📚 Investigation 1 – Merits of the Mean and Median | 调查1：平均值与中位数的优点
In statistics, measures of central tendency provide a single value that summarises the centre of a data set. The arithmetic mean and the median are the most frequently used measures, each with distinct merits. Understanding their strengths helps in selecting the appropriate summary for a given dataset, ensuring accurate interpretation. This investigation explores the relative advantages of the mean and median, focusing on how data characteristics such as symmetry, outliers, and measurement scale influence their suitability.
在统计学中，集中趋势度量通过一个单一数值概括数据集的中心。算术平均值和中位数是最常用的两种度量，各具独特的优点。理解它们的优势有助于为给定的数据集选择合适的概括方式，从而保证解读的准确性。本次调查探讨平均值与中位数的相对优势，重点关注数据的对称性、异常值和测量尺度等特征如何影响它们的适用性。
1. The Concept of Central Tendency | 集中趋势的概念
A measure of central tendency identifies the typical or central value in a set of observations. It reduces the complexity of raw data to a single representative number, facilitating comparisons and decision-making. While several measures exist, the mean and median dominate due to their interpretability and mathematical properties.
集中趋势度量确定一组观测值中的典型值或中心值。它将原始数据的复杂性简化为一个代表性数字，便于进行比较和决策。尽管存在多种度量，但平均值和中位数因其可解释性和数学性质而占据主导地位。
2. Defining the Arithmetic Mean | 算术平均值的定义
The arithmetic mean, often simply called the mean, is the sum of all data values divided by the number of observations. For a sample of size n with values x₁, x₂, …, xₙ, the sample mean x̄ is calculated as:
算术平均值（通常简称为平均值）是所有数据值之和除以观测值个数。对于样本量为 n、取值分别为 x₁, x₂, …, xₙ 的样本，样本均值 x̄ 的计算公式为：
x̄ = (x₁ + x₂ + … + xₙ) / n = Σ xᵢ / n
The mean incorporates every data point equally, making it sensitive to all values in the set. This property can be both an advantage and a limitation, depending on the nature of the data.
平均值同等地纳入每一个数据点，因此对数据集中的所有值都敏感。这一特性既可以是优点，也可以是局限，具体取决于数据的性质。
3. Defining the Median | 中位数的定义
The median is the middle value when the data are arranged in ascending or descending order. If the number of observations n is odd, the median is the (n+1)/2-th value. If n is even, it is the average of the n/2-th and (n/2 + 1)-th values. Unlike the mean, the median depends only on the rank-order of the data, not on their actual magnitudes beyond the central position.
中位数是将数据按升序或降序排列后位于中间位置的值。如果观测值个数 n 为奇数，中位数是第 (n+1)/2 个值；如果 n 为偶数，则是第 n/2 个和第 (n/2 + 1) 个值的平均数。与平均值不同，中位数仅依赖于数据的排列顺序，而不受中心位置以外具体数值大小的影响。
4. Sensitivity to Outliers | 对异常值的敏感性
One of the most important differences between the mean and median lies in their response to extreme values. Because the mean uses every value in its calculation, a single outlier can drastically shift it, whereas the median remains largely unaffected. Consider the data set: 3, 4, 5, 6, 102. The mean is (3+4+5+6+102)/5 = 24, but the median is 5. In this case, the median better reflects the central cluster of the data.
平均值与中位数最重要的区别之一在于它们对极端值的反应。由于平均值在计算时使用了每一个数值，单个异常值就能使其发生显著偏移，而中位数则基本不受影响。以数据集 3, 4, 5, 6, 102 为例：平均值为 (3+4+5+6+102)/5 = 24，而中位数为 5。在这个例子中，中位数更准确地反映了数据的中心聚集状态。
This robustness makes the median the preferred measure when data contain recording errors, measurement anomalies, or naturally occurring outliers, such as income or property prices.
这种稳健性使得中位数在数据包含记录错误、测量异常或天然存在极端值（如收入或房产价格）时成为首选度量。
5. Merits of the Mean in Symmetric Distributions | 平均值在对称分布中的优点
When the data follow a symmetric distribution, such as the normal distribution, the mean and median coincide. In these situations, the mean is favoured because it uses all available information and has smaller sampling variability than the median. For a perfectly symmetric bell-shaped curve, the mean is the maximum likelihood estimator of the population centre and forms the basis of classical statistical inference.
当数据服从对称分布（如正态分布）时，平均值与中位数重合。在这种情况下，平均值更受青睐，因为它利用了所有的信息，且其抽样变异性比中位数更小。对于完全对称的钟形曲线，平均值是总体中心的极大似然估计，并构成经典统计推断的基础。
6. Merits of the Median in Skewed Distributions | 中位数在偏态分布中的优点
Real-world data are frequently skewed. For instance, household incomes, housing prices, and reaction times often exhibit right-skewness. In a skewed distribution, the mean is pulled in the direction of the tail, away from the bulk of the data, while the median stays near the peak. Because the median represents the 50th percentile, exactly half of the observations lie below it and half above, giving a more accurate sense of a “typical” value.
现实世界中的数据常常是偏态的。例如，家庭收入、房价和反应时间通常呈现右偏分布。在偏态分布中，平均值被拉向尾部方向，远离大部分数据，而中位数则停留在峰值附近。由于中位数代表第50百分位数，恰好有一半的观测值在它之下、一半在上，从而更准确地传达了“典型”值的概念。
Consequently, official statistics on income and expenditure often report median values to avoid distortion by extreme wealth.
因此，收入和支出的官方统计常报告中位数，以避免极端财富造成扭曲。
7. Mathematical Tractability of the Mean | 平均值的数学可处理性
The mean possesses algebraic properties that make it indispensable for further statistical analysis. It is linear: the mean of a sum is the sum of means, and the mean of a transformed variable aX + b is a·x̄ + b. This linearity underpins methods such as analysis of variance (ANOVA), linear regression, and the calculation of standard deviation, where the mean plays a central role. The median does not share these convenient algebraic features.
平均值具有代数性质，这使其在进一步的统计分析中不可或缺。它是线性的：和的平均值等于各平均值的和，经变换 aX + b 后的均值等于 a·x̄ + b。这种线性性质奠定了方差分析、线性回归以及标准差计算等方法的基础，而这些方法都以平均值为核心。中位数并不具备这些便捷的代数特征。
Furthermore, the mean minimizes the sum of squared deviations: Σ(xᵢ − c)² is minimized when c = x̄. This property leads to the concept of variance and the widespread use of least squares estimation.
此外，平均值使偏差平方和最小化：当 c = x̄ 时，Σ(xᵢ − c)² 取得最小值。这一性质导出了方差的概念以及最小二乘估计的广泛应用。
8. Stability under Sampling and the Central Limit Theorem | 抽样稳定性与中心极限定理
The sampling distribution of the sample mean is well understood. According to the Central Limit Theorem, for sufficiently large samples, the sample mean is approximately normally distributed regardless of the population distribution, with a standard error σ/√n. This allows statisticians to construct confidence intervals and perform hypothesis tests for population means with great reliability. The sample median, while having its own asymptotic normality, typically exhibits a larger standard error than the mean in symmetric populations, making the mean a more efficient estimator.
样本均值的抽样分布已被充分理解。根据中心极限定理，对于足够大的样本，无论总体分布如何，样本均值都近似服从正态分布，其标准误为 σ/√n。这使得统计人员能够可靠地构建总体均值的置信区间并进行假设检验。样本中位数虽然也具有渐近正态性，但在对称总体中通常表现出比均值更大的标准误，因此均值是更有效的估计量。
9. Ease of Interpretation and Communication | 解释与沟通的方便性
The mean is intuitive: the idea of sharing equally or “splitting the total” is easily grasped by the public. The median, defined as the middle point, is also straightforward—50% of the data fall on either side. In communication, however, the median sometimes requires explaining the concept of ordering and ranking, whereas the mean is immediately associated with the notion of an “average”. Yet, when a dataset contains extremes, explaining the mean can be more challenging because it may not correspond to any actual observation.
平均值直观易懂：平摊或“均分总数”的概念容易被公众理解。中位数作为中间点也同样直接——各50% 的数据分布在两侧。但在沟通中，中位数有时需要解释排序和排名的概念，而平均值则立刻让人联想到“平均”的概念。然而，当数据集中包含极端值时，解释平均值可能更具挑战性，因为它可能不与任何一个实际观测值对应。
10. Real-World Applications | 实际应用场景
In education, mean test scores are commonly used to assess overall class performance because scores often follow a roughly symmetric distribution. In economics, median income or median house price is reported to present a picture unaffected by the ultra-rich. In quality control, the mean of product measurements is tracked to detect shifts in manufacturing processes, while the median might be used in non-parametric tests when data are not normal. The choice between mean and median is thus highly domain-dependent.
在教育领域，通常使用平均考试成绩来评估班级整体表现，因为分数往往接近对称分布。在经济领域，报告中位数收入或中位数房价是为了呈现不受超级富豪影响的情况。在质量控制中，跟踪产品测量值的均值以检测制造过程的变化，而当数据非正态时，或许会使用中位数进行非参数检验。因此，平均值与中位数的选择高度依赖于应用领域。
11. Choosing the Right Measure for Your Investigation | 为调查选择合适的度量
When conducting a statistical investigation, the decision to use the mean or the median should be guided by:
- Data shape: For symmetric data, the mean is more informative; for skewed data, the median is more representative.
- Presence of outliers: If outliers are errors, removing them and using the mean may be best; if they are genuine, the median offers protection.
- Purpose of analysis: If further inference or mathematical modelling is required, the mean is usually essential. If a simple resistant summary is needed, the median suffices.
- Measurement scale: The mean requires interval or ratio data; the median can be used with ordinal data as well.
Reporting both measures alongside the shape of the distribution often provides the most complete picture.
在进行统计调查时，选择均值还是中位数应遵循以下指导：
- 数据形态：对于对称数据，平均值信息更丰富；对于偏态数据，中位数更具代表性。
- 异常值的存在：若异常值是错误，最好剔除后使用平均值；若为真实值，则中位数能提供保护。
- 分析目的：若需进一步推断或数学建模，平均值通常是必需的。若仅需简单的抵抗性概括，中位数即可满足。
- 测量尺度：平均值需要定距或定比数据；中位数还可用于定序数据。
同时报告两个度量并附上分布形态，往往能呈现最完整的图景。
12. Summary of Merits | 优点总结
The mean and median are complementary tools in descriptive statistics. The mean’s merits include full utilisation of data, algebraic simplicity, optimality under normality, and essential role in inferential statistics. The median’s merits lie in its robustness to outliers, meaningfulness in skewed distributions, and applicability to ordinal data. An informed analyst weighs these merits against the characteristics of the dataset and the goals of the investigation to select the most appropriate measure—or employs both for a more comprehensive summary.
平均值和中位数是描述统计学中互补的工具。平均值的优点包括：充分利用数据、代数简洁、在正态性下的最优性以及在推断统计中的核心作用。中位数的优点则在于对异常值的稳健性、在偏态分布中的意义以及对定序数据的适用性。明智的分析者会结合数据集的特征与研究目标权衡这些优点，从而选择最合适的度量——或同时采用两者以求更全面的概括。
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