How to Choose the Appropriate Statistical Method | 如何选择合适的统计方法

📚 How to Choose the Appropriate Statistical Method | 如何选择合适的统计方法

Selecting the correct statistical method is one of the most frequently tested skills in A-Level Mathematics. Examiners reward not just correct calculation but also sound reasoning about why a method is appropriate. This article provides a structured decision framework aligned with the requirements of Edexcel, CIE, AQA and OCR, covering data classification, summary measures, distributions, hypothesis tests and sampling.

选择合适的统计方法是A-Level数学中最高频考查的技能之一。考官奖励的不仅是通过计算得出的正确答案，更是关于某种方法为何适用的合理推理。本文提供了一套结构化的决策框架，与爱德思、CIE、AQA和OCR考试局的要求保持一致，涵盖数据分类、汇总度量、分布、假设检验和抽样方法。

1. Types of Data | 数据类型

Every statistical decision begins with identifying the data type. Data may be categorical (qualitative) or numerical (quantitative). Numerical data is further split into discrete and continuous types. Discrete data takes only isolated values, such as counts of students; continuous data can take any value in an interval, such as height or time.

每一个统计决策都始于对数据类型的判断。数据可以分为类别型（定性）和数值型（定量）。数值型数据进一步分为离散型和连续型。离散数据只能取孤立的值，如学生人数；连续数据可以在一个区间内取任何值，如身高或时间。

• Categorical data: use counts, modes, bar charts and chi-squared tests.

  类别型数据：使用计数、众数、条形图和卡方检验。

• Discrete numerical data: use means, medians, and binomial or Poisson distributions.

  离散数值数据：使用平均数、中位数，以及二项分布或泊松分布。

• Continuous numerical data: use histograms, box plots, the normal distribution and correlation measures.

  连续数值数据：使用直方图、箱线图、正态分布和相关度量。

A common exam trap is treating a discrete variable as continuous. For example, ‘number of cars passing per minute’ is discrete, even though it is counted over a time interval. Always ask: can this variable take fractional values?

常见的考试陷阱是把离散变量当作连续变量。例如，”每分钟经过的汽车数量”是离散的，即使它是在时间段内计数的。始终要问：这个变量能否取分数值？

2. Measures of Central Tendency | 集中趋势度量

Choosing between the mean, median and mode depends on the shape of the distribution and the presence of outliers.

在平均数、中位数与众数之间进行选择，取决于分布的形态以及是否存在异常值。

• Mean: use for symmetric distributions without outliers; it is the most mathematically powerful and supports further inference.

  平均数：适用于无异常值的对称分布；数学上最有力，并支持进一步的推断。

• Median: use for skewed distributions or when outliers are present, because the median is resistant to extreme values.

  中位数：适用于偏态分布或存在异常值时，因为中位数对极端值具有抗性。

• Mode: use for categorical data or when locating the most frequent value is meaningful.

  众数：适用于类别型数据，或当寻找最频繁出现的值具有实际意义时。

For grouped continuous data, the midpoint of the modal class approximates the mode, and the mean is estimated using midpoint × frequency.

对于分组连续数据，众数所在组的组中值可以近似众数，而平均数使用组中值 × 频数来估计。

x̄ = Σ(fx)/Σf

Published by TutorHao | Mathematics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）