are conditionally independent given the hypothesis.
司法审查案件很少依赖于单一证据。相反，许多线索交织在一起。组合独立证据的数学方法是通过顺序应用贝叶斯定理，或者，如果各证据在给定假设下条件独立，则通过似然比相乘。
Given n pieces of evidence E₁, E₂, …, Eₙ, the posterior odds can be expressed as:
给定 n 份证据 E₁, E₂, …, Eₙ，后验几率可以表示为：
Posterior Odds = Prior Odds × LR₁ × LR₂ × … × LRₙ
This product form is powerful: even weak evidence can accumulate to a compelling case if the LRs are consistently greater than 1. Conversely, a single piece of strong exculpatory evidence (LR << 1) can outweigh many weakly incriminating ones.
这种乘积形式很强大：如果似然比始终大于1，即使是弱证据也能累积成一个有说服力的案件。相反，单个强有力的开脱证据（LR << 1）可以胜过许多微弱的入罪证据。
In judicial review, courts look for a 'cumulative effect' of multiple procedural irregularities. The mathematics of likelihood ratio products gives a rigorous basis for such reasoning, though independence assumptions must be checked carefully to avoid double-counting.
在司法审查中，法院会寻找多个程序违规的“累积效应”。似然比乘积的数学为这种推理提供了严格基础，但必须仔细检查独立性假设以避免重复计算。
11. Judicial Review as a Hypothesis Test | 作为假设检验的司法审查
We can formalise the judicial review process as a classical hypothesis test with the following structure:
我们可以将司法审查过程形式化为具有如下结构的经典假设检验：
- Null hypothesis H₀: The decision is lawful. / 零假设 H₀：决定合法。
- Alternative hypothesis H₁: The decision is unlawful. / 备择假设 H₁：决定不合法。
- Test statistic: A composite score derived from the evidence, reflecting the strength of the case. / 检验统计量：从证据中得出的一个综合得分，反映案件强度。
- Critical region: If the test statistic exceeds a critical value, H₀ is rejected; the decision is quashed. / 临界域：如果检验统计量超过临界值，则拒绝 H₀；决定被撤销。
The critical value is set by the standard of review. For irrationality, the critical value is positioned far into the tail of the distribution under H₀, making it difficult to reject lawfulness. This corresponds to a very low significance level, often less than 1%.
临界值由审查标准设定。对于非合理性标准，临界值被置于 H₀ 下分布的远端尾部，使得拒绝合法性变得困难。这对应于非常低的显著性水平，通常低于1%。
Edexcel students familiar with one-tailed and two-tailed tests will recognise that judicial review typically uses a one-tailed test: the court only rejects the decision if the evidence strongly points towards unlawfulness, not mere oddity.
熟悉单尾和双尾检验的爱德思学生会认识到，司法审查通常使用单尾检验：只有当证据强烈指向不合法性时，法院才会推翻决定，而不仅仅是看似奇怪。
12. Conclusion: Mathematics Ensuring Fairness | 结论：数学保障公平
The judicial review process, at its heart, is an exercise in structured reasoning under uncertainty – the very domain of A-Level Mathematics. By applying the concepts of probability, hypothesis testing, Bayesian updating, and decision theory, we gain a clearer, more consistent model of what courts are doing when they scrutinise public decision-making.
司法审查过程，就其核心而言，是在不确定条件下进行结构化推理的一种实践——这正是 A-Level 数学的领域。通过应用概率、假设检验、贝叶斯更新和决策理论等概念，我们获得了一个更清晰、更一致的模型，用以理解法院在审查公共决策时实际上在做什么。
While judges rarely use mathematical notation, they constantly weigh probabilities, evaluate likelihood ratios, and set thresholds for error. Understanding these mathematical underpinnings not only deepens your appreciation of the rule of law but also demonstrates the real-world power of the statistics and probability you study in the classroom.
尽管法官很少使用数学符号，但他们不断在权衡概率、评估似然比和设定错误阈值。理解这些数学基础不仅能加深你对法治的理解，也展示了你在课堂上学到的统计和概率知识在现实世界中的力量。
Whether you are aiming for a top grade in Edexcel Mathematics or curious about the intersection of law and logic, seeing judicial review through a mathematical lens transforms an abstract process into a concrete, analysable system.
无论你是以爱德思数学取得高分作为目标，还是对法律与逻辑的交汇感到好奇，通过数学镜头看待司法审查，都会将这一抽象过程转化为一个具体、可分析的系统。
Published by TutorHao | Mathematics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图