Psychology Research Methods: Experimental Design and Variable Control | 心理学研究方法：实验设计与变量控制

📚 Psychology Research Methods: Experimental Design and Variable Control | 心理学研究方法：实验设计与变量控制

Psychology aims to understand behaviour and mental processes. Among all research methods, the experiment is the most powerful tool for identifying cause-and-effect relationships. It allows researchers to manipulate specific factors and observe the consequences while carefully controlling other influences.

心理学旨在理解行为和心理过程。在所有研究方法中，实验是确定因果关系最有力的工具。它让研究者能够操控特定因素，观察其结果，同时谨慎控制其他影响。

1. What is an Experiment? | 什么是实验？

An experiment is a research method in which the researcher deliberately changes one or more independent variables and measures the effect on one or more dependent variables. The key feature is control: the experimenter controls the environment, the procedure, and the assignment of participants to conditions.

实验是一种研究方法，研究者有目的地改变一个或多个自变量，并测量其对因变量的影响。其关键特征是控制：实验者控制环境、程序以及被试在不同条件下的分配。

In a true experiment, participants are randomly assigned to different conditions. This random assignment helps to ensure that the groups are equivalent before the manipulation, so any later differences can be attributed to the independent variable rather than pre-existing differences.

在真正的实验中，被试被随机分配到不同条件下。这种随机分配有助于确保各组在操作前是等价的，因此之后的差异可以归因于自变量，而非先前存在的差异。

2. Independent and Dependent Variables | 自变量与因变量

The independent variable (IV) is the factor that the researcher manipulates. It is the presumed cause. The dependent variable (DV) is the factor that is measured; it is the presumed effect. For example, if a researcher wants to test the effect of caffeine on reaction time, caffeine dose is the IV and reaction time is the DV.

自变量（IV）是研究者操控的因素，它是被认定的原因。因变量（DV）是被测量的因素，它是被认定的结果。例如，如果研究者想测试咖啡因对反应时间的影响，那么咖啡因剂量是自变量，反应时间是因变量。

An operational definition is essential for both variables. It specifies how the IV is manipulated and how the DV is measured. For instance, “caffeine dose” might be operationally defined as 0 mg, 50 mg, or 100 mg in a pill; “reaction time” might be defined as milliseconds to press a button after a stimulus.

操作性定义对两个变量都至关重要。它规定了自变量如何被操控、因变量如何被测量。例如，“咖啡因剂量”可操作性地定义为药片中的0毫克、50毫克或100毫克；“反应时间”可定义为刺激出现后按下按钮所需的毫秒数。

3. Control of Extraneous Variables | 额外变量的控制

Extraneous variables are any variables other than the IV that could affect the DV. If they are not controlled, they become alternative explanations for the results. There are several strategies to control them: holding variables constant, randomizing across conditions, and using counterbalancing.

额外变量是除自变量之外可能影响因变量的任何变量。如果不加以控制，它们就会成为对结果的替代解释。控制额外变量的策略有多种：使变量保持恒定、随机化分配到不同条件、以及使用平衡设计。

Holding constant: keep the environment (e.g., room temperature, time of day) identical for all participants.

保持恒定：让环境（如室温、一天中的时间）对所有被试完全相同。
Randomization: random assignment spreads the influence of individual differences across groups.

随机化：随机分配使个体差异的影响在各组之间被分散。
Counterbalancing: in within-subjects designs, the order of conditions is varied to prevent order effects.

平衡：在被试内设计中，改变条件的呈现顺序以预防顺序效应。

4. Experimental vs. Control Groups | 实验组与对照组

The experimental group receives the treatment or the active level of the IV. The control group does not receive the treatment, or receives a placebo. The comparison between these groups reveals the effect of the IV.

实验组接受处理或自变量的活跃水平。对照组不接受处理，或接受安慰剂。两组之间的比较揭示了自变量的效应。

A placebo control is especially important in psychology and medicine. If participants know they are in the experimental group, they may expect to improve or change, which could bias the results. A placebo (inert substance or fake procedure) helps to eliminate this expectation effect.

安慰剂对照在心理学和医学中尤为重要。如果被试知道自己处于实验组，他们可能预期改善或变化，从而偏倚结果。安慰剂（惰性物质或虚假程序）有助于消除这种预期效应。

Experimental group → Treatment → Measure DV
Control group → Placebo/No treatment → Measure DV

5. Random Assignment vs. Random Sampling | 随机分配与随机抽样

Random sampling refers to how participants are selected from the population. It enhances external validity, or the ability to generalise findings. Random assignment refers to how participants are placed into groups once they are in the study. It enhances internal validity, or the ability to draw causal conclusions.

随机抽样指被试如何从总体中被选择。它提高外部效度，即研究结果的可推广性。随机分配指被试进入研究后如何被分到各组。它提高内部效度，即得出因果结论的能力。

Method	Purpose	Threat addressed
Random sampling	Representative sample	External validity
Random assignment	Equivalent groups	Internal validity

Both processes are important, but they serve different purposes. A researcher may randomly assign participants from a clinical population, but the results might only generalise to that specific population.

这两种过程都很重要，但目的不同。研究者可以对来自临床总体的被试进行随机分配，但结果可能只适用于该特定总体。

6. Between-Subjects vs. Within-Subjects Designs | 被试间设计与被试内设计

In a between-subjects design, each participant experiences only one level of the IV. The groups are compared. This design avoids order effects but requires more participants and assumes groups are equivalent before the study.

在被试间设计中，每位被试只经历自变量的一个水平，各组进行比较。这种设计避免了顺序效应，但需要更多被试，并假定各组在研究前是等价的。

In a within-subjects design (also called repeated-measures design), each participant experiences all levels of the IV. This design is more powerful because individual differences are held constant. However, it may suffer from order effects, such as practice or fatigue. Counterbalancing is used to reduce these effects.

在被试内设计（又称重复测量设计）中，每位被试经历自变量的所有水平。这种设计更有效，因为个体差异被恒定控制。然而，它可能受到顺序效应的影响，如练习效应或疲劳效应。平衡设计用于减少这些影响。

Between-subjects: A → IV level 1, B → IV level 2
Within-subjects: Same participant → level 1 then level 2

7. Confounding Variables and How to Avoid Them | 混淆变量及其避免方法

A confounding variable is an extraneous variable that changes systematically with the IV. If a confounding variable is present, the researcher cannot determine whether the effect on the DV is caused by the IV or by the confounding variable. For example, if reaction time is measured at different times of day for different conditions, time becomes a confound.

混淆变量是随自变量系统性变化的额外变量。如果存在混淆变量，研究者就无法确定对因变量的效应是由自变量还是由混淆变量引起的。例如，如果不同条件在不同时间段测量反应时间，那么时间就成了混淆变量。

Strategies to avoid confounding variables include:

避免混淆变量的策略包括：

Use a double-blind procedure so that neither the participant nor the experimenter knows which condition is administered.

使用双盲程序，使被试和实验者都不知道处于哪种条件下。
Hold the administration procedure constant for all groups.

对所有组保持相同的实施程序。
Use matching or block randomisation when sample sizes are small.

当样本量较小时，使用匹配或区组随机化。

8. Internal and External Validity | 内部效度与外部效度

Internal validity refers to the degree to which the observed effect is due to the IV and not some other factor. High internal validity allows causal statements. Random assignment, controlled procedures, and standardised instructions all increase internal validity.

内部效度指观察到的效应在多大程度上是由自变量而非其他因素导致的。高内部效度允许作出因果陈述。随机分配、受控程序和标准化指导语都能提高内部效度。

External validity refers to the degree to which results can be generalised to other people, settings, and times. Mundane realism (how realistic the experiment appears) and ecological validity (how closely the task matches real life) are important considerations. There is often a trade-off: tightly controlled laboratory experiments have high internal validity but lower external validity.

外部效度指结果能在多大程度上推广到其他人、其他情境和其他时间。日常真实性（实验看起来多真实）和生态效度（任务与现实生活的匹配程度）是重要考量。两者常常存在权衡：严格控制下的实验室实验具有高内部效度，但外部效度较低。

9. Common Threats to Validity | 常见威胁效度的因素

Several known threats can undermine an experiment. History effects are events that occur during the experiment that are not part of the IV. Maturation refers to natural changes in participants over time (e.g., growing tired, becoming older). Testing effects occur when taking a pretest changes performance on a later test.

若干已知威胁会削弱实验。历史效应指实验期间发生的、不属于自变量的事件。成熟效应指被试随时间发生的自然变化（如变累、年龄增长）。测试效应指前测改变后测表现的情况。

Instrumentation occurs when the measuring instrument changes during the study, such as an observer becoming more lenient or a machine recalibrating. Attrition is the loss of participants from the study, which can make the final sample non-representative. Selection bias happens when groups are not equivalent from the start, often due to non-random assignment.

测量工具效应指研究中测量工具发生变化，如观察者变得更宽松或设备重新校准。被试流失指研究过程中被试的减少，这可能使最终样本不具有代表性。选择偏误指各组一开始不等价，通常因非随机分配导致。

Internal validity = control over confounds
External validity = generalisability of results

10. Ethical Considerations in Experimental Research | 实验研究中的伦理考量

Ethical guidelines govern psychological experiments. Participants must give informed consent, meaning they understand the purpose, procedures, and risks before agreeing to take part. Deception is only permitted when no alternative exists and the study has clear scientific benefit. After the study, participants must be debriefed — fully informed about the true nature of the research.

伦理规范约束心理学实验。被试必须签署知情同意，即在同意参与前了解目的、程序和风险。欺骗只有在没有替代方案且研究具有明确科学价值时才被允许。研究结束后，被试必须接受事后说明——即充分了解研究的真实性质。

Participants also have the right to withdraw at any time without penalty. Their data must remain confidential. Researchers must avoid causing physical or psychological harm. Ethical committees review protocols before experiments can begin.

被试还有权在任何时候无惩罚地退出。其数据必须保密。研究者必须避免造成身体或心理伤害。伦理委员会在实验开始前审查研究方案。

11. Practical Example: Designing a Simple Experiment | 实践示例：设计一个简单实验

Suppose a researcher wants to test whether listening to calm music improves memory recall. The IV is the type of music (calm music vs. silence). The DV is the number of words correctly recalled from a list.

假设研究者想测试聆听平静音乐是否提高记忆回忆成绩。自变量是音乐类型（平静音乐 vs. 安静）。因变量是正确回忆的单词数量。

Steps:

步骤：

Recruit 60 participants from a university sample.

从大学样本中招募60名被试。
Randomly assign them to one of two groups (30 each).

将他们随机分配到两组之一（每组30人）。
Both groups study the same list of 20 words for two minutes. The calm-music group studies while music plays; the silence group studies in silence.

两组都用两分钟学习同一份20个单词的词表。平静音乐组在音乐播放时学习；安静组在安静中学习。
After a five-minute filler task, all participants are asked to write down as many words as they can recall.

经过五分钟的填充任务后，所有被试被要求写下他们能回忆出的尽可能多的单词。
Compare the mean recall scores using an appropriate statistical test (e.g., t-test).

使用适当的统计检验（如t检验）比较平均回忆成绩。

This experiment controls many extraneous variables: identical word list, identical study time, identical filler task, and random assignment. It establishes a clear comparison and allows a causal conclusion about the effect of music on memory.

这个实验控制了许多额外变量：相同的词表、相同的学习时间、相同的填充任务以及随机分配。它建立了清晰的比较，并允许就音乐对记忆的影响得出因果结论。

12. Conclusion | 结论

Experimental design is fundamental to psychological research. The careful manipulation of independent variables, accurate measurement of dependent variables, and rigorous control of extraneous variables enable psychologists to answer causal questions. Random assignment, appropriate group designs, and attention to validity and ethics are essential for producing trustworthy and reproducible findings.

实验设计是心理学研究的基础。精细地操控自变量、准确地测量因变量以及严格地控制额外变量，使心理学家能够回答因果问题。随机分配、合适的组间设计以及对效度和伦理的关注，是产生可信且可重复结果的关键。

Mastering these methods not only helps in evaluating published research but also in designing one’s own studies. A well-designed experiment is a powerful tool to reveal the hidden causes of behaviour.

掌握这些方法不仅有助于评估已发表的研究，也有助于设计我们自己的研究。一项设计良好的实验是揭示行为背后隐藏原因的强有力工具。

Published by TutorHao | Psychology Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Psychology Research Methods: Experimental Design and Variable Control | 心理学研究方法：实验设计与变量控制