📚 Linear Modelling in IB Mathematics | IB数学中的线性建模
Linear modelling sits at the heart of the IB Mathematics: Applications and Interpretation syllabus, bridging real-world data with algebraic simplicity. It equips learners to identify trends, quantify relationships, and make justified predictions using straight-line equations. Whether you are analysing the correlation between study hours and exam scores or modelling temperature change over time, a firm grasp of linear models is essential for both internal assessments and final examinations.
线性建模是IB数学：应用与解释课程的核心内容，它将现实数据与代数的简洁性连接起来。它帮助学习者识别趋势、量化关系，并使用直线方程做出合理的预测。无论你是在分析学习时长与考试成绩之间的关联，还是模拟温度随时间的变化，掌握线性模型对于内部评估和最终考试都是至关重要的。
1. What Is Linear Modelling? | 什么是线性建模？
onal to changes in the other, resulting in a mathematical equation of the form y = mx + c.
线性建模是为一组双变量数据拟合一条直线的过程，用以描述、解释和预测两个变量之间的关系。该模型假定一个变量的变化与另一个变量的变化成比例，从而得到形如 y = mx + c 的数学方程。
In IB mathematics, linear models are explored both graphically and algebraically. Students learn to construct them by eye, using technology, and through rigorous least-squares regression. The emphasis is not only on computation but also on meaningful interpretation of slope, intercept, and the strength of the relationship.
在IB数学中，线性模型从图形和代数两个角度来探索。学生要学会通过目测、使用技术工具以及严格的最小二乘回归来构建模型。重点不仅在于计算，还在于有意义地解释斜率、截距以及关系的强度。
2. The Straight Line Equation: y = mx + c | 直线方程：y = mx + c
The backbone of any linear model is the slope– of y when x = 0). In statistics-focused settings, you may also see the equation written as ŷ = a + bx, where b is the slope and a is the intercept.
任何线性模型的支柱都是斜截式：y = mx + c，其中 m 表示斜率（变化率），c 给出 y 轴截距（x = 0 时的 y 值）。在以统计为重点的语境中，你也会看到方程写为 ŷ = a + bx，其中 b 是斜率，a 是截距。
+ 200, then the slope 25 means that for every additional degree Celsius, sales increase by $25 on average; the intercept 200 suggests that at 0 °C, sales are expected to be $200.
对这些参数的解释是关键。例如，如果连接冰淇淋销售额（S 美元）和温度（t °C）的模型为 S = 25t + 200，那么斜率 25 表示平均每升高1摄氏度，销售额增加25美元；截距 200 表明在0 °C时，预期销售额为200美元。
3. Slope and Intercept in Context | 斜率与截距的实际含义
Slope m describes the estimated change in the response (dependent) variable for a one-unit increase in the predictor (independent) variable. A positive slope implies a positive association; a negative slope indicates an inverse relationship. A slope of zero suggests no linear association at all.
斜率 m 描述了当预测变量（自变量）增加一个单位时，响应变量（因变量）的估计变化量。正斜率意味着正相关；负斜率表示负相关；零斜率意味着完全没有线性关联。
The intercept c is the predicted value of the response when the predictor equals zero. It may or may not have a sensible real-world interpretation, especially if zero lies far outside the observed data range. In such cases, the intercept is merely an anchor for the line.
截距 c 是当预测变量等于零时响应变量的预测值。它可能有也可能没有合理的现实意义，尤其是当零值远在观测数据范围之外时。在这种情况下，截距只是直线的一个定位点。
Always examine whether the signs and magnitudes of m and c make sense in the given scenario. An impossible negative length or an unrealistic boundary value may signal that the linear model is inappropriate beyond the data range.
始终要检查 m 和 c 的符号和大小在给定情境中是否合理。一个不可能出现的负长度或一个不切实际的边界值可能表明线性模型在数据范围之外是不合适的。
4. Scatter Plots and the Eye Test for Association | 散点图与关联性的目测检验
Before any numerical modelling, a scatter plot should be the first step. By plotting one variable against the other, you can visually assess the direction (positive or negative), form (linear or non-linear), and strength (tight or scattered) of the relationship.
在进行任何数值建模之前，散点图应该是第一步。通过将一个变量相对于另一个变量绘图，你可以直观地评估关系的方向（正或负）、形态（线性或非线性）和强度（紧密还是分散）。
If the points roughly follow a straight-line pattern, a linear model is suitable. Curved patterns, clusters, or fanning shapes suggest that linear modelling may be misleading. IB exam questions often ask you to judge whether a linear model is appropriate based on a scatter plot.
如果数据点大致沿着一条直线分布，那么线性模型是合适的。弯曲的形状、聚类或扇形发散则表明线性建模可能产生误导。IB考题经常会让你根据散点图判断线性模型是否合适。
5. Drawing a Line of Best Fit by Eye | 目测绘制最佳拟合线
When technology is not available, an approximate linear model can be drawn manually. The line of best fit by eye should pass through the "centre" of the data points, balancing the number of points above and below the line along its entire length.
当无法使用技术工具时，可以手动绘制近似的线性模型。目测的最佳拟合线应穿过数据点的“中心”，在其整个长度上平衡线上方和下方的点数。
pt where the line crosses the y-axis. This method gives a rough but valuable model for simple predictions.
在这条手绘直线上，选取两个相距较远的点（不一定是数据点），计算斜率 m = Δy / Δx。读取直线与 y 轴相交处的截距。这种方法能为简单预测提供一个粗略但有价值的模型。
6. Least Squares Regression: The Best Mathematical Fit | 最小二乘回归：最佳的数学拟合
The most widely used objective method for linear modelling is least squares regression. It finds the unique line that minimises the sum of squared vertical distances (residuals) from the data points to the line. This line is often called the regression line of y on x.
最广泛使用的客观线性建模方法是最小二乘回归。它找到唯一一条直线，使数据点到直线的竖直距离（残差）的平方和最小。这条直线通常称为 y 对 x 的回归线。
For a set of n paired values (x₁, y₁), (x₂, y₂), …, (xₙ, yₙ), the regression line ŷ = a + bx has slope b = Sₓy / Sₓₓ, where Sₓy = Σ(xᵢ – x̄)(yᵢ – ȳ) and Sₓₓ = Σ(xᵢ – x̄)². The intercept a = ȳ – b x̄.
对于一组 n 对数值 (x₁, y₁)，(x₂, y₂)，…，(xₙ, yₙ)，回归线 ŷ = a + bx 的斜率 b = Sₓy / Sₓₓ，其中 Sₓy = Σ(xᵢ – x̄)(yᵢ – ȳ)，Sₓₓ = Σ(xᵢ – x̄)²。截距 a = ȳ – b x̄。
These formulas ensure that the line passes through the mean point (x̄, ȳ). In IB examinations, you may be expected to compute a and b from summary statistics using a calculator or, for small datasets, directly from the formulas.
这些公式确保直线通过均值点 (x̄, ȳ)。在IB考试中，你可能需要根据摘要统计量使用计算器计算 a 和 b，或者对于小数据集，直接根据公式计算。
7. Pearson’s Correlation Coefficient r | 皮尔逊相关系数 r
by Pearson’s product-moment correlation coefficient, denoted by r. It always lies between –1 and 1, where r = 1 indicates perfect positive linear correlation, r = –1 indicates perfect negative linear correlation, and r = 0 suggests no linear correlation.
线性关系的强度和方向由皮尔逊积矩相关系数衡量，记作 r。它的取值总是在 –1 和 1 之间，其中 r = 1 表示完全正线性相关，r = –1 表示完全负线性相关，r = 0 表明没有线性相关。
The formula for r is r = Sₓy / √Sₓₓ × Syy, where Syy = Σ(yᵢ – ȳ practice, most IB students obtain r directly from their GDC (graphic display calculator).
r 的公式为 r = Sₓy / √Sₓₓ × Syy，其中 Syy = Σ(yᵢ – ȳ)²。注意 r 与斜率 b 符号相同。在实践中，大多数IB学生直接从图形计算器获取 r。
Interpretation must be nuanced. An r close to zero does not necessarily mean there is no relationship; it only means there is no linear relationship. A curved pattern could still have a very strong non-linear association.
解释必须细致入微。r 接近零并不一定意味着没有关系；它只表示没有线性关系。一个弯曲的形态仍然可能具有很强的非线性关联。
8. The Coefficient of Determination r² | 决定系数 r²
While r measures strength, the coefficient of determination r² tells us the proportion of variation in the dependent variable that is accounted for by the linear model with the independent variable. An r² of 0.85 means that 85% of the variability in y can be explained by the variation in x.
虽然 r 衡量强度，但决定系数 r² 告诉我们因变量的变化中有多大比例可以被包含自变量的线性模型所解释。r² 为 0.85 意味着 y 中 85% 的变异性可以被 x 的变化所解释。
This interpretation is immensely useful for comparing models. A higher r² suggests that the regression line fits the observed data more closely. However, a high r² does not guarantee the model will predict well outside the collected range.
这种解释对于比较模型非常有用。较高的 r² 表明回归线对观测数据的拟合更紧密。然而，高 r² 并不能保证模型在收集范围之外的良好预测效果。
9. Residuals: The Errors We Make | 残差：我们所犯的误差
between an observed y-value and the value predicted by the regression line: residual = yᵢ – ŷ ssing model adequacy. The least squares line makes the sum of squared residuals a minimum.
残差是观测到的 y 值与回归线预测值之间的竖直差：残差 = yᵢ – ŷᵢ。残差是评估模型适宜性的基础。最小二乘直线使残差平方和最小。
Plotting residuals against the x-values (or against predicted values) helps check whether a linear model is appropriate. Ideally, residuals should be randomly scattered around zero with no obvious pattern. A curved pattern suggests the relationship may be non-linear; a fan shape indicates non-constant variance.
将残差相对于 x 值（或预测值）作图有助于检查线性模型是否合适。理想情况下，残差应随机散布在零附近，没有明显的模式。弯曲的模式表明关系可能是非线性的；扇形发散则表明方差不是常数。
10. Using Linear Models for Interpolation and Extrapolation | 使用线性模型进行内插与外推
Once a reliable linear model has been established, it can be used for prediction. Interpolation involves estimating values within the observed x-range; it is generally considered safe if the model fits well. Extrapolation predicts beyond the data range and carries far greater risk.
一旦建立了可靠的线性模型，就可以用它进行预测。内插是指在观测到的 x 范围内进行估计；若模型拟合良好，通常被认为是安全的。外推则预测数据范围之外的值，风险要大得多。
Extrapolation assumes that the linear trend continues indefinitely, which may be unrealistic. Many physical and social phenomena level off or change direction outside the measured interval. Always treat extrapolated predictions with caution and comment on their reliability.
外推假设线性趋势无限延续，这可能不现实。许多物理和社会现象在测量区间之外会趋于平稳或改变方向。对待外推预测始终要保持谨慎，并对其可靠性加以评论。
11. Technology Tips: GDC and Spreadsheet Modelling | 技术技巧：图形计算器与电子表格建模
IB candidates are expected to use a graphic display calculator efficiently. Enter bivariate data into lists, produce a scatter plot, run linear regression (LinReg), and record the equation together with r and r². Knowing how to store the regression equation and generate a residual plot is essential.
IB考生需要高效使用图形计算器。将双变量数据输入列表，生成散点图，运行线性回归（LinReg），并记录方程以及 r 和 r²。了解如何存储回归方程并生成残差图是必不可少的。
For internal assessment tasks, spreadsheet tools such as Excel or Google Sheets can also be used. They provide regression output, residual analysis, and clear graphing options. Documenting technology steps and interpreting the output clearly earns high marks in criterion C and D.
对于内部评估任务，也可以使用电子表格工具，如Excel或Google Sheets。它们提供回归输出、残差分析以及清晰的作图选项。记录技术步骤并清晰地解读输出，可以在标准C和D中获得高分。
12. Common Pitfalls and Best Practices in Linear Modelling | 线性建模中的常见陷阱与最佳实践
A recurring mistake is confusing correlation with causation. Even a very high r does not prove that changes in x cause changes in y. There may be a lurking variable or simply coincidence. Always phrase conclusions carefully and consider the broader context.
一个反复出现的错误是混淆相关与因果。即使很高的 r 也不能证明 x 的变化会导致 y 的变化。可能存在潜在变量或仅仅是巧合。始终要仔细措辞结论，并考虑更广泛的背景。
Another trap is relying on r² without checking the residual plot. A high r² can occur with a curved pattern if the points happen to cluster near a straight line at two ends. The residual plot reveals such hidden curvature. Additionally, avoid using the regression equation if the scatter plot shows no clear linear trend.
另一个陷阱是依赖 r² 而不检查残差图。如果数据点恰好在两端聚集在一条直线附近，即使存在弯曲模式，也可能出现高 r²。残差图能揭示这种隐藏的曲率。此外，如果散点图没有显示出清晰的线性趋势，就避免使用回归方程。
Finally, always round coefficients to a sensible degree of accuracy—usually three significant figures—and label units when interpreting slope and intercept. Clear communication of mathematical thinking is the hallmark of a top-scoring IB student.
最后，始终要将系数四舍五入到合理的精度——通常是三位有效数字——并在解释斜率和截距时标注单位。清晰地交流数学思想是高分IB学生的标志。
Published by TutorHao | 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