IB数学AI Statistics 正态分布 二项分布 检验 | IB Math AI Statistics: Normal Distribution, Binomial Distribution, Hypothesis Testing

IB数学AI Statistics 正态分布 二项分布 检验

SL 水平

描述性统计, 概率基础, 正态分布, 二项分布, 卡方检验

HL 水平

t检验, 置信区间, 泊松分布, 误差类型, 假设检验深入

Statistics is the beating heart of IB Mathematics: Applications and Interpretation (AI). Whether you are an SL student navigating descriptive statistics and basic probability, or an HL candidate wrestling with hypothesis testing and confidence intervals, a solid grasp of statistical reasoning separates a grade 5 from a grade 7. 统计是IB数学AI的核心模块。无论你是SL学生探索描述性统计与基础概率，还是HL考生迎战假设检验与置信区间，扎实的统计推理能力是5分与7分之间的分水岭。 This article walks you through every major topic in the IB Math AI statistics syllabus, from the normal and binomial distributions to chi-square and t-tests, with worked examples and exam strategy tips. 本文将带你系统梳理IB数学AI统计大纲中的每一个核心主题，从正态分布与二项分布到卡方检验与t检验，配以例题解析与应试策略。

🛍️ 微信小店好物推荐

fx-991CN X 中文版计算器-老王推荐

¥168.00 🛒 微信小店购买 ### 描述性统计 Descriptive Statistics 描述性统计是所有数据分析的起点。IB数学AI要求学生熟练掌握数据的集中趋势和离散程度度量。Descriptive statistics is the starting point of all data analysis. IB Math AI requires students to master measures of central tendency and dispersion. **Central tendency 集中趋势**: Mean (平均数), median (中位数), and mode (众数). The mean is sensitive to outliers; the median is robust. When data is skewed, report the median alongside the IQR. 均值易受离群值影响，中位数具有稳健性。当数据偏斜时，应报告中位数与四分位距。 **Dispersion 离散程度**: Range (极差), interquartile range IQR (四分位距), variance (方差), and standard deviation (标准差). For grouped data, use the midpoint of each interval to estimate the mean and standard deviation. 对于分组数据，使用每个区间的中点来估计均值和标准差。 **GDC shortcuts**: On the TI-Nspire, enter data into a Lists and Spreadsheets page, then use `Menu > Statistics > Stat Calculations > One-Variable Statistics`. On the Casio fx-991CN X, use MODE 2 (STAT) and select 1-VAR. 在TI-Nspire上，将数据输入Lists and Spreadsheets页面，使用菜单统计单变量统计。Casio用户使用MODE 2进入统计模式。 ### 概率基础 Probability Fundamentals Probability in IB Math AI covers conditional probability, Venn diagrams, tree diagrams, and probability distributions. IB数学AI的概率涵盖条件概率、韦恩图、树状图和概率分布。 **Key formulas**: - P(A ∪ B) = P(A) + P(B) - P(A ∩ B) - P(A | B) = P(A ∩ B) / P(B), provided P(B) > 0 - For independent events: P(A ∩ B) = P(A) × P(B) The most common exam trap is confusing P(A | B) with P(B | A). Remember: the condition comes after the vertical bar. Bayes' theorem formalizes this relationship: P(A | B) = P(B | A) × P(A) / P(B). 最常见的考试陷阱是将P(A | B)与P(B | A)混淆。记住：竖线后面的是条件。贝叶斯定理形式化了这个关系。 **Tree diagrams 树状图**: Always multiply along branches and add across branches. Label every branch with its probability before doing any calculation. A two-stage tree with 3 branches at each stage yields 9 terminal outcomes. 始终沿分支相乘，跨分支相加。在计算前标注每条分支的概率。 ### 正态分布 Normal Distribution The normal distribution is the most important continuous probability distribution in the IB syllabus. Its bell-shaped curve is symmetric about the mean μ, with approximately 68% of data within ±1σ, 95% within ±2σ, and 99.7% within ±3σ. 正态分布是IB大纲中最重要的连续概率分布。钟形曲线关于均值μ对称，约68%的数据落在±1σ内，95%在±2σ内，99.7%在±3σ内。 **Standardization and z-scores 标准化与z分数**: To compare values from different normal distributions, convert raw scores to z-scores: z = (x - μ) / σ. A z-score tells you how many standard deviations a value is from the mean. Negative z-scores indicate values below the mean. 要将不同正态分布的数值进行比较，将原始分数转换为z分数。负z分数表示数值低于均值。 **GDC calculation 计算器操作**: For normal probability calculations, use `normCdf(lower, upper, μ, σ)` on the TI-Nspire. To find a value given a probability (inverse normal), use `invNorm(area, μ, σ)`. On the Casio fx-991CN X: MODE 3 (DIST), then select Normal CD for cumulative probability or Inverse Normal for quantile. 在TI-Nspire上用normCdf计算正态概率，用invNorm反查数值。Casio上用MODE 3进入分布模式。 68%

μ ± 1σ

95%

μ ± 2σ

99.7%

μ ± 3σ

**Worked example 例题**: The heights of IB students are normally distributed with mean 168 cm and standard deviation 8 cm. What proportion of students are taller than 180 cm? z = (180 - 168) / 8 = 1.5. P(Z > 1.5) = 1 - P(Z < 1.5) = 1 - 0.9332 = 0.0668. Approximately 6.68% of students are taller than 180 cm. IB学生身高服从正态分布，均值168cm，标准差8cm。身高超过180cm的学生比例是多少？z = 1.5，P(Z > 1.5) = 0.0668，约6.68%。 ### 二项分布 Binomial Distribution The binomial distribution models the number of successes in n independent trials, each with probability p of success. Notation: X ~ B(n, p). 二项分布建模n次独立试验中成功的次数，每次成功概率为p。记作X ~ B(n, p)。 **Conditions 条件**: 1. Fixed number of trials n (固定试验次数) 2. Each trial is independent (每次试验独立) 3. Two possible outcomes per trial: success or failure (每次试验只有成功或失败) 4. Constant probability p for each trial (每次试验概率p恒定) **Formula**: P(X = k) = C(n, k) × p^k × (1-p)^(n-k), where C(n, k) = n! / [k!(n-k)!]. **Mean and variance 均值与方差**: E(X) = np, Var(X) = np(1-p). These are the most frequently tested properties. 这是最常考的性质。For large n, when np > 5 and n(1-p) > 5, the binomial distribution can be approximated by a normal distribution N(np, np(1-p)). 当np > 5且n(1-p) > 5时，二项分布可用正态分布N(np, np(1-p))近似。 **GDC usage 计算器操作**: On the TI-Nspire, use `Menu > Probability > Distributions > Binomial Pdf` for P(X = k) and `Binomial Cdf` for P(X ≤ k). Remember: Cdf gives cumulative "≤", so for P(X ≥ k), compute 1 - binomCdf(n, p, k-1). TI-Nspire上用Binomial Pdf计算恰好k次成功的概率，Binomial Cdf计算累积概率。 **Worked example 例题**: A multiple-choice quiz has 10 questions, each with 4 options. A student guesses all answers. Find the probability of getting exactly 3 correct. X ~ B(10, 0.25). P(X = 3) = C(10,3) × (0.25)^3 × (0.75)^7 = 120 × 0.015625 × 0.13348 = 0.2503. 学生全部猜题，恰好答对3题的概率为0.2503，约25%。 ### 假设检验 Hypothesis Testing Hypothesis testing is the crown jewel of IB Math AI statistics. It provides a formal framework for making decisions based on data. 假设检验是IB数学AI统计的皇冠明珠，为基于数据的决策提供了形式化框架。 **Core concepts 核心概念**: - **H₀ (null hypothesis 原假设)**: The default position, usually "no effect" or "no difference" - **H₁ (alternative hypothesis 备择假设)**: What you are trying to prove - **p-value (p值)**: The probability of observing the sample result (or more extreme) if H₀ is true - **Significance level α (显著性水平)**: The threshold for rejecting H₀, typically 5% (0.05) in IB exams **Decision rule 决策规则**: If p-value < α, reject H₀. If p-value ≥ α, do not reject H₀. Critical: "do not reject H₀" is NOT the same as "accept H₀." You never "accept" the null hypothesis -- you only fail to find sufficient evidence against it. 如果p值 < α，拒绝H₀。如果p值 ≥ α，不拒绝H₀。关键："不拒绝H₀"不等于"接受H₀"。你永远不能"接受"原假设，只是未能找到足够证据拒绝它。 **Type I and Type II errors 第一类与第二类错误**: - Type I error (第一类错误): Rejecting H₀ when it is actually true. Probability = α. - Type II error (第二类错误): Failing to reject H₀ when it is actually false. Probability = β. - Power (检验功效): 1 - β, the probability of correctly rejecting a false H₀. Type I Error (α)

False positive. Reject true H₀. 假阳性，拒绝了真实的原假设。

Type II Error (β)

False negative. Miss a real effect. 假阴性，错过了真实效应。

### 卡方检验 Chi-Square Tests The chi-square (χ²) test is used for categorical data. IB covers two types: goodness of fit and test for independence. 卡方检验用于分类数据。IB涵盖两种类型：拟合优度检验和独立性检验。 **Chi-square goodness of fit 拟合优度检验**: Tests whether observed frequencies match expected frequencies from a theoretical distribution. H₀: the data follows the specified distribution. 检验观察频率是否符合理论分布的期望频率。 Formula: χ² = Σ [(O_i - E_i)² / E_i], where O_i is the observed frequency and E_i is the expected frequency. Degrees of freedom (df) = number of categories - 1. For the test of independence in a contingency table, df = (rows - 1) × (columns - 1). 对于列联表中的独立性检验，自由度 = (行数-1) × (列数-1)。 **GDC procedure 计算器步骤**: On the TI-Nspire, enter observed frequencies into a matrix, then use `Menu > Statistics > Stat Tests > χ² GOF` for goodness of fit, or `χ² 2-way` for independence. The calculator returns χ², p-value, and df. CASIO fx-991CN X does not have built-in χ² test functionality; use the formula manually. TI-Nspire上用矩阵输入观察频率，使用χ² GOF进行拟合优度检验或χ² 2-way进行独立性检验。 **Key exam tip 考试关键**: Always check that all expected frequencies E_i ≥ 5. If any expected frequency is less than 5, combine categories. The IB mark scheme explicitly penalizes ignoring this condition. 务必检查所有期望频率E_i ≥ 5。如果任何期望频率小于5，合并类别。IB评分标准明确惩罚忽视此条件的行为。 ### t检验 t-Tests (HL Only) The t-test is an HL-only topic that extends hypothesis testing to situations where the population standard deviation σ is unknown. t检验是仅HL涉及的主题，将假设检验扩展到总体标准差σ未知的情况。 **One-sample t-test 单样本t检验**: Tests whether a population mean equals a hypothesized value. t = (x̄ - μ₀) / (s / √n), where s is the sample standard deviation. df = n - 1. 检验总体均值是否等于假设值。t = (x̄ - μ₀) / (s / √n)，df = n - 1。 **Two-sample t-test 双样本t检验**: Compares the means of two independent groups. The IB typically uses the pooled (equal variance) version: t = (x̄₁ - x̄₂) / [s_p × √(1/n₁ + 1/n₂)], where s_p² = [(n₁-1)s₁² + (n₂-1)s₂²] / (n₁ + n₂ - 2). 比较两个独立组的均值。IB通常使用合并方差版本。 **GDC shortcut**: On the TI-Nspire, use `Menu > Statistics > Stat Tests > t Test` and select data or stats input. The calculator produces t-statistic, p-value, and confidence interval automatically. TI-Nspire上用t Test菜单直接得出t统计量、p值和置信区间。 **Worked example 例题 (two-sample t-test)**: A researcher tests two teaching methods. Group A (n=25): mean score 72, SD 8. Group B (n=25): mean score 78, SD 7. Test whether the difference is significant at α = 0.05. H₀: μ_A = μ_B, H₁: μ_A ≠ μ_B. s_p² = (24×64 + 24×49) / 48 = 56.5, s_p ≈ 7.52. t = (72-78) / (7.52 × √(2/25)) = -6 / 2.127 = -2.82. For df = 48, p ≈ 0.007. Since p < 0.05, reject H₀. There is significant evidence of a difference. 研究者测试两种教学方法。A组平均72分，B组平均78分。t = -2.82, p ≈ 0.007 < 0.05，拒绝H₀，存在显著差异。 ### 置信区间 Confidence Intervals (HL Only) A confidence interval provides a range of plausible values for a population parameter. 置信区间为总体参数提供了一个合理的取值范围。 **For a population mean (σ unknown)**: x̄ ± t* × (s / √n), where t* is the critical t-value for the desired confidence level with n-1 degrees of freedom. 对于总体均值（σ未知）：x̄ ± t* × (s / √n)，t*是所需置信水平下自由度为n-1的临界t值。 **Interpretation 解读**: A 95% confidence interval means: if we repeated the sampling process many times, 95% of the constructed intervals would contain the true population mean. It does NOT mean there is a 95% probability that the population mean lies in this particular interval. The mean is fixed; the interval varies. 95%置信区间的含义是：如果我们多次重复抽样过程，95%构造的区间将包含真实总体均值。这并不意味着有95%的概率总体均值在这个特定区间内。 **GDC on TI-Nspire**: Use `Menu > Statistics > Confidence Intervals > t Interval`. On the Casio fx-991CN X, use `INTR` mode then select `t`. 在TI-Nspire上用t Interval，Casio上用INTR模式选择t。 ### 常见陷阱与应试技巧 Common Pitfalls and Exam Tips ❌ 常见错误 Common Mistake

混淆 population SD (σ) 与 sample SD (s)。当数据来自样本时使用s，来自总体时使用σ。注意GDC上的σx与sx区分。

✅ 得分技巧 Scoring Tip

写出完整的假设：H₀和H₁，显著性水平α，检验统计量，p值，以及上下文中的结论。即使计算正确，缺少结论也会丢分。

1. **"Accept H₀" trap**: Never write "accept H₀." Always write "do not reject H₀" or "there is insufficient evidence to reject H₀." 永远不要写"接受H₀"，始终写"不拒绝H₀"或"没有足够证据拒绝H₀"。 2. **p-value vs α comparison**: Compare p-value to α, NOT to 0.5 or 0.1 unless those are your α. The IB sometimes uses α = 0.01 in addition to the standard 0.05. 将p值与α比较，不要与0.5或0.1比较，除非它们是设定的α。 3. **One-tailed vs two-tailed 单尾与双尾**: Check the wording of H₁. "Greater than" or "less than" → one-tailed. "Different from" or "not equal to" → two-tailed. For two-tailed tests, remember to double the p-value from the GDC output if it only gives a one-tailed result. 检查H₁的措辞。"大于"或"小于"→单尾；"不同于"或"不等于"→双尾。 4. **Degrees of freedom**: χ² GOF: df = categories - 1. χ² independence: df = (r-1)(c-1). One-sample t: df = n-1. Two-sample t (pooled): df = n₁ + n₂ - 2. 自由度：χ²拟合优度df = 类别数-1；独立性df = (行-1)(列-1)；单样本t: n-1；双样本t: n₁+n₂-2。 5. **Outliers 离群值**: For normally distributed data, values beyond μ ± 3σ are potential outliers. For box plots, values beyond Q1 - 1.5×IQR or Q3 + 1.5×IQR are flagged. IB questions often ask you to discuss the effect of removing an outlier on the mean and standard deviation. 对于正态分布数据，超出μ ± 3σ的值可能是离群值。IB题目常要求讨论移除离群值对均值和标准差的影响。 ### 考试策略与复习清单 Exam Strategy and Revision Checklist 📊 描述性统计

mean, median, SD, IQR, box plot

🎲 概率

Venn, tree, Bayes, conditional

🔔 正态分布

z-scores, normCdf, invNorm

🎯 二项分布

B(n,p), binomPdf, binomCdf

🧪 假设检验

H₀/H₁, p-value, α, Type I/II

📐 卡方检验

χ² GOF, independence, df

📏 t检验 (HL)

one-sample, two-sample, CI

🧮 GDC操作

TI-Nspire, Casio shortcuts

**Paper 1 tips 试卷一技巧**: Paper 1 is calculator-free for SL and allows GDC for HL. For SL, practice computing binomial coefficients and normal probabilities by hand using the formula sheet. For HL Paper 1 (Section B), expect extended-response statistics questions that combine multiple topics -- for example, a binomial distribution question that leads into a hypothesis test. 试卷一SL不用计算器，HL可以用。SL要练习手算二项系数和正态概率。HL试卷一B部分常见综合题。 **Paper 2 tips 试卷二技巧**: Paper 2 is GDC-active for both SL and HL. Master your GDC shortcuts. Know how to enter data, produce summary statistics, run hypothesis tests, and generate confidence intervals without hunting through menus. Time saved on calculations is time earned for interpretation and conclusion writing. 试卷二SL和HL均可用计算器。熟练掌握GDC快捷操作。节省计算时间就是赢得解读和结论写作时间。 **IA connections IA关联**: Statistics is the most popular choice for the IB Math AI Internal Assessment. Possible IA topics include: comparing exam scores between two groups (two-sample t-test), analyzing whether a die is fair (χ² GOF), modeling traffic arrival times (Poisson distribution), or correlating study hours with grades (regression analysis). Remember that your IA must include a clear rationale, data collection methodology, statistical analysis, and critical evaluation of limitations. 统计是IB数学AI内部评估中最受欢迎的选择。IA必须包含清晰的依据、数据收集方法、统计分析和局限性评估。 ### 总结与建议 Conclusion and Final Tips IB Math AI Statistics rewards systematic thinking over memorization. The key to a grade 7 is not knowing more formulas -- you have a formula booklet -- but knowing when and why to apply each test. Ask yourself for every question: What type of data do I have? Is it quantitative or categorical? Am I describing, comparing, or predicting? Every statistical method has its place; the exam tests your judgment in choosing the right one. IB数学AI统计奖励系统性思维而非死记硬背。7分的关键不是记住更多公式（你有公式手册），而是知道何时以及为何应用每种检验。每道题问自己：我有什么类型的数据？定量还是分类？我在描述、比较还是预测？每种统计方法都有其用武之地，考试考的是你选择正确方法的分辨力。 Practice with past papers, master your GDC, and always write a conclusion in context. Confidence comes from repetition. 用历年真题练习，精通GDC操作，始终在上下文中写出结论。信心来自重复。

🎓 Need one-on-one tutoring? 需要一对一辅导?

📱 16621398022 同微信

Follow tutorhao on WeChat for more learning resources 关注公众号获取更多学习资源

更多咨询请联系16621398022（同微信）