Experimental Design in Biology: Defining Variables Clearly | 生物实验设计：如何明确实验变量

📚 Experimental Design in Biology: Defining Variables Clearly | 生物实验设计：如何明确实验变量

Biology is an empirical science: every conclusion about living systems must be supported by carefully designed experiments. The single most important step in designing a meaningful experiment is to decide, before touching a pipette or a microscope, exactly what you are changing, what you are measuring, and what you are keeping constant. These are the variables of your experiment.

生物学是一门以实验为基础的学科：关于生命系统的每一个结论都必须由精心设计的实验来支撑。而设计一个有意义的实验，最重要的一步就是在动手使用移液管或显微镜之前，先明确你要改变什么、测量什么、保持什么不变。这些要素就是实验中的变量。

1. What Is a Variable? | 什么是变量？

A variable is any factor, trait, or condition that can exist in differing amounts or types in an experiment. In biological experiments, variables can be continuous (e.g., temperature in °C), discrete (e.g., number of offspring), or categorical (e.g., genotype: wild type versus mutant). Understanding the type of variable you are working with guides both your experimental design and your choice of statistical test.

变量是指在实验中可以存在不同数量或类型的任何因素、特征或条件。在生物实验中，变量可以是连续型的（如以℃为单位的温度）、离散型的（如后代数量），也可以是类别型的（如基因型：野生型相对于突变型）。理解变量的类型，既有助于设计实验，也有助于选择合适的统计检验方法。

Continuous variable: can take any numerical value within a range, e.g., height or reaction rate.
Discrete variable: takes countable integer values, e.g., number of colonies on an agar plate.
Categorical variable: belongs to a set of categories, e.g., species or colour.

连续变量：在一定范围内可取任意数值，例如高度或反应速率。
离散变量：取可数的整数值，例如琼脂平板上的菌落数。
类别变量：属于某一组类别，例如物种或颜色。

2. The Three Core Variables | 三类核心变量

Every well-designed biological experiment contains three core variables: the independent variable, the dependent variable, and controlled variables. These three work together to reveal cause-and-effect relationships. If you fail to identify all three clearly, your results may be meaningless or impossible to interpret.

每一个设计良好的生物实验都包含三类核心变量：自变量（独立变量）、因变量（依赖变量）和控制变量。三者协同作用，才能揭示因果关系。如果不能清楚区分这三者，实验结果可能毫无意义或难以解释。

Variable type	Definition	Example
Independent	The factor deliberately altered by the researcher	Light intensity (lux)
Dependent	The factor measured to see the effect	Rate of oxygen production (mL/min)
Controlled	Factors kept constant for all groups	Temperature, CO₂ concentration, plant species

变量类型	定义	例子
自变量	研究者有意改变的因素	光照强度（勒克斯）
因变量	为观察效果而测量的因素	产氧速率（mL/min）
控制变量	所有组均保持恒定的因素	温度、CO₂浓度、植物种类

3. Defining the Independent Variable Operationally | 操作性地定义自变量

The independent variable is what the experimenter intentionally changes between groups. However, a vague description such as “more light” is not scientific. You must provide an operational definition, which states exactly how the variable is set, manipulated, and recorded. For example, “light intensity” can be operationally defined as the distance between a lamp and the plant, converted to lux using the inverse square law.

自变量是实验者有意在不同组之间改变的因素。然而，像“更多光照”这种模糊描述并不科学。你必须给出一个操作性定义，即明确说明该变量如何设置、如何操作以及如何记录。例如，“光照强度”可以操作性地定义为灯与植物之间的距离，并根据平方反比定律换算成勒克斯。

Illuminance = light source intensity × (1 / distance²)

Operational definitions also require clear units and specific levels. Instead of “high, medium, low” temperature, use 15 °C, 25 °C, and 35 °C. This allows other scientists to replicate your experiment exactly and allows you to construct a quantitative graph.

操作性定义还要求明确的单位和具体水平。不要使用“高、中、低”温度，而应使用15°C、25°C和35°C。这既能让其他科学家精确重复你的实验，也能让你绘制出定量关系图。

4. Choosing a Valid Dependent Variable | 选择合适的因变量

The dependent variable is the response you measure. It must be directly linked to the biological process you are studying, and it must be measured objectively. For instance, if you want to test the effect of temperature on enzyme activity, the dependent variable might be the rate of product formation (mg of product per minute), not merely a colour change described as “more pink.”

因变量是你测量的反应。它必须与你所研究的生物学过程直接相关，并且必须被客观地测量。例如，如果我们要检验温度对酶活性的影响，因变量可以是产物生成的速率（每分钟产物的毫克数），而不能仅仅是“更粉红色”这样的颜色变化描述。

Validity: Does the measurement reflect the biological process?
Precision: Can the measurement be taken with sufficient accuracy?
Repeatability: Would repeated measurements give similar values?

有效性：测量是否真的反映生物学过程？
精确性：测量能否达到足够的准确度？
可重复性：重复测量是否会得到相近的值？

Whenever possible, choose a quantitative dependent variable. Data such as “mass change in grams” or “time in seconds” can be statistically analysed. Qualitative observations, such as “cells looked healthier,” should be recorded only as supporting evidence.

只要可能，应选择定量的因变量。像“以克为单位的质量变化”或“以秒为单位的时间”这类数据可以用于统计分析。而“细胞看起来更健康”这类定性观察只能作为辅助证据。

5. Controlled Variables: The Foundation of Fair Testing | 控制变量：公平实验的基础

Controlled variables are all the other factors that could influence the dependent variable if they were allowed to vary. In a biology experiment, these are often more numerous than the independent variable itself. For a plant growth experiment, controlled variables include soil volume, water volume, temperature, humidity, light quality, pot size, and initial seedling mass.

控制变量是指所有其他可能影响因变量、但如果不加限制就会发生变化的因素。在生物实验中，控制变量往往比自变量本身还要多。以植物生长实验为例，控制变量包括土壤体积、浇水量、温度、湿度、光质、花盆大小以及幼苗初始质量。

Why is controlling variables so important? If a control variable changes accidentally along with the independent variable, you cannot tell which variable caused the observed effect. This is called confounding, and it destroys the internal validity of the experiment.

为什么控制变量如此重要？因为如果一个控制变量随着自变量偶然地一起变化，你就无法判断究竟是哪个变量引起了观察到的效应。这就叫作混杂，它会破坏实验的内部效度。

Effect on dependent variable = effect of independent variable + effect of uncontrolled confounders

6. Identifying and Minimising Confounding Variables | 识别并减少混杂变量

A confounding variable is a factor that varies systematically with the independent variable and also affects the dependent variable. For example, in an experiment testing the effect of a new fertiliser on crop yield, if the plants with fertiliser also receive more sunlight because of their position in the greenhouse, then sunlight is a confounding variable.

混杂变量是指与自变量同步变化、并且同时影响因变量的因素。例如，在一个检验新型肥料对作物产量影响的实验中，如果施加肥料的植物因为位于温室中不同的位置而接受更多阳光，那么阳光就是一个混杂变量。

To minimise confounding variables, you can use the following strategies:

为了减少混杂变量，可以采用以下策略：

Standardisation: Keep all known conditions identical for all groups.
Randomisation: Assign subjects or treatments to positions/order randomly.
Blocking: Group subjects with similar characteristics (e.g., same age) and treat them together.
Blinding: Ensure that the person measuring the result does not know which treatment was applied.

标准化：使所有已知条件对所有组完全一致。
随机化：随机分配样本或处理的位置与顺序。
区组化：将具有相似特征的样本（如相同年龄）分成组并一起处理。
盲法：确保测量结果的人不知道样本接受的是哪种处理。

7. Positive and Negative Controls | 阳性对照与阴性对照

Controls are groups that help you confirm that your experimental system is working correctly. A negative control is a group in which no response is expected. If the negative control shows a response, then your measuring system is contaminated or flawed. A positive control is a group in which a known response is expected; it proves that the experiment is capable of detecting the effect.

对照组是帮助你确认实验系统正常工作的组别。阴性对照是预期不会出现反应的组。如果阴性对照出现了反应，就说明测量系统被污染或有缺陷。阳性对照是预期出现已知反应的组，它证明实验有能力检测到效应。

Type	Expected result	Interpretation if unexpected
Negative control	No response	Contamination or measurement error
Positive control	Known response	Experimental system may be unable to detect the effect

对照类型	预期结果	出现意外时的解释
阴性对照	无反应	污染或测量误差
阳性对照	已知反应	实验系统可能无法检测到该效应

For example, in a food test for glucose with Benedict’s reagent, distilled water is the negative control (should remain blue), while a known glucose solution is the positive control (should turn brick red).

例如，在用本尼迪特试剂检测葡萄糖的食物实验中，蒸馏水是阴性对照（应保持蓝色），而已知葡萄糖溶液是阳性对照（应变成砖红色）。

8. From Hypothesis to Testable Prediction | 从假设到可检验的预测

Before defining variables, you need a clear hypothesis. A hypothesis is a tentative explanation for a biological phenomenon. It must be falsifiable: there must be an experimental result that could prove it wrong. From the hypothesis, you can derive a specific prediction that links the independent variable to the dependent variable.

在定义变量之前，你需要一个清晰的假设。假设是对某一生物学现象的尝试性解释，而且必须是可以被证伪的，也就是说，必须存在一个可能证明它错误的结果。从假设出发，你可以推导出一个将自变量与因变量联系起来的特定预测。

Weak hypothesis: “Light affects plant growth.”

较弱的假设：“光照影响植物生长。”

Strong hypothesis: “Increasing light intensity from 200 lux to 600 lux will increase the rate of oxygen production in Elodea by at least 50%.”

较强的假设：“将光照强度从200勒克斯增加到600勒克斯，会使伊乐藻的产氧速率至少提高50%。”

A strong prediction specifies the direction and sometimes the magnitude of the expected change, making it clear what data would support or reject the hypothesis.

一个强有力的预测会指明预期变化的方向，有时还包括大小，从而让数据清楚地表明结果是支持还是否定假设。

9. Sample Size and Replication | 样本量与重复

Replication means repeating the same treatment on multiple biological subjects or multiple samples. It is not the same as taking repeated measurements of the same subject. For example, measuring the same leaf ten times only tells you about measurement error; using ten different leaves tells you about biological variation.

重复是指在多个生物学个体或多个样本上进行相同的处理。它不等于对同一个体进行重复测量。例如，同一片叶子测量十次只能反映测量误差；而使用十片不同的叶子才能反映生物变异。

Larger sample sizes reduce the impact of individual variation and increase the reliability of statistical conclusions. However, in real biological research and in school practicals, sample size is limited by time, cost, and ethics. You should always choose the largest sample size that is practical while ensuring that all samples receive identical treatment apart from the independent variable.

更大的样本量可以降低个体变异的影响，提高统计结论的可靠性。然而，在真实的生物学研究和学校实验中，样本量受到时间、成本和伦理的限制。在保证除自变量外所有样本都接受相同处理的前提下，你应该选择实际可行范围内最大的样本量。

n = 1 is never enough; n ≥ 3 is a minimum; n ≥ 5 or more is preferred for statistical analysis.

10. Randomisation and Blinding | 随机化与盲法

Randomisation involves assigning subjects or experimental units to treatment groups purely by chance. This prevents bias, such as placing the healthiest plants in the group receiving the new fertiliser. Randomisation distributes unknown and unavoidable variation evenly across groups.

随机化是指完全随机地将样本或实验单元分配到不同处理组中。这可以防止偏差，例如把最健康的植物都放到施加新肥料的那一组。随机化会把未知且无法避免的变异均匀地分布到各组中。

Blinding ensures that the person measuring the dependent variable does not know which treatment each subject received. In a drug trial, the researcher measuring blood pressure should not know whether a patient received the drug or a placebo. Blinding prevents conscious or unconscious bias in data collection and interpretation.

盲法确保测量因变量的人不知道每个样本接受的是哪种处理。在药物试验中，测量血压的研究者不应该知道患者服用的是药物还是安慰剂。盲法可以防止数据收集和解释中出现有意或无意的偏差。

11. A Complete Example: Investigating Enzyme Activity | 完整示例：探究酶活性

Let us apply these principles to a classic experiment: the effect of pH on the activity of catalase. Catalase breaks down hydrogen peroxide into water and oxygen. The experimenter needs to define all variables before starting.

让我们把这些原则应用于一个经典实验：pH对过氧化氢酶活性的影响。过氧化氢酶将过氧化氢分解为水和氧气。实验者需要在开始之前定义所有变量。

Independent variable: pH of the buffer solution, using pH 4, 6, 7, 8, and 10.
Dependent variable: initial rate of oxygen gas production, measured as volume of gas collected in the first 30 seconds (mL/min).
Controlled variables: temperature (25 °C), enzyme concentration (1 mg/mL), substrate concentration (2% hydrogen peroxide), total volume of reaction mixture (10 mL), and the type of enzyme source (same batch of catalase solution).
Negative control: buffer without enzyme; no oxygen should be produced.
Positive control: catalase at its known optimum pH (pH 7); a high rate of oxygen production is expected.

自变量：缓冲液的pH值，设置为pH 4、6、7、8和10。
因变量：氧气产生的初始速率，以前30秒内收集到气体的体积（mL/min）表示。
控制变量：温度（25°C）、酶浓度（1 mg/mL）、底物浓度（2%过氧化氢）、反应混合物体积（10 mL）以及酶来源类型（同一批过氧化氢酶溶液）。
阴性对照：不加酶的缓冲液；预计不会产生氧气。
阳性对照：在已知最适pH（pH 7）下的过氧化氢酶；预期产生大量氧气。

Each pH treatment must be performed in at least three replicate trials. The order of trials should be randomised, and the person recording gas volume should be blinded to the pH condition. Only then can the results be used to construct a graph of reaction rate versus pH.

每种pH处理都必须至少进行三次重复试验。试验顺序应随机化，记录气体体积的人应对pH条件保持盲法。只有这样，结果才能用于绘制反应速率相对于pH的曲线图。

12. Common Mistakes and How to Avoid Them | 常见错误及避免方法

Students often lose marks in practical exams because of simple variable errors. The most common mistake is confusing the independent and dependent variables. Another frequent issue is failing to state controlled variables in enough detail, for example writing “same conditions” instead of listing temperature, time, concentration, and volume.

在实验考试中，学生常常因为简单的变量错误而失分。最常见的错误是混淆自变量和因变量。另一个常见问题是控制变量写得不够具体，例如只写“相同条件”，而没有列出温度、时间、浓度和体积。

Common mistake	Why it is a problem	Better approach
Changing two variables at once	Cannot determine cause and effect	Change only the independent variable
Using a subjective measurement	Results are biased and unclear	Use instruments with clear units
No negative control	Cannot detect contamination	Include a no-treatment group
Only one trial per group	Results may be due to chance	Repeat at least three times

常见错误	为什么有问题	更好的做法
同时改变两个变量	无法确定因果关系	只改变一个自变量
使用主观测量	结果有偏差且不明确	使用带有明确单位的仪器
缺少阴性对照	无法发现污染	加入不进行处理的组
每组只做一次	结果可能只是偶然	至少重复三次

Another common error is ignoring sample age or genetic background as controlled variables. In a study using seedlings, all seeds should come from the same parent plant or the same commercial batch. This reduces genetic variation and makes the experiment more sensitive to the independent variable.

另一个常见错误是忽略样本年龄或遗传背景作为控制变量。在使用幼苗的实验中，所有种子应来自同一母株或同一商业批次。这样可以减少遗传变异，使实验对自变量更加敏感。

In summary, clearly defining variables is not a bureaucratic formality; it is the logical skeleton of every biological experiment. A precise independent variable, a valid dependent variable, rigorous controlled variables, and appropriate controls allow you to answer a biological question with confidence. Always ask yourself before starting: What exactly am I changing? What exactly am I measuring? What else must stay the same?

总之，明确界定变量不是一种形式上的手续，而是每一个生物学实验的逻辑骨架。精确的自变量、有效的因变量、严格的控制变量以及适当的对照，能让你自信地回答生物学问题。在开始之前，永远要问自己：我究竟要改变什么？我究竟要测量什么？还有哪些条件必须保持不变？

Published by TutorHao | Biology Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Experimental Design in Biology: Defining Variables Clearly | 生物实验设计：如何明确实验变量