📚 Correlation in Biology | 生物学中的相关性
Correlation is a fundamental statistical tool used across the biological sciences to explore and quantify the relationship between two continuous variables. Whether you are investigating how enzyme activity changes with temperature, examining the link between smoking and lung cancer incidence, or testing whether leaf length is related to leaf width, correlation analysis provides a numerical and visual summary of the association. In A-level Biology, students are expected not only to interpret correlation coefficients from given data but also to critically evaluate the nature of the relationship, recognising that correlation does not automatically imply a causal link.
相关是生物科学中用于探索和量化两个连续变量之间关系的基本统计工具。无论你是在研究酶活性如何随温度变化，还是在检查吸烟与肺癌发病率之间的联系，或是在检验叶片长度是否与叶片宽度有关，相关分析都能为这种关联提供数值和视觉上的概括。在 A-level 生物学中，学生不仅要学会解读给定数据中的相关系数，还要能够批判性地评价这种关系的性质，认识到相关性并不自动意味着因果联系。
1. What is Correlation? | 什么是相关？
Correlation refers to a statistical measure that describes the extent to which two variables change together. When one variable tends to increase as the other increases, we speak of a positive association; when one increases while the other decreases, that is a negative association. If the variables move independently, there is no correlation.
相关是一种统计量度，用来描述两个变量共同变化的程度。当一个变量随着另一个变量的增加而倾向于增加时，我们称其为正相关；当一个增加而另一个减少时，则是负相关。如果变量独立变化，则无相关。
In biology, correlation helps researchers detect patterns in experimental and observational data. For example, a positive correlation might exist between the concentration of a substrate and the initial rate of an enzyme-catalysed reaction up to a saturation point. However, the presence of a correlation does not, on its own, explain the biological mechanism.
在生物学中，相关帮助研究人员发现实验和观察数据中的模式。例如，在达到饱和点之前，底物浓度与酶催化反应的初始速率之间可能存在正相关。然而，相关性的存在本身并不能解释背后的生物学机制。
The most common type of correlation assessed at A-level is linear correlation, which assumes that when plotted on a graph, the points roughly follow a straight line. Non-linear relationships, such as exponential or sigmoidal curves, can exhibit very low correlation coefficients even though a strong functional link exists.
在 A-level 阶段评估的最常见相关类型是线性相关，它假定在图上描点时，数据点大致沿一条直线分布。非线性关系，如指数或S形曲线，即使存在很强的函数联系，也可能表现出很低的相关系数。
2. The Pearson Correlation Coefficient (r) | 皮尔逊相关系数 (r)
The Pearson product-moment correlation coefficient, denoted by the symbol r, is the standard measure of the strength and direction of a linear relationship between two variables measured on an interval or ratio scale. It is calculated from the covariance of the two variables divided by the product of their standard deviations, giving a dimensionless number.
皮尔逊积矩相关系数，用符号 r 表示，是测量两个在间隔或比率尺度上度量的变量之间线性关系强度与方向的标准方法。它由两个变量的协方差除以它们标准差的乘积计算得出，得到一个无量纲的数字。
When computed from a sample, r estimates the population correlation coefficient ρ (rho). For A-level Biology purposes, you will normally be given the value of r or asked to interpret it from a scatter graph; you are unlikely to be required to compute r from raw data in the exam, but understanding the formula can aid interpretation.
当从样本中计算时，r 估计总体相关系数 ρ。就 A-level 生物学而言，你通常会被给出 r 的值，或要求从散点图中解读它；你在考试中不太可能需要从原始数据计算 r，但理解公式有助于解释。
The formula for r is:
r 的计算公式为：
r = Σ((xi – x)(yi – y)) / √[Σ(xi – x)2 Σ(yi – y)2]
3. Range of r and Interpretation | r 的范围与解释
The correlation coefficient r always lies between –1 and +1. A value of +1 indicates a perfect positive linear correlation: all data points lie exactly on a straight line with a positive slope. A value of –1 indicates a perfect negative linear correlation: the points lie on a line with a negative slope. An r of 0 suggests no linear correlation, meaning the points show no tendency to follow a straight line.
相关系数 r 始终介于 –1 到 +1 之间。+1 表示完全正线性相关：所有数据点恰好落在一条具有正斜率的直线上。–1 表示完全负线性相关：点落在一条负斜率的直线上。r 为 0 表明没有线性相关，意味着这些点没有沿直线分布的趋势。
The following table summarises the conventional interpretation of r magnitudes in biological contexts:
下表总结了在生物学情境中对 r 大小的一般解释：
| r value | Interpretation | r 值 | 解释 |
| +0.9 to +1.0 | Strong positive correlation | +0.9 到 +1.0 | 强正相关 |
| +0.6 to +0.89 | Moderate positive correlation | +0.6 到 +0.89 | 中等正相关 |
| 0 to +0.59 | Weak positive correlation | 0 到 +0.59 | 弱正相关 |
| 0 | No linear correlation | 0 | 无线性相关 |
| –0.01 to –0.59 | Weak negative correlation | –0.01 到 –0.59 | 弱负相关 |
| –0.6 to –0.89 | Moderate negative correlation | –0.6 到 –0.89 | 中等负相关 |
| –0.9 to –1.0 | Strong negative correlation | –0.9 到 –1.0 | 强负相关 |
It is crucial to note that the boundaries between ‘weak’, ‘moderate’ and ‘strong’ are arbitrary and can depend on the scientific context. In field ecology, an r of 0.5 might be viewed as quite meaningful, whereas in a tightly controlled biochemical assay, researchers may expect r > 0.95.
需要注意的是，“弱”、“中等”和“强”之间的界限是主观的，并且可能取决于科学背景。在野外生态学中，r 值为 0.5 可能被视为很有意义，而在严格控制的生化测定中，研究人员可能期望 r > 0.95。
4. Positive, Negative and Zero Correlation | 正相关、负相关和零相关
Positive correlation means that high values of one variable are associated with high values of the other, and low with low. In a scatter diagram, this appears as an upward trend. A biological example is the relationship between stomatal density and transpiration rate, where, under comparable conditions, leaves with more stomata tend to lose water faster.
正相关意味着一个变量的高值与另一个变量的高值相关联，低值与低值相关联。在散点图中，这表现为上升趋势。一个生物学实例是气孔密度与蒸腾速率之间的关系：在可比条件下，气孔更多的叶片往往失水更快。
Negative correlation occurs when high values of one variable pair with low values of the other. The scatter plot slopes downwards. An example might be the link between ambient light intensity and cockroach activity: as light intensity increases, activity often decreases in nocturnal species.
负相关发生在一个变量的高值与另一个变量的低值成对出现时。散点图向下倾斜。例如，环境光强度与蟑螂活动之间的联系：随着光强度增加，夜行性物种的活动往往减少。
Zero correlation means there is no discernible linear pattern; the data points form a roughly circular cloud. However, zero correlation does not imply that the variables are unrelated – they may have a strong non-linear relationship, such as the inverted U-shaped curve of an enzyme’s activity with temperature beyond its optimum.
零相关意味着不存在明显的线性模式；数据点形成大致圆形的散点云。然而，零相关并不意味着变量没有关系——它们可能具有强烈的非线性关系，比如酶活性与温度在超过其最适温度后的倒 U 形曲线。
5. Scatter Diagrams: Visualising Correlation | 散点图：可视化相关性
Before calculating a correlation coefficient, a scatter diagram (scatter plot) should always be drawn. It displays each data pair (x, y) as a point on a Cartesian plane. The pattern of points reveals whether a linear relationship is plausible and can help identify outliers or clusters that might distort r.
在计算相关系数之前，应始终先绘制散点图。它将每个数据对 (x, y) 显示为笛卡尔平面上的一个点。点的分布模式可以揭示线性关系是否合理，并有助于识别可能扭曲 r 的异常值或数据聚类。
When interpreting a scatter diagram, look for the overall trend. If the points are tightly clustered around a straight line, correlation is strong; if they are widely scattered, it is weak. A plot that curves suggests that Pearson’s r may be misleading, and a different statistical approach or data transformation might be needed.
在解读散点图时，要观察整体趋势。如果点紧密聚集在一条直线附近，相关性就强；如果点广泛散布，相关性就弱。弯曲的图形提示皮尔逊 r 可能产生误导，可能需要不同的统计方法或数据转换。
In your A-level Biology practical write-ups, always include a scatter graph when discussing a possible correlation between two measured variables. You may also sketch a line of best fit (either by eye or using a regression line) to illustrate the direction and approximate strength of the association.
在你的 A-level 生物学实验报告中，当讨论两个测量变量之间可能存在相关时，始终应包含散点图。你还可以画出最佳拟合线（通过目测或使用回归线）来说明关联的方向和大致强度。
6. Strength of Correlation: From Weak to Strong | 相关强度：从弱到强
The absolute value of r (ignoring the sign) indicates strength. An r of 0.2 represents a weak correlation; changes in one variable are only loosely related to changes in the other. An r of 0.9 signals a very strong linear association, with one variable predicting the other with high accuracy.
r 的绝对值（忽略符号）表示强度。r = 0.2 代表弱相关，一个变量的变化与另一个变量的变化只有松散的联系。r = 0.9 表明非常强的线性关联，一个变量可以高精度地预测另一个变量。
It is important to remember that the strength of a correlation also depends on sample size. A moderate r of 0.5 calculated from 8 data points may not be statistically significant, whereas the same r from 100 data points almost certainly is. Always consider the number of replicates when drawing conclusions.
重要的是要记住，相关的强度也取决于样本量。从 8 个数据点计算出的中等 r 值 0.5 可能不具有统计显著性，而从 100 个数据点计算出的相同 r 值几乎肯定是显著的。在得出结论时，始终要考虑重复的数量。
In addition, the coefficient of determination, r², indicates the proportion of variation in y that can be explained by variation in x. For instance, if r = 0.8, then r² = 0.64, meaning 64% of the variance in y is accounted for by its linear relationship with x. The remaining 36% is due to other factors or random variation.
此外，决定系数 r² 表示 y 的变异中能被 x 的变异解释的比例。例如，如果 r = 0.8，那么 r² = 0.64，意味着 y 的方差中有 64% 可由其与 x 的线性关系解释。其余 36% 归因于其他因素或随机变异。
7. Calculating r: A Brief Overview | 计算 r：概览
While you are not required to memorise the full Pearson formula for the A-level exam, it is helpful to understand the logic behind it. The formula compares how much each point deviates from the mean of x and the mean of y. When x and y consistently deviate in the same direction, the product of the deviations is positive, leading to a positive r. When they deviate in opposite directions, the product is negative.
虽然 A-level 考试不要求你记住完整的皮尔逊公式，但理解其背后的逻辑很有帮助。该公式比较每个点偏离 x 平均值和 y 平均值的程度。当 x 和 y 始终朝同一方向偏离时，偏差的乘积为正，从而得到正的 r。当它们朝相反方向偏离时，乘积为负。
Calculation involves summing the products of paired deviations and dividing by the square root of the product of the sums of squared deviations. Most biologists use statistical software or a spreadsheet to compute r; in the laboratory, a calculator with two-variable statistics mode can be used on prepared data sets.
计算涉及求配对偏差乘积的总和，再除以偏差平方和乘积的平方根。大多数生物学家使用统计软件或电子表格来计算 r；在实验室中，可以使用具有双变量统计模式的计算器处理准备好的数据集。
8. Null Hypothesis and Statistical Significance | 零假设与统计显著性
In biological research, finding a non-zero r in a sample is not enough to prove that a true correlation exists in the population. The null hypothesis (H₀) states that the population correlation coefficient ρ is zero – that is, there is no linear relationship. The alternative hypothesis (H₁) can be two-sided (ρ ≠ 0) or one-sided (ρ > 0 or ρ < 0), depending on the prediction.
在生物学研究中，在样本中发现一个非零的 r 值并不足以证明总体中存在真实的线性相关。零假设 (H₀) 假定总体相关系数 ρ 为零——即不存在线性关系。备择假设 (H₁) 可以是双侧的 (ρ ≠ 0) 或单侧的 (ρ > 0 或 ρ < 0)，具体取决于预测。
To test significance, the calculated r is compared with a critical value obtained from a Spearman/Pearson table at a given significance level (usually α = 0.05) and n – 2 degrees of freedom. If |r| exceeds the critical value, we reject H₀ and conclude that the correlation is statistically significant. This means the observed association is unlikely to have arisen by random chance alone.
为了检验显著性，将计算出的 r 与在给定显著性水平（通常 α = 0.05）和自由度 n – 2 下从斯皮尔曼/皮尔逊表中获得的临界值进行比较。如果 |r| 超过临界值，我们就拒绝 H₀，并得出结论说该相关具有统计显著性。这意味着观察到的关联不太可能仅由随机偶然产生。
9. Correlation Does Not Equal Causation | 相关不等于因果
Perhaps the most important lesson in interpreting correlation is that a strong r, even if highly significant, does not prove that x causes y or that y causes x. Two variables may be correlated because both are influenced by a common third factor, or the association might be entirely coincidental.
在解读相关性时，最重要的一课或许就是：即便 r 很强且高度显著，也不能证明 x 导致 y，或者 y 导致 x。两个变量可能相关，是因为它们都受到同一个第三因素的影响，或者这种关联可能完全是巧合。
A classic biological example is the positive correlation between the number of storks and the human birth rate in some regions. The underlying variable is the size of the rural area: larger areas support more storks and also have more people, leading to a spurious correlation. To establish causation, controlled experiments, manipulation of the independent variable, and evidence of a plausible mechanism are required.
一个经典的生物学例子是某些地区鹳的数量与人类出生率之间的正相关。潜在的变量是乡村地区的面积：更大的地区能养活更多的鹳，也有更多的人口，从而导致了虚假相关。要确立因果关系，需要对照实验、对自变量的操控以及可信机制的证据。
10. Confounding Variables: The Hidden Third Factor | 混淆变量：隐藏的第三因素
A confounding variable is an extraneous variable that correlates with both the independent and dependent variables, creating a false impression of a direct relationship. In ecological field studies, confounding variables are very common because many environmental factors covary.
混淆变量是一个外来变量，它与自变量和因变量都相关，从而造成两者之间直接关系的假象。在生态学野外研究中，混淆变量非常普遍，因为许多环境因子是共同变化的。
For instance, a student measuring the effect of light intensity on photosynthesis might find a positive correlation between distance from a lamp and the rate of oxygen production. However, lamp distance also changes temperature, a known confounding variable. Unless temperature is kept constant, any observed correlation cannot be solely attributed to light intensity.
例如，一位测量光强度对光合作用影响的学生可能会发现灯的距离与氧气产生速率之间存在正相关。然而，灯的距离也会改变温度，这是一个已知的混淆变量。除非温度保持不变，否则观察到的任何相关性都不能仅归因于光强度。
When planning an investigation, biologists try to identify potential confounding variables in advance and control them through standardisation, blocking, or randomisation. In the analysis phase, techniques such as partial correlation can mathematically remove the influence of a confounding variable.
在设计研究时，生物学家会尝试提前识别潜在的混淆变量，并通过标准化、区组化或随机化来控制它们。在分析阶段，诸子偏相关等技术可以在数学上去除混淆变量的影响。
11. Biological Examples of Correlation | 生物学中的相关实例
Correlation analyses appear frequently in A-level Biology investigations. A few typical and safe examples include: (i) the length and width of leaves from the same plant species – often show a strong positive linear correlation; (ii) body mass and resting heart rate in mammals – a negative correlation is often observed; and (iii) the dissolved oxygen concentration in a pond and the number of mayfly nymphs – may show a positive correlation because both are indicators of water quality.
相关分析经常出现在 A-level 生物学研究中。几个典型且稳妥的例子包括：(i) 同种植物叶片的长度和宽度——通常表现出较强的正线性相关；(ii) 哺乳动物的体重和静息心率——常观察到负相关；(iii) 池塘中的溶解氧浓度与蜉蝣稚虫的数量——因为两者都是水质的指标，可能呈现出正相关。
In human biology, the correlation between tar content of cigarettes and incidence of lung cancer has been extensively documented. While the correlation is very strong, proving causation required decades of epidemiological studies, animal experiments and biochemical identification of carcinogens.
在人类生物学中，香烟的焦油含量与肺癌发病率之间的相关性已有广泛记录。虽然该相关性非常强，但要证明因果关系需要数十年的流行病学研究、动物实验和致癌物的生化鉴定。
An important application of correlation in genetics is in the study of heritability: the correlation between the phenotype of parents and offspring or between twins gives an estimate of the genetic contribution to a trait. However, these correlations can be inflated by shared environment, a classic confounding factor.
相关在遗传学中的一个重要应用是遗传力的研究：亲本与后代或双胞胎之间表型的相关给出了性状遗传贡献的估计。然而，这些相关可能会因共同环境（一个经典的混淆因素）而被夸大。
12. Common Pitfalls and Summary | 常见误区与总结
Students frequently misinterpret correlation in the following ways: (a) assuming that a high r guarantees a causal link; (b) ignoring non-linear patterns when they only look
Published by TutorHao | A-Level Biology 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Cancel reply