Statistical Models and Their Applications | 统计模型及其应用

📚 Statistical Models and Their Applications | 统计模型及其应用

Statistical models are mathematical frameworks that help us understand, describe, and make predictions from data. In A-Level mathematics, mastering these models is essential for interpreting real-world phenomena and solving examination problems effectively.

统计模型是帮助我们理解、描述数据并从中进行预测的数学框架。在A-Level数学中，掌握这些模型对于解释现实世界现象和有效解决考试问题至关重要。

1. Types of Statistical Models | 统计模型的类型

Statistical models are broadly classified into two categories: descriptive models and inferential models. Descriptive models summarise data using measures such as mean, median, and standard deviation. Inferential models use probability theory to draw conclusions about a population from a sample.

统计模型大致可分为两类：描述性模型和推断性模型。描述性模型通过均值、中位数和标准差等指标来概括数据。推断性模型利用概率理论从样本中得出关于总体的结论。

In the A-Level syllabus, you will encounter probability distributions such as the binomial distribution, Poisson distribution, and normal distribution. Each of these serves as a theoretical model for different types of real-world data.

在A-Level课程大纲中，你将接触到二项分布、泊松分布和正态分布等概率分布。每一种分布都作为不同类型现实世界数据的理论模型。

Binomial distribution models the number of successes in a fixed number of independent trials.
二项分布模拟固定次数独立试验中成功的次数。
Poisson distribution models the number of events occurring in a fixed interval of time or space.
泊松分布模拟在固定时间或空间间隔内发生的事件次数。
Normal distribution models continuous data that clusters around a central mean.
正态分布模拟围绕中心均值聚集的连续数据。

2. The Normal Distribution Model | 正态分布模型

The normal distribution is one of the most important models in statistics. It is a continuous probability distribution defined by the mean μ and the standard deviation σ. Its probability density function is given by:

正态分布是统计学中最重要的模型之一。它是一种由均值μ和标准差σ定义的连续概率分布。其概率密度函数为：

f(x) = (1/(σ√(2π))) × e^(−(x−μ)²/(2σ²))

The curve is symmetric about the mean and follows the empirical rule: approximately 68% of data lies within one standard deviation, 95% within two, and 99.7% within three standard deviations.

该曲线关于均值对称，并遵循经验法则：约68%的数据位于一个标准差内，95%位于两个标准差内，99.7%位于三个标准差内。

To solve problems involving the normal distribution, convert the variable X to the standard normal variable Z using the transformation:

为求解涉及正态分布的问题，需通过以下变换将变量X转换为标准正态变量Z：

Z = (X − μ) / σ

For example, if the weights of apples are normally distributed with mean 150 g and standard deviation 20 g, then the probability that a randomly selected apple weighs more than 180 g is computed as P(Z > (180 − 150)/20) = P(Z > 1.50). Using standard normal tables, this equals 1 − 0.9332 = 0.0668.

例如，若苹果重量服从均值为150克、标准差为20克的正态分布，则随机选取一个苹果重量超过180克的概率计算为P(Z > (180 − 150)/20) = P(Z > 1.50)。查标准正态分布表可得1 − 0.9332 = 0.0668。

3. The Binomial Distribution Model | 二项分布模型

The binomial distribution applies when a scenario involves a fixed number n of independent trials, each with the same probability p of success. The probability of exactly k successes is given by:

二项分布适用于固定次数n次独立试验、每次成功概率p相同的情形。恰好k次成功的概率为：

P(X = k) = ⁿCₖ × pᵏ × (1−p)ⁿ⁻ᵏ

The mean and variance of a binomial distribution are:

二项分布的均值和方差为：

E(X) = np, Var(X) = np(1−p)

The binomial model is widely used in quality control, medical trials, and opinion polls. For instance, if a drug is effective 80% of the time and is tested on 10 patients, the probability that exactly 7 patients recover is ¹⁰C₇ × 0.8⁷ × 0.2³ ≈ 0.2013.

二项模型广泛应用于质量控制、医学试验和民意测验中。例如，若某种药物有效率为80%，在10名患者身上进行试验，恰好7名患者康复的概率为¹⁰C₇ × 0.8⁷ × 0.2³ ≈ 0.2013。

When n is large and p is small, the binomial distribution can be approximated by the Poisson distribution, which simplifies calculations.

当n较大且p较小时，二项分布可以用泊松分布近似，从而简化计算。

4. The Poisson Distribution Model | 泊松分布模型

The Poisson distribution models the number of events occurring in a fixed interval of time or space, provided the events occur independently and at a constant average rate λ. Its probability mass function is:

泊松分布模拟在固定时间或空间间隔内发生的事件次数，前提是事件独立发生且以恒定的平均速率λ发生。其概率质量函数为：

P(X = k) = (λᵏ × e⁻λ) / k!

For this distribution, the mean and variance are both equal to λ:

对于该分布，均值和方差都等于λ：

E(X) = λ, Var(X) = λ

Real-world applications include modelling the number of telephone calls received per minute at a switchboard, the number of radioactive decays per second, or the number of accidents at a junction per month. For example, if calls arrive at a rate of 3 per minute, the probability of receiving exactly 5 calls in a minute is (3⁵ × e⁻³) / 5! ≈ 0.1008.

现实应用包括模拟电话总机每分钟接到的电话数量、每秒放射性衰变次数或某路口每月事故发生次数。例如，若电话以每分钟3个的速率到达，一分钟内恰好接到5个电话的概率为(3⁵ × e⁻³) / 5! ≈ 0.1008。

5. Sampling Methods and the Estimation of Parameters | 抽样方法与参数估计

Sampling is the process of selecting a subset of individuals from a population to estimate population parameters. Common sampling methods include simple random sampling, stratified sampling, systematic sampling, and cluster sampling.

抽样是从总体中选择个体子集以估计总体参数的过程。常见抽样方法包括简单随机抽样、分层抽样、系统抽样和整群抽样。

Simple random sampling: every member of the population has an equal chance of being selected.
简单随机抽样：总体中每个成员被选中的机会均等。
Stratified sampling: the population is divided into strata, and a proportional sample is drawn from each.
分层抽样：将总体划分为若干层，从每层中按比例抽取样本。
Systematic sampling: individuals are selected at regular intervals from an ordered list.
系统抽样：从有序列表中按固定间隔选取个体。

The sample mean x̄ is an unbiased estimator of the population mean μ. The sample standard deviation s, however, should be corrected using n − 1 in the denominator to obtain an unbiased estimate of the population variance.

样本均值x̄是总体均值μ的无偏估计量。然而，样本标准差s在计算时需使用n − 1作为分母，才能获得总体方差的无偏估计。

s² = Σ(xᵢ − x̄)² / (n − 1)

6. Confidence Intervals | 置信区间

A confidence interval provides a range of values within which the true population parameter is likely to lie, with a certain level of confidence. For a large sample with known population variance, the confidence interval for a population mean is given by:

置信区间提供了真实总体参数可能所在的取值范围，并带有一定的置信水平。对于大样本且总体方差已知的情况，总体均值的置信区间为：

x̄ ± z(α/2) × (σ/√n)

Here, z(α/2) is the critical value from the standard normal distribution (e.g., 1.96 for a 95% confidence level). A 95% confidence interval means that if we repeated the sampling process many times, approximately 95% of the intervals constructed would contain the true population mean.

其中z(α/2)是标准正态分布的临界值（例如，95%置信水平下为1.96）。95%置信区间意味着若重复多次抽样，约95%构建的区间将包含真实总体均值。

When the population variance is unknown and the sample size is small, the t-distribution is used instead of the normal distribution.

当总体方差未知且样本量较小时，应使用t分布而非正态分布。

7. Hypothesis Testing | 假设检验

Hypothesis testing is a formal procedure for deciding whether a claim about a population parameter is supported by the data. The null hypothesis H₀ represents the status quo, while the alternative hypothesis H₁ represents the claim we wish to test.

假设检验是一种正式程序，用于判断关于总体参数的声明是否得到数据支持。零假设H₀代表现状，而备择假设H₁代表我们希望检验的声明。

The test statistic is calculated from the sample data and compared with a critical value or used to compute a p-value. If the p-value is less than the significance level α (usually 0.05), the null hypothesis is rejected.

检验统计量根据样本数据计算，并与临界值比较或用于计算p值。若p值小于显著性水平α（通常为0.05），则拒绝零假设。

Z = (x̄ − μ₀) / (σ/√n)

For example, a manufacturer claims that the mean lifetime of its light bulbs is 1000 hours. A sample of 50 bulbs gives a mean of 980 hours with a known standard deviation of 60 hours. At α = 0.05, the test statistic is Z = (980 − 1000) / (60/√50) ≈ −2.36. Since this is less than the critical value of −1.96, we reject the null hypothesis and conclude that the mean lifetime is significantly different from 1000 hours.

例如，某制造商声称其灯泡的平均寿命为1000小时。抽取50个灯泡的样本，均值为980小时，已知标准差为60小时。在α = 0.05下，检验统计量为Z = (980 − 1000) / (60/√50) ≈ −2.36。由于该值小于临界值−1.96，我们拒绝零假设，并得出结论：平均寿命与1000小时有显著差异。

8. Correlation and Regression | 相关与回归

Correlation measures the strength and direction of a linear relationship between two variables. The Pearson product-moment correlation coefficient r ranges from −1 to +1.

相关衡量两个变量之间线性关系的强度和方向。皮尔逊积矩相关系数r的取值范围为−1到+1。

r = Σ(xᵢ − x̄)(yᵢ − ȳ) / √(Σ(xᵢ − x̄)² × Σ(yᵢ − ȳ)²)

A value of r close to +1 indicates a strong positive linear relationship, while a value close to −1 indicates a strong negative linear relationship. A value near 0 suggests little or no linear relationship.

r值接近+1表示强正线性相关，接近−1表示强负线性相关，接近0表示几乎没有线性关系。

The least-squares regression line, which models the relationship between a response variable y and an explanatory variable x, is given by:

最小二乘回归线用于模拟响应变量y与解释变量x之间的关系，其方程为：

y = a + bx

where the slope b and intercept a are calculated from the data. This model is used for prediction and trend analysis. For instance, a regression of exam score on study hours can help predict the expected score for a given number of study hours.

其中斜率b和截距a根据数据计算。该模型用于预测和趋势分析。例如，考试成绩对学习时数的回归可以帮助预测给定学习时数下的期望分数。

9. Model Selection and Goodness of Fit | 模型选择与拟合优度

Choosing the appropriate statistical model is critical. The chi-square (χ²) goodness-of-fit test is used to determine whether observed frequencies differ significantly from expected frequencies under a theoretical model.

选择适当的统计模型至关重要。卡方(χ²)拟合优度检验用于判断观察频数与理论模型下的期望频数是否有显著差异。

χ² = Σ (Oᵢ − Eᵢ)² / Eᵢ

where Oᵢ is the observed frequency and Eᵢ is the expected frequency for category i. The degrees of freedom equal the number of categories minus the number of estimated parameters minus one.

其中Oᵢ是第i类的观察频数，Eᵢ是期望频数。自由度等于类别数减去估计参数个数减一。

This test can verify, for example, whether a die is fair (uniform distribution) or whether data follows a normal distribution. A small chi-square value suggests a good fit, while a large value suggests the model is inappropriate.

该检验可验证例如骰子是否均匀（均匀分布）或数据是否服从正态分布。较小的卡方值表明拟合良好，而较大的值表明模型不适合。

10. Applications in Real-World Contexts | 现实世界中的应用

Statistical models are applied across many disciplines. In medicine, they are used to analyse the effectiveness of treatments and the spread of diseases. In economics, they model consumer behaviour and market trends. In engineering, they monitor product quality and process control.

统计模型广泛应用于众多学科。在医学中，用于分析治疗效果和疾病传播。在经济学中，用于模拟消费者行为和市场趋势。在工程中，用于监控产品质量和过程控制。

Field	Application	领域	应用
Quality control	Control charts monitor defect rates	质量控制	控制图监控缺陷率
Finance	Binomial model prices stock options	金融	二项模型为股票期权定价
Insurance	Poisson model estimates claims frequency	保险	泊松模型估计索赔频率
Education	Regression predicts student performance	教育	回归预测学生表现

Understanding the assumptions behind each model is essential. For example, the binomial model requires independent trials, and the Poisson model requires events to occur at a constant average rate. Violating these assumptions leads to inaccurate conclusions.

理解每个模型背后的假设至关重要。例如，二项模型要求试验独立，泊松模型要求事件以恒定平均速率发生。违反这些假设将导致不准确的结论。

11. Common Examination Pitfalls | 常见考试误区

Students often lose marks in statistics examinations due to common errors. These include confusing the sample standard deviation with the population standard deviation, using the wrong continuity correction, and misidentifying which distribution applies to a given scenario.

学生在统计考试中常因常见错误而失分。这些错误包括混淆样本标准差与总体标准差、使用错误的连续性修正，以及误判适用哪种分布。

Always state the distribution and its parameters clearly before calculating probabilities.
在计算概率之前，务必清楚说明分布及其参数。
Check whether the question asks for P(X = k) or P(X ≤ k).
检查题目要求的是P(X = k)还是P(X ≤ k)。
Apply the continuity correction when using a normal approximation to a discrete distribution.
使用正态近似离散分布时，务必进行连续性修正。
Draw a diagram of the normal curve to support your calculations.
绘制正态曲线图以支持你的计算。
In hypothesis testing, always write down H₀ and H₁ before calculating the test statistic.
在假设检验中，务必先写出H₀和H₁再计算检验统计量。

12. Exam-Style Worked Example | 典型考题示例

A café records the number of customers arriving every 10 minutes. Over a long period, the mean arrival rate is found to be 4.5 customers per 10 minutes. Assuming a Poisson model, find the probability that in a randomly chosen 10-minute interval, (a) exactly 5 customers arrive, and (b) at most 2 customers arrive.

某咖啡馆记录每10分钟到达的顾客人数。较长时间内，平均到达率为每10分钟4.5人。假设采用泊松模型，求在随机选取的10分钟间隔内：(a)恰好到达5名顾客的概率；(b)最多到达2名顾客的概率。

For part (a), use P(X = 5) = (4.5⁵ × e⁻⁴·⁵) / 5! . First compute 4.5⁵ = 1845.28125, and e⁻⁴·⁵ ≈ 0.0111089. Multiplying gives 1845.28125 × 0.0111089 ≈ 20.499, and dividing by 120 gives approximately 0.1708.

对于(a)，使用P(X = 5) = (4.5⁵ × e⁻⁴·⁵) / 5!。首先计算4.5⁵ = 1845.28125，e⁻⁴·⁵ ≈ 0.0111089。相乘得1845.28125 × 0.0111089 ≈ 20.499，除以120得约0.1708。

For part (b), P(X ≤ 2) = P(X = 0) + P(X = 1) + P(X = 2). These are e⁻⁴·⁵ ≈ 0.0111089, 4.5 × e⁻⁴·⁵ ≈ 0.049990, and (4.5² × e⁻⁴·⁵)/2 ≈ 0.112477. The sum is 0.0111089 + 0.049990 + 0.112477 ≈ 0.173576, so P(X ≤ 2) ≈ 0.174.

对于(b)，P(X ≤ 2) = P(X = 0) + P(X = 1) + P(X = 2)。分别为e⁻⁴·⁵ ≈ 0.0111089、4.5 × e⁻⁴·⁵ ≈ 0.049990和(4.5² × e⁻⁴·⁵)/2 ≈ 0.112477。三者之和为0.0111089 + 0.049990 + 0.112477 ≈ 0.173576，因此P(X ≤ 2) ≈ 0.174。

This example illustrates the direct application of the Poisson model to a real-life context, a skill frequently tested in examinations.

此示例说明了泊松模型在现实情境中的直接应用，这是考试中经常考查的技能。

Published by TutorHao | Mathematics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Statistical Models and Their Applications | 统计模型及其应用