📚 Summarizing Data | 数据总结
In statistics, summarizing data is the essential process of turning raw numbers into meaningful information. Whether you are analysing exam scores, measuring reaction times, or tracking climate data, raw data on its own tells you very little. To draw conclusions, make comparisons, or test hypotheses, you need to describe the central tendency, spread, and shape of the data using a handful of numerical measures and graphical displays. This chapter covers the core techniques: frequency tables, histograms, cumulative frequency, measures of central tendency (mean, median, mode), measures of dispersion (range, interquartile range, variance, standard deviation), box plots, skewness, and outlier detection. For IB Mathematics students, fluency with both ungrouped and grouped data is essential, as is the ability to interpret these summaries in context.
在统计学中，数据总结是将原始数字转化为有意义信息的关键过程。无论是分析考试成绩、测量反应时间，还是追踪气候数据，单看原始数据本身信息量极低。要得出结论、进行比较或检验假设，就需要用几个核心的数值度量和图表来描述数据的集中趋势、离散程度和分布形状。本章将涵盖频率表、直方图、累积频数、集中趋势的度量（平均数、中位数、众数）、离散程度的度量（极差、四分位距、方差、标准差）、箱线图、偏度和异常值检测等核心技术。对 IB 数学学生来说，熟练掌握未分组数据和分组数据的处理至关重要，同时要能够结合实际解读这些总结信息。
1. Types of Data | 数据类型
Before any summarising can begin, it is crucial to identify the type of data at hand. Data can be categorical (qualitative), such as colours or types of fruit, or numerical (quantitative). Numerical data is further split into discrete data, which can only take certain isolated values (e.g. number of students in a class), and continuous data, which can take any value within a range (e.g. height, time, mass). The methods used for summarising depend heavily on the data type: categorical data is often summarised with frequency tables, bar charts, and mode, while numerical data allows the full range of averages and measures of spread. In IB examinations, you will frequently be asked to choose and justify an appropriate summary method based on the data type and its distribution.
在开始任何总结工作之前，关键是确定手头的数据类型。数据可以是分类数据（定性），如颜色或水果种类，也可以是数值数据（定量）。数值数据进一步分为离散数据，只能取特定的孤立值（例如一个班级的学生人数），以及连续数据，可以在一个区间内取任意值（例如身高、时间、质量）。总结的方法很大程度上取决于数据类型：分类数据通常用频率表、条形图和众数来总结，而数值数据允许使用各种平均数及离散度量指标。在 IB 考试中，经常会要求你根据数据类型及其分布来选择并证明合适的总结方法。
2. Frequency Distributions for Ungrouped Data | 未分组数据的频数分布
When you first collect a set of discrete or categorical data, the simplest summary is a frequency table. You list each distinct value alongside its count (frequency) and sometimes the relative frequency (proportion or percentage). For small datasets, this instantly reveals the mode and the spread of values. It is also the starting point for calculating the mean of ungrouped data using the formula Σx/n for raw data or Σfx/Σf for a frequency distribution. A frequency table organises the data so you can quickly see which values are most common and whether there are any unusual gaps.
当你首次收集一组离散或分类数据时，最简单的总结就是频数表。你列出每个不同的值，并旁边写出它的计数（频数），有时也会记录相对频数（比例或百分比）。对于小型数据集，这能立刻揭示出众数和数值的分布范围。它也是基于原始数据使用公式 Σx/n，或基于频数分布使用 Σfx/Σf 来计算未分组数据平均数的起点。频数表将数据组织起来，让你能快速看出哪些值最常见，以及是否存在异常的空档。
3. Grouped Frequency Tables | 分组频数表
With continuous data or large discrete datasets, listing every individual value is impractical. Instead, we group the data into class intervals. A grouped frequency table records how many observations fall into each interval. When constructing such a table, intervals should be equal in width where possible, and boundaries must be clearly defined to avoid ambiguity. For example, if the class is 10–19, the next class is 20–29, assuming data are recorded to the nearest whole number. In IB problems, you may need to work backwards from a grouped table to estimate the mean, median, or percentiles, using the midpoint of each interval as a representative value.
对于连续数据或大型离散数据集，列出每个单独的值是不切实际的。因此，我们将数据分组到若干类区间中。分组频数表记录落在每个类区间内的观测值数量。在构建这样的表时，区间宽度应尽可能相等，且边界必须明确定义以避免歧义。例如，如果一类是 10–19，下一类是 20–29，这假设数据是四舍五入到整数记录的。在 IB 考题中，你可能需要根据分组表逆向估算平均值、中位数或百分位数，使用每个区间的中点作为代表值。
4. Histograms and Frequency Density | 直方图与频率密度
s straightforward. However, when intervals have unequal widths, you must calculate frequency density = frequency ÷ class width. In IB Mathematics, you are expected to draw histograms with frequency density on the vertical axis and to use area to determine frequencies, especially when asked to estimate totals or proportions from a histogram with varying bin widths.
直方图是连续数据分组频数分布的图形表示。与条形图不同，直方图中每个长方形的面积代表频数，而非高度。当所有类区间的宽度相等时，高度与频数成正比，绘制直方图很简单。然而，当区间宽度不相等时，必须计算频率密度 = 频数 ÷ 区间宽度。在 IB 数学中，你应该能够以频率密度为纵轴绘制直方图，并利用面积来确定频数，尤其是在面对不同箱子宽度的直方图，需要估算总数或比例时。
5. Cumulative Frequency Curves | 累积频数曲线
A cumulative frequency curve, or ogive, shows the accumulation of frequencies as you move through the class intervals. To construct it, you add up the frequencies from the first interval onwards, plotting the cumulative total against the upper boundary of each class. The resulting curve is used to estimate medians, quartiles, and percentiles by reading values from the graph. For instance, the median corresponds to the 50th percentile, found by drawing a horizontal line from the halfway point on the cumulative frequency axis to the curve, then dropping vertically to the data axis. IB exam questions often provide a partially completed cumulative frequency table and ask you to complete the graph, then estimate key values and interpret them in real-life terms.
累积频数曲线是一种展示随着类区间推移频数累积情况的图形。要构建它，你需要从第一个区间开始累加频数，并将累积总数对应于每个类的上界进行描点。由此得到的曲线可用于估算中位数、四分位数和百分位数，通过读取图上的数值来实现。例如，中位数对应第 50 百分位数，只需从累积频数轴的中点画一条水平线到曲线，再垂直向下画到数据轴。IB 考试题目经常给出一个部分完成的累积频数表，要求你完成图形，然后估算关键值，并结合实际进行解读。
6. Measures of Central Tendency: Mean, Median, Mode | 集中趋势的度量：平均数、中位数、众数
The three most common averages serve different purposes. The mean (x̄) is the arithmetic average, highly sensitive to every single data point, including outliers. It is calculated as Σx/n for raw data or Σfx/Σf for frequency distributions. The median is the middle value when data are ordered, resistant to extreme values, and often a better measure of centre for skewed distributions. The mode is the most frequently occurring value; a dataset can have no mode, one mode (unimodal), or more than one (bimodal). Understanding when to use each average is a key statistical skill. In IB, you may be asked to explain why the median is more appropriate than the mean for income data, or why the mode is best for categorical shoe sizes.
最常见的三种平均数用途各异。平均数（x̄）是算术平均值，对每一个数据点（包括异常值）都非常敏感。对于原始数据，计算公式为 Σx/n；对于频数分布，计算公式为 Σfx/Σf。中位数是将数据排序后位于中间的值，能够抵抗极端值的影响，对于偏态分布往往是更好的中心度量。众数是出现频率最高的值；一个数据集可能没有众数、有一个众数（单峰），或者有多个众数（双峰）。理解每种平均数的适用场景是一项关键的统计技能。在 IB 中，你可能会被要求解释为什么收入数据用中位数比平均数更合适，或者为什么鞋码的众数最适合鞋类尺码这样的分类数据。
7. Quartiles and the Interquartile Range (IQR) | 四分位数与四分位距
ian, and the third quartile (Q₃) is the 75th percentile. The interquartile range, IQR = Q₃ – Q₁, measures the spread of the middle 50% in constructing box plots and in formal outlier detection rules, where any value below Q₁ – 1.5 × IQR or above Q₃ + 1.5 × IQR is flagged as a potential outlier.
四分位数将一个有序数据集分成四个相等的部分。第一四分位数（Q₁）对应第 25 百分位数，第二四分位数（Q₂）即是中位数，第三四分位数（Q₃）是第 75 百分位数。四分位距 IQR = Q₃ – Q₁，衡量的是中间 50% 数据的离散程度，完全不收异常值影响。这使得 IQR 成为一个非常稳健的离散度量指标。当数据被分组时，你可以通过累积频数图估算四分位数。IQR 还用于构建箱线图以及正式的异常值检测规则中：任何低于 Q₁ – 1.5 × IQR 或高于 Q₃ + 1.5 × IQR 的值被标记为潜在的异常值。
8. Variance and Standard Deviation | 方差与标准差
verage of the squared deviations from the mean. For a population, the variance is σ² = Σ(x – μ)² / N. For a sample, we use s² = Σ(x – x̄)² / (n – ectly interpretable. In IB, you will use the formula that best fits the context: usually the sample standard deviation for data drawn from a larger population. Standard deviation is particularly useful when data follow an approximately normal distribution, as it links directly to percentages within 1, 2, or 3 standard deviations from the mean.
IQR 关注中间一半的数据，而方差和标准差则考虑了每一个数据点。方差是各个数据与平均数之差的平方的平均值。对于总体，方差 σ² = Σ(x – μ)² / N。对于样本，我们使用 s² = Σ(x – x̄)² / (n – 1) 以提供无偏估计。标准差就是方差的平方根，恢复了原始单位，使其可直接解释。在 IB 中，你会根据上下文选用最适合的公式：对于从较大总体中抽取的数据，通常使用样本标准差。当数据近似服从正态分布时，标准差尤为有用，因为它直接与平均数加减 1、2 或 3 个标准差范围内的百分比相联系。
9. Box-and-Whisker Plots (Box Plots) | 箱线图
A box plot is a compact graphical summary based on the five-number summary: minimum, Q₁, median, Q₃, and maximum. The box spans from Q₁ to Q₃, with a line inside marking the median. Whiskers extend to the smallest and largest data points that are not outliers. Outliers are plotted individually as points or asterisks. Box plots are especially useful for comparing two or more datasets side by side because they reveal differences in median, spread, skewness, and the presence of outliers at a glance. In IB investigations, you might be required to construct parallel box plots and then write a comparative analysis, referring to both measures of centre and spread.
箱线图是一种基于五数概括的紧凑型图形总结：最小值、Q₁、中位数、Q₃ 和最大值。箱体从 Q₁ 延伸到 Q₃，中间有一条线标记中位数。箱须从箱体延伸到非异常值的最小和最大数据点。异常值会单独绘制为点或星号。箱线图对于并列比较两个或多个数据集尤为有用，因为它能让你一眼看出中位数、离散程度、偏度以及异常值存在的差异。在 IB 探究任务中，你可能需要构建平行箱线图，然后撰写比较分析，同时论及中心和离散的度量。
10. Skewness and Its Interpretation | 偏度及其解读
The shape of a distribution tells you a great deal about the underlying data. A symmetric distribution has the mean and median approximately equal. In a positively skewed (right-skewed) distribution, the mean is greater than the median, and the tail extends to the right; typical examples include household income or reaction times to a stimulus. In a negatively skewed (left-skewed) distribution, the mean is less than the median, and the tail stretches to the left (e.g. age at retirement or test scores where most students score highly). Recognising skewness helps in choosing appropriate summaries: for skewed data, the median and IQR are preferable to the mean and standard deviation. IB questions often show a histogram or box plot and ask you to infer the skewness and justify your reasoning using the positions of the mean and median.
分布的形状能告诉你许多关于数据的信息。对称分布的平均数和中位数大致相等。在正偏（右偏）分布中，平均数大于中位数，尾部向右延伸；典型的例子包括家庭收入或对刺激的反应时间。在负偏（左偏）分布中，平均数小于中位数，尾部向左延伸（例如退休年龄或多数学生得分较高的考试成绩）。识别偏度有助于选择合适的总结方法：对于偏态数据，中位数和 IQR 比平均数和标准差更可取。IB 题目经常给出一个直方图或箱线图，要求你推断偏度，并运用平均数和中位数的位置来论证你的推理。
11. Outliers and Their Impact | 异常值及其影响
s the standard method for identifying potential outliers. When outliers are present, they can dramatically inflate the mean, variance, and standard deviation, making the dataset appear more spread out than it truly is. On the other hand, the median and IQR remain stable. In IB, you should never simply remove an outlier without justification. You must discuss its possible cause, its effect on summary statistics, and whether it should be retained or excluded from the analysis based on context.
异常值是与其它值存在异常距离的数据点。它们可能源于测量误差、数据录入错误或真实的自然变异。1.5 × IQR 规则是识别潜在异常值的标准方法。当存在异常值时，它们会显著拉高平均值、方差和标准差，使数据集看起来比实际上更分散。相反，中位数和 IQR 保持稳定。在 IB 中，你不应该在没有理由的情况下简单地删除一个异常值。你必须讨论它可能的原因、它对总结统计量的影响，以及根据上下文是应该保留还是从分析中剔除它。
12. Choosing the Right Summary and Comparing Datasets | 选择合适的总结与数据集比较
When presented with real data, summarising effectively means choosing the most appropriate measures. For symmetric distributions with no outliers, the mean and standard deviation provide a complete description. For skewed distributions or those with outliers, the median and IQR are more robust. When comparing two or more datasets, always comment on both a measure of centre (mean or median) and a measure of spread (standard deviation or IQR), and refer to the shape (skewness). Graphical tools like histograms, box plots, and cumulative frequency curves allow you to visualise the entire distribution, not just the numeric summaries. IB examiners reward responses that combine numerical comparisons with insightful contextual interpretations, such as linking a lower IQR in one group to more consistent performance.
在处理真实数据时，有效的数据总结意味着选择最合适的度量。对于没有异常值的对称分布，平均数和标准差可以提供一个完整的描述。对于偏态分布或含有异常值的数据，中位数和 IQR 更加稳健。当比较两个或多个数据集时，要始终同时评述中心度量（平均数或中位数）和离散度量（标准差或 IQR），并提及形状（偏度）。像直方图、箱线图和累积频数曲线这样的图形工具可以让你看到整个分布的全貌，而不仅仅是数值总结。IB 阅卷人青睐那些将数值比较与富有洞察力的实际解读相结合的答案，例如将某一组较低的 IQR 与更稳定的表现联系起来。
Published by TutorHao | IB 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