Conditional Probability | 条件概率

📚 Conditional Probability | 条件概率

Conditional probability is one of the most powerful concepts in the IB Mathematics syllabus, forming the backbone of many real-world applications from medical testing to risk assessment. It allows us to revise the likelihood of an event occurring after we have learned that another event has taken place. Understanding this topic not only helps you solve intricate probability problems but also sharpens your logical reasoning skills for unseen scenarios in Paper 2 and the Internal Assessment.

条件概率是 IB 数学大纲中最强大的概念之一，构成了从医学检测到风险评估等众多现实应用的基石。它使我们能够在得知另一事件发生后，重新修正某个事件发生的可能性。理解这一话题不仅有助于解决复杂的概率问题，还能为试卷二和内部评估中出现的陌生情境提供逻辑推理的锐利工具。

1. What Is Conditional Probability? | 什么是条件概率？

Suppose you draw a card from a standard deck of 52 cards. The probability of drawing a King is 4/52. But if your friend glances at the card and tells you it is a face card, suddenly the sample space shrinks to 12 cards, and the probability becomes 4/12. This updated chance is a conditional probability.

假设你从一副标准的 52 张扑克牌中抽一张牌，抽到 K 的概率是 4/52。但如果你的朋友瞥了一眼牌然后告诉你这是一张人头牌，样本空间立刻缩小到 12 张牌，概率就变成了 4/12。这个更新后的概率就是条件概率。

In formal terms, conditional probability is the probability of an event A occurring given that event B has already occurred. It is written as P(A | B) and read as “the probability of A given B”. The vertical bar is not a division symbol but a conditioning sign that signals we restrict our attention to the sub-universe where B is known to be true.

从形式上讲，条件概率是在事件 B 已经发生的条件下，事件 A 发生的概率。记作 P(A | B)，读作“给定 B 时 A 的概率”。竖线不是除号，而是一个条件符号，表明我们将关注点限制在已知 B 成立的那个子空间内。

2. Notation and Formal Definition | 符号与形式化定义

The defining formula for conditional probability is:

条件概率的定义公式为：

P(A | B) = P(A ∩ B) / P(B), provided P(B) > 0

P(A | B) = P(A ∩ B) / P(B), 当 P(B) > 0 时

Here, A ∩ B represents the intersection of events A and B, i.e., the probability that both occur. The denominator P(B) serves as a normalising factor, ensuring that all probabilities within the new sample space B sum to 1. If P(B) = 0, then P(A | B) is undefined, as it would require division by zero.

这里，A ∩ B 表示事件 A 与 B 的交集，即两者同时发生的概率。分母 P(B) 充当归一化因子，确保在 B 这个新样本空间内的所有概率之和为 1。如果 P(B) = 0，则 P(A | B) 无定义，因为意味着除以零。

The complement rule also adapts naturally under conditioning: P(A’ | B) = 1 – P(A | B). This can save time when directly calculating the complement is easier. Keep in mind that P(A | B) and P(B | A) are generally different numbers; confusing them is one of the most common errors in probability.

补集规则在条件背景下也自然成立：P(A’ | B) = 1 – P(A | B)。当直接计算补集更容易时，这可以节省时间。务必记住 P(A | B) 和 P(B | A) 通常是不同的数值；混淆它们是概率中最常见的错误之一。

3. The Multiplication Rule | 乘法法则

Rearranging the definition of conditional probability yields the multiplication rule, which allows us to find the joint probability P(A ∩ B) when a conditional probability is known:

将条件概率的定义变形就得到了乘法法则，它使得我们在已知条件概率时可以求得联合概率 P(A ∩ B)：

P(A ∩ B) = P(A | B) × P(B) = P(B | A) × P(A)

The two equivalent forms give us flexibility: if information is presented as P(A | B) and P(B), the first form is natural; if we know P(B | A) and P(A), the second form is preferred. This rule generalises to more than two events via a chain of conditionals, e.g., P(A ∩ B ∩ C) = P(A) × P(B | A) × P(C | A ∩ B), which is particularly useful in tree diagram problems.

这两种等价形式给了我们灵活性：如果信息以 P(A | B) 和 P(B) 的形式呈现，就用第一种形式；如果我们知道 P(B | A) 和 P(A)，则采用第二种形式。这个法则可以通过一连串条件推广到多于两个事件，例如 P(A ∩ B ∩ C) = P(A) × P(B | A) × P(C | A ∩ B)，这在树形图问题中特别有用。

In many IB exam questions, the multiplication rule is the hidden engine behind seemingly complicated scenarios. You may be asked to find the probability that two events occur in sequence, such as selecting two defective items without replacement. Sketching a tree and applying the multiplication rule along the branches keeps your working transparent.

在许多 IB 考试题中，乘法法则是看似复杂情境背后的隐形引擎。你可能会被要求找出两个事件依序发生的概率，比如无放回地抽出两个次品。画出树形图并沿着分支应用乘法法则可以让你的解题过程清晰透明。

4. Interplay with Venn Diagrams | 与维恩图的相互作用

Venn diagrams provide a vivid geometric interpretation of conditional probability. When we condition on B, we effectively zoom into the circle representing B and ignore the rest of the universe. The fraction P(A ∩ B) / P(B) corresponds to the area of the overlap relative to the entire area of B.

维恩图为条件概率提供了形象的几何解释。当我们以 B 为条件时，实际上就是把视野缩放到代表 B 的圆圈内，忽略掉其余部分。分数 P(A ∩ B) / P(B) 就对应于重叠部分面积占整个 B 面积的比例。

To solve a problem with a Venn diagram, start by labeling the probabilities of each disjoint region (A only, B only, both A and B, neither). The values must sum to 1. Once the diagram is complete, P(A | B) can be read off as (probability in the intersection) ÷ (probability in the B circle). This method is excellent for two-event systems and serves as a visual check.

要用维恩图解解题，首先标出各个互斥区域的概率（只有 A、只有 B、A 与 B 交集、两者皆非）。这些值之和必须为 1。一旦图完成，P(A | B) 就可以直接从图中读取，即（交集区域内的概率）÷（B 圆内的概率）。这种方法对于两事件系统非常出色，并可作为视觉检验。

Be careful with notation: some students mistakenly write P(A ∩ B) as P(A) × P(B) even when the events are not independent. Venn diagrams remind you that the intersection is a raw probability that must be measured or given, not automatically the product of marginal probabilities.

注意符号：有些学生即便事件不独立，也错误地将 P(A ∩ B) 写成 P(A) × P(B)。维恩图提醒你，交集的概率必须是实测或给定的原始概率，不能自动等于边缘概率的乘积。

5. Tree Diagrams and Conditional Probabilities | 树形图与条件概率

Tree diagrams are the IB candidate’s best friend when dealing with multi-stage experiments. Each set of branches stems from a node and carries the conditional probability of the next outcome given everything that has occurred up to that point. The probabilities on the first‑level branches are ordinary (unconditional) probabilities; thereafter every branch is a conditional probability.

在处理多阶段试验时，树形图是 IB 考生最好的帮手。每一组分支都从一个节点发出，并附带了在已发生全部事件条件下下一结果的概率。第一层的分支概率是普通（无条件）概率；自此往后的每一条分支都是条件概率。

The key property is that the probabilities of all branches emanating from a given node must sum to 1, because they represent all possible outcomes from that state. The end of each path corresponds to a sequence of events; the probability of that whole path is the product of the branch probabilities along it, as prescribed by the multiplication rule.

关键性质是，从同一节点发出的所有分支的概率之和必须为 1，因为它们代表了从该状态可能出发的全部结果。每条路径的末端对应一个事件序列；整条路径的概率就是沿途分支概率的乘积，正如乘法法则所规定的那样。

Often you need to reverse the condition, i.e., given a final outcome, what is the probability that a particular earlier branch was taken? This is where the tree diagram feeds into Bayes’ type reasoning. Extract every path that leads to the final outcome, sum their probabilities, and then take the ratio of the desired path to that total.

你常常需要逆转条件，也就是说，给定最终结果，求某条特定早期分支被选中的概率是多少？这正是树形图通向贝叶斯类推理之处。提取所有通往该最终结果的路径，将它们概率相加，然后取所需路径与总和的比值。

6. Two‑Way Tables and Conditional Probability | 双向表与条件概率

A two‑way table (also called a contingency table) organises data by two categorical variables. The interior cells contain joint frequencies or joint probabilities. The marginal totals appear in the last row and last column. This structure makes computing conditional probabilities entirely mechanical: to find P(A | B), locate the cell where A and B intersect, then divide by the row or column total corresponding to B.

双向表（也称列联表）通过两个分类变量来组织数据。内部单元格包含联合频数或联合概率。边际总和出现在最后一行和最后一列。这种结构使得计算条件概率完全机械化：要计算 P(A | B)，先找到 A 与 B 交叉处的单元格，然后除以与 B 对应的行或列的总和。

For instance, a table of gender versus preferred subject might show 30 females who like Mathematics out of 80 females. If we choose a student at random and learn that she is female, the probability she likes Mathematics is 30/80. The denominator changes depending on which variable is conditioned upon, so read the question carefully.

例如，一份性别与偏好科目的表格可能显示 80 名女生中有 30 名喜欢数学。如果我们随机选择一名学生并得知她是女性，则她喜欢数学的概率为 30/80。分母会根据条件设定在哪个变量上而改变，因此请仔细读题。

IB problems may require you to convert frequencies to probabilities by dividing by the grand total first, or to complete missing entries using additive reasoning. Always double‑check that row and column totals sum to the grand total, otherwise conditional probabilities will be corrupted.

IB 问题可能要求你先将频数除以总计以转换为概率，或利用加法推理补全缺失项。务必反复检查行总和与列总和是否等于总计，否则条件概率会出错。

7. Independence and Its Condition | 独立事件及其条件

Two events A and B are independent if the occurrence of one does not affect the probability of the other. The formal condition for independence can be expressed in three equivalent ways:

如果两个事件 A 与 B 中一个的发生不影响另一个的概率，则它们是独立的。独立性的形式化条件可以用三种等价方式表达：

P(A | B) = P(A) (and symmetrically P(B | A) = P(B))
P(A ∩ B) = P(A) × P(B)
P(A | B) = P(A | B’) (rarely used but logically equivalent)

P(A | B) = P(A)（且对称地 P(B | A) = P(B)）
P(A ∩ B) = P(A) × P(B)
P(A | B) = P(A | B’)（较少使用但逻辑等价）

The most practical test in IB is to compare P(A ∩ B) with P(A) × P(B). If the two quantities are equal, the events are independent; otherwise they are dependent. Be careful: mutual exclusivity and independence are completely different ideas. Mutually exclusive events with non‑zero probabilities cannot be independent because P(A ∩ B) = 0 while P(A) × P(B) > 0.

IB 中最实用的检验是比较 P(A ∩ B) 与 P(A) × P(B)。若两者相等，则事件独立；否则不独立。请注意：互斥与独立是完全不同的概念。非零概率的互斥事件不可能独立，因为 P(A ∩ B) = 0 而 P(A) × P(B) > 0。

Conditional probability also lets us define conditional independence. Events A and B might be dependent overall yet become independent when conditioned on a third event C. This idea appears in more advanced IB HL questions and requires careful symbolic reasoning.

条件概率还允许我们定义条件独立。事件 A 与 B 可能在总体上相关，但当以第三个事件 C 为条件时就变得独立了。这一思想出现在更高级的 IB HL 问题中，需要仔细的符号推理。

8. The Law of Total Probability | 全概率公式

Often we know conditional probabilities across a partition of the sample space and need an unconditional probability. The law of total probability provides the bridge. If events B₁, B₂, …, Bₙ form a partition of the sample space (they are mutually exclusive and exhaustive), then for any event A:

我们常常已知样本空间一个划分上的条件概率，而需要一个无条件概率。全概率公式提供了这道桥梁。若事件 B₁, B₂, …, Bₙ 构成样本空间的一个划分（它们互斥且穷尽），则对任意事件 A：

P(A) = Σ P(A | Bᵢ) × P(Bᵢ)

This formula says: the total probability of A is the weighted average of the conditional probabilities, with weights given by the probabilities of the partitioning events. In a tree diagram, this corresponds to summing the probabilities of all paths that end at A.

该公式表示：A 的全概率是各条件概率的加权平均，权重由划分事件的概率给出。在树形图中，这对应于将所有以 A 为终点的路径概率相加。

A classic IB question involves a factory with three machines producing components at different rates and with different defect rates. The law of total probability gives the overall defect rate of a randomly chosen component. The calculation is straightforward once the partition is clearly identified.

经典的 IB 问题涉及一家工厂有三台机器，以不同速率生产零件且次品率不同。全概率公式可以给出随机抽取一个零件的总体次品率。一旦清晰识别出划分，计算就非常直接了。

9. Bayes’ Theorem | 贝叶斯定理

Bayes’ Theorem is the natural consequence of combining the definition of conditional probability with the law of total probability. It allows us to “invert” conditional probabilities: if we know P(A | Bᵢ) and the prior probabilities P(Bᵢ), we can find the posterior probability P(Bₖ | A). The formula is:

贝叶斯定理是将条件概率的定义与全概率公式相结合的自然结果。它使我们能够“反转”条件概率：如果我们知道 P(A | Bᵢ) 以及先验概率 P(Bᵢ)，就可以求出后验概率 P(Bₖ | A)。公式为：

P(Bₖ | A) = P(A | Bₖ) × P(Bₖ) / [ Σ P(A | Bᵢ) × P(Bᵢ) ]

IB examiners often set questions that require Bayesian reasoning but do not necessarily ask for the formula by quote. The safest approach is to draw a tree diagram, compute all path probabilities, identify the relevant paths, and form the fraction (probability of path through Bₖ to A) ÷ (total probability of A). This method reduces the risk of substitution errors.

IB 考官常常设置需要贝叶斯推理的问题，但不一定要求引述公式。最稳妥的方法是画出树形图，计算所有路径概率，识别相关路径，并构建分式（途经 Bₖ 到达 A 的路径概率）÷（A 的总概率）。这种方法降低了代入出错的风险。

Interpretation is crucial. The posterior probability P(Bₖ | A) blends prior belief with new evidence. In a medical testing context, a positive test result from a highly specific test still might not imply a high probability of disease if the disease is extremely rare. This counter‑intuitive result is a hallmark of Bayesian thinking and frequently appears in IB analysis questions.

解释至关重要。后验概率 P(Bₖ | A) 将先验信念与新证据融合。在医学检验情境中，即便使用了高特异性的测试，如果疾病本身极为罕见，阳性结果仍可能不意味着患病的概率很高。这个反直觉的结果是贝叶斯思维的标志，并频繁出现在 IB 分析题中。

10. Common Pitfalls and IB Exam Tips | 常见陷阱与 IB 考试技巧

One common trap is confusing P(A | B) with P(B | A). Always check which event is the condition. A quick mental substitution can help: “Probability of rain given clouds” is not the same as “Probability of clouds given rain”.

一个常见陷阱是将 P(A | B) 与 P(B | A) 混淆。始终要核查哪个事件是条件。快速的心理代入会有所帮助：“有云时下雨的概率”与“下雨时有云的概率”是不同的。

Another pitfall is assuming events are independent without verification. Unless the problem explicitly states independence or the physical setup implies it (e.g., replacement in selection), you must compute using given probabilities. The same caution applies to assuming mutual exclusivity.

另一个陷阱是未经验证就假设事件独立。除非题目明确陈述独立性，或实际设置隐含了独立性（如抽取时放回），否则你必须使用给定的概率进行计算。同样的警惕也适用于假设互斥性。

When using formulas, always check that P(B) > 0 before writing P(A | B). In many IB mark schemes, simply noting this condition is a mark‑worthy step. For the law of total probability and Bayes’ Theorem, ensure your partition is exhaustive and mutually exclusive; otherwise the formulas break down.

在使用公式时，在写下 P(A | B) 之前务必检查 P(B) > 0。在许多 IB 评分方案中，仅注明这一条件就是值得给分的一步。对于全概率公式和贝叶斯定理，要确保你的划分是穷尽且互斥的；否则公式将失效。

Finally, practise interpreting the word “given” in prose questions and translating it into the notation P( … | … ). When the question asks “Find the probability that a randomly chosen student is male, given that they study Physics”, write P(Male | Physics) and then decide whether to use a table, tree, or formula. Clarity of notation earns marks and prevents confusion.

最后，练习如何解读文字题中的“给定”一词，并将其转化为符号 P( … | … )。当题目要求“求随机选出一名学习物理的学生是男性的概率”时，写成 P(男 | 物理)，然后决定使用表格、树形图还是公式。清晰的符号能为你赢得分数并防止混乱。

Published by TutorHao | Mathematics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Conditional Probability | 条件概率