IB Mathematics: Foundations and Key Points of Regression Analysis | IB数学：回归分析基础与要点

📚 IB Mathematics: Foundations and Key Points of Regression Analysis | IB数学：回归分析基础与要点

Regression analysis is one of the most practical and frequently tested topics in IB Mathematics, appearing in both Analysis and Approaches (AA) and Applications and Interpretation (AI) courses. This article covers the essential foundations, formulas, and exam-focused strategies to help you master this topic with confidence.

回归分析是IB数学中最实用且高频考查的主题之一，在分析与方法（AA）和应用与解释（AI）课程中均有涉及。本文系统梳理回归分析的基础知识、核心公式和考试要点，助你从容应对各类题型。

1. What Is Regression Analysis? | 什么是回归分析？

Regression analysis is a statistical method used to model the relationship between two or more variables. In its simplest form, we study how a dependent variable (often denoted as y) changes in response to an independent variable (often denoted as x). The goal is to find a mathematical function that best fits the observed data points.

回归分析是一种用于建立两个或多个变量之间关系的统计方法。在最简单的形式中，我们研究因变量（通常记为 y）如何随自变量（通常记为 x）的变化而变化。目标是找到一条最能拟合观测数据点的数学函数。

In IB Mathematics, the most common form is linear regression, where we model the relationship using a straight line. However, you may also encounter non-linear forms such as exponential or quadratic regression, especially in the AI course.

在IB数学中，最常见的形式是线性回归，即用一条直线来拟合变量间的关系。当然，在AI课程中也可能遇到指数回归、二次回归等非线性形式。

2. Scatter Diagrams and Correlation | 散点图与相关性

Before performing regression, we first plot the data on a scatter diagram. A scatter diagram is a graph that displays each pair of values (x, y) as a point on the coordinate plane. It visually reveals the direction, form, and strength of the relationship.

在进行回归分析之前，我们首先将数据绘制成散点图。散点图是将每一对数据 (x, y) 在平面直角坐标系中描点得到的图形，能够直观地展示变量间关系的方向、形式和强度。

Positive correlation: as x increases, y tends to increase.
Negative correlation: as x increases, y tends to decrease.
No correlation: there is no apparent linear pattern.

正相关：x 增大时，y 总体呈增大趋势。
负相关：x 增大时，y 总体呈减小趋势。
无相关：数据点没有明显的线性规律。

The scatter diagram also helps identify outliers — extreme values that may disproportionately influence the regression line.

散点图还有助于识别异常值——那些可能对回归线产生过大影响的极端数据点。

3. The Linear Regression Model | 线性回归模型

The linear regression model assumes that the relationship between x and y can be approximated by a straight line of the form:

线性回归模型假设 x 与 y 之间的关系可以用一条直线近似描述，其一般形式为：

y = a + bx

Here, a is the intercept (the value of y when x = 0), and b is the slope (the change in y for each unit increase in x). The values of a and b are estimated from the sample data using the least squares method, which we will explain in the next section.

其中，a 为截距（即 x = 0 时的 y 值），b 为斜率（即 x 每增加一个单位时 y 的平均变化量）。a 和 b 的值通过最小二乘法从样本数据中估计得出，我们将在下一节详细说明。

In the IB syllabus, the regression line is often written as y on x, meaning we predict y from x. This is important: the line y = a + bx is not the same as the line of x on y.

在IB课程中，回归线通常写成 y on x 的形式，即用 x 来预测 y。这一点非常重要：y on x 的回归线与 x on y 的回归线并不相同。

4. The Method of Least Squares | 最小二乘法

The least squares method is the standard technique for finding the line of best fit. It minimises the sum of the squares of the vertical distances between each observed point and the regression line. Mathematically, we minimise:

最小二乘法是寻找最优拟合线的标准方法。它通过最小化每个观测点到回归线的垂直距离的平方和来实现。数学上，我们最小化：

S = Σ(yᵢ − (a + bxᵢ))²

Using calculus, the values of a and b that minimise S can be shown to be:

利用微积分可以证明，使 S 达到最小的 a 和 b 分别为：

b = (Σ(xᵢ − x̄)(yᵢ − ȳ)) / Σ(xᵢ − x̄)²

a = ȳ − b·x̄

where x̄ and ȳ are the means of the x-values and y-values respectively. Notice that the regression line always passes through the point (x̄, ȳ), the centroid of the data.

其中 x̄ 和 ȳ 分别为 x 和 y 的平均值。注意，回归线总是通过数据点的质心 (x̄, ȳ)。

In the IB exam, you are not expected to compute these by hand in most cases. Instead, you should be proficient in using your GDC (graphical display calculator) to obtain a and b instantly from a given data set.

在IB考试中，大多数情况下你不需要手算这些公式。你应该熟练使用图形计算器（GDC），从给定数据集中快速求出 a 和 b。

5. Interpreting the Regression Line | 回归线的解释

Once the regression line is obtained, it is used for two main purposes: describing the relationship and making predictions.

得到回归线后，它主要用于两个目的：描述关系以及进行预测。

Description: The slope b tells us the average rate of change of y per unit change in x. The intercept a gives the predicted value of y when x = 0.

描述：斜率 b 表示 x 每变化一个单位时 y 的平均变化量。截距 a 给出当 x = 0 时 y 的预测值。

Prediction: Substituting a given x-value into the equation y = a + bx yields the predicted y-value. This process is called interpolation when the x-value lies within the observed range, and extrapolation when it lies outside. Extrapolation is risky because the relationship may not hold beyond the data range.

预测：将给定的 x 值代入方程 y = a + bx 即可得到 y 的预测值。若 x 值在观测范围内，这一过程称为内插；若超出观测范围，则称为外推。外推具有较大风险，因为变量间的关系在数据范围之外可能不再成立。

6. Pearson’s Correlation Coefficient r | 皮尔逊相关系数 r

While the regression line describes how y changes with x, the correlation coefficient r measures the strength and direction of the linear relationship between x and y.

回归线描述的是 y 如何随 x 变化，而相关系数 r 则衡量 x 与 y 之间线性关系的强度和方向。

r = Σ(xᵢ − x̄)(yᵢ − ȳ) / √[Σ(xᵢ − x̄)² · Σ(yᵢ − ȳ)²]

Key properties of r:

r 的关键性质如下：

−1 ≤ r ≤ 1
r close to 1 implies a strong positive linear relationship.
r close to −1 implies a strong negative linear relationship.
r close to 0 implies a weak or no linear relationship.
r is dimensionless and symmetric (it does not distinguish between x and y).

−1 ≤ r ≤ 1
r 接近 1 表示强正线性相关。
r 接近 −1 表示强负线性相关。
r 接近 0 表示线性关系很弱或不存在。
r 是无量纲的且对称（它不区分 x 和 y）。

Note: a large r does not necessarily imply causation. Correlation is not causation — two variables may be strongly correlated simply by coincidence or due to a hidden third variable.

请注意：r 很大并不一定意味着因果关系。相关不等于因果——两个变量可能只是碰巧高度相关，或者受到某个隐藏的第三变量的影响。

7. Coefficient of Determination r² | 决定系数 r²

Another important statistic is the coefficient of determination, denoted r². It represents the proportion of the variance in y that can be explained by the linear relationship with x.

另一个重要的统计量是决定系数，记作 r²。它表示 y 的方差中能被 x 的线性关系所解释的比例。

For example, if r = 0.8, then r² = 0.64, meaning that 64% of the total variation in y is explained by the regression line, while the remaining 36% is due to other factors or random error.

例如，若 r = 0.8，则 r² = 0.64，意味着 y 的总变异性中有 64% 可以由回归线解释，其余 36% 由其他因素或随机误差引起。

In IB exam questions, you are often asked to interpret r or r² in context. Always refer to the actual variables (e.g., “64% of the variation in exam score is accounted for by study time”), not just a generic statistical statement.

在IB考题中，你经常需要结合具体情境来解释 r 或 r²。永远要联系实际变量进行说明（例如“考试成绩的变异中64%可由学习时间解释”），而不仅仅是给出一个笼统的统计学表述。

8. Residuals and the Goodness of Fit | 残差与拟合优度

A residual is the difference between an observed y-value and the value predicted by the regression line. In symbols:

残差是观测到的 y 值与回归线预测值之间的差。用符号表示为：

residual = yᵢ − ŷᵢ

where ŷᵢ is the fitted value from the regression line. Residuals are central to assessing whether a linear model is appropriate.

其中 ŷᵢ 为回归线的拟合值。残差是评估线性模型是否合适的关键工具。

A residual plot (residuals against x) is a useful diagnostic tool. If the residual plot shows a random pattern, the linear model is appropriate. If it shows a curved pattern, a non-linear model may be more suitable.

残差图（以 x 为横轴、残差为纵轴的图形）是有效的诊断工具。若残差图呈现随机分布，说明线性模型是合适的；若呈现弯曲趋势，则说明非线性模型可能更合适。

9. Non-Linear Regression and Transformations | 非线性回归与数据变换

In IB AI (Applications and Interpretation), you may need to fit non-linear models such as exponential, logarithmic, or power functions to data. A common technique is to transform the data so that the transformed variables have a linear relationship.

在IB AI课程中，你可能需要将指数函数、对数函数或幂函数等非线性模型拟合到数据中。一种常用方法是先对数据进行变换，使变换后的变量之间呈现线性关系。

For example, if the data follow an exponential relationship y = c·dˣ, taking the natural logarithm of both sides gives:

例如，若数据符合指数关系 y = c·dˣ，对两边取自然对数可得：

ln y = ln c + x · ln d

This is now a linear relationship between ln y and x, with slope ln d and intercept ln c. With your GDC, you can perform a linear regression on (x, ln y) to estimate c and d.

这就是 ln y 与 x 之间的线性关系，斜率为 ln d，截距为 ln c。使用图形计算器，你可以对 (x, ln y) 进行线性回归，从而估算出 c 和 d。

The GDC also provides a direct option for fitting several regression models (linear, quadratic, exponential, power, sinusoidal, etc.). Always check which model gives the highest r² value, but also ensure the model makes sense in the context of the problem.

图形计算器还提供了多种回归模型的直接拟合选项（线性、二次、指数、幂、正弦等）。通常应选择 r² 值最高的模型，但也要确保该模型在实际问题情境中有意义。

10. Key Exam Tips and Common Pitfalls | 考试要点与常见错误

Here are the most important tips and pitfalls to keep in mind when answering regression questions in the IB exam.

以下是在IB考试中回答回归分析问题时最重要的技巧和常见误区。

Use GDC correctly: In AA and AI exams, you are expected to obtain regression equations from your calculator. Enter the data as lists, run the correct regression function, and clearly write the equation in the form y = a + bx.
Circle the right variables: Pay attention to which variables are defined as x and y in the question. Do not swap them.
Interpolation vs extrapolation: Interpolating within the data range is safe; extrapolating outside it is unreliable. Always state this limitation when making predictions.
Do not confuse r and r²: Both are important but measure different things. r gives direction and strength, r² gives explanatory power.
Watch significant figures: IB questions often specify “give your answer correct to 3 significant figures.” Follow this instruction exactly.
Interpret in context: Always link your interpretation of slope, intercept, and correlation to the specific variables in the question.

正确使用GDC：AA和AI考试要求你用计算器求回归方程。将数据输入列表，运行正确的回归功能，并以 y = a + bx 的形式清晰书写方程。
明确变量：注意题目中 x 和 y 分别代表什么，不要交换位置。
内插与外推：在数据范围内内插是安全的；超出范围外推则不可靠。进行预测时一定要说明这一局限。
不要混淆 r 和 r²：两者都重要但含义不同。r 表示方向和强度，r² 表示解释能力。
注意有效数字：IB题目常要求“答案保留三位有效数字”，务必按要求作答。
结合情境解释：解释斜率、截距和相关系数时，务必将结论与题目中的具体变量联系起来。

11. Worked Example | 例题解析

Consider the following exam-style question: A student records the number of hours studied (x) and the corresponding test scores (y) for 8 students. The results are shown in the table below. Find the regression line of y on x and use it to predict the test score for a student who studies for 6.5 hours.

来看一道典型考题：某学生记录了8名同学的学习时间（x，单位：小时）与对应的测验成绩（y）。数据如下表所示。求 y on x 的回归线，并用它预测学习时间为6.5小时的同学的测验成绩。

x (hours)	2	3	4	5	6	7	8	9
y (score)	45	50	55	63	70	75	82	88

Solution: Using the GDC, we obtain the regression line:

解：使用图形计算器，可得回归线：

y = 33.1 + 6.18x

(correct to 3 significant figures for the coefficients). Substituting x = 6.5 gives:

（系数保留三位有效数字）。代入 x = 6.5，可得：

y = 33.1 + 6.18 × 6.5 = 73.3

Thus, the predicted score is approximately 73 (rounded to the nearest integer). Since x = 6.5 lies within the observed range of x (2 to 9), this prediction is an interpolation and is considered reliable.

因此预测成绩约为73分（四舍五入到整数）。由于 x = 6.5 在观测范围（2至9）之内，该预测属于内插，具有较高可靠性。

12. Summary | 总结

Regression analysis is a core tool in IB Mathematics that links statistics, algebra, and real-world applications. The key components are scatter diagrams, the least squares line, the correlation coefficient r, the coefficient of determination r², and residual analysis. Being able to use your GDC fluently and to interpret results in context are the two most valuable exam skills.

回归分析是IB数学中将统计、代数与实际应用紧密结合的核心工具。关键组成部分包括散点图、最小二乘回归线、相关系数 r、决定系数 r² 以及残差分析。熟练使用图形计算器并结合情境解释结果，是考试中最有价值的两项技能。

Always remember: correlation does not imply causation, extrapolation is dangerous, and a well-interpreted answer earns more marks than a merely calculated one.

请时刻记住：相关不等同于因果，外推需要谨慎，一个解释到位的答案比单纯计算的结果更能得分。

Published by TutorHao | Mathematics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

IB Mathematics: Foundations and Key Points of Regression Analysis | IB数学：回归分析基础与要点