📚 Experimental Design: Key Principles in Psychological Research | 心理学研究：实验设计的关键要点
Experimental design is the backbone of psychological research. It provides a structured framework for testing hypotheses, establishing cause-and-effect relationships, and ensuring that findings are reliable and valid. Understanding the key principles of experimental design is essential for any psychology student who wishes to critically evaluate research or conduct their own studies.
实验设计是心理学研究的基石。它为检验假设、建立因果关系以及确保研究结果的可靠性和有效性提供了结构化的框架。对于任何希望批判性评估研究或开展自己研究的心理学学生而言，理解实验设计的关键原则至关重要。
1. What Is Experimental Design? | 什么是实验设计？
Experimental design refers to the overall plan or blueprint that guides a researcher in collecting and analysing data. It specifies how participants are allocated to different conditions, which variables are manipulated and measured, and how potential confounding factors are controlled. A well-designed experiment allows the researcher to make strong inferences about whether changes in one variable cause changes in another.
实验设计是指指导研究者收集和分析数据的整体计划或蓝图。它规定了参与者如何被分配到不同条件、哪些变量被操纵和测量，以及如何控制潜在的混淆因素。一个设计良好的实验使研究者能够对某一变量的变化是否导致另一变量的变化做出强有力的推断。
The defining feature of a true experiment is the manipulation of an independent variable (IV) while holding all other factors constant, combined with random allocation of participants to conditions. This allows the researcher to conclude that the IV, rather than extraneous factors, is responsible for any observed changes in the dependent variable (DV).
真正实验的定义特征是在保持所有其他因素恒定的情况下操纵自变量（IV），并将参与者随机分配到各条件中。这使研究者能够得出结论：是自变量而非额外因素导致了因变量（DV）的任何观察到的变化。
2. Independent and Dependent Variables | 自变量与因变量
The independent variable (IV) is the factor that the researcher deliberately manipulates. It is the presumed cause in the cause-and-effect relationship. In a study on the effects of caffeine on reaction time, for example, the amount of caffeine consumed would be the IV. The IV must have at least two levels or conditions — for instance, a caffeine group and a placebo group.
自变量（IV）是研究者刻意操纵的因素，它被认为是因果关系中的原因。例如，在一项关于咖啡因对反应时间影响的研究中，摄入的咖啡因量就是自变量。自变量必须至少有两个水平或条件——例如，咖啡因组和安慰剂组。
The dependent variable (DV) is the outcome that the researcher measures to assess the effect of the IV. It is the presumed effect. In the caffeine example, the DV would be reaction time, measured in milliseconds. The DV must be operationalised — defined in a way that is precise, observable, and measurable, such as “time taken to press a button after a visual stimulus appears.”
因变量（DV）是研究者为评估自变量效果而测量的结果，它被认为是因果关系中的效应。在咖啡因的例子中，因变量就是反应时间，以毫秒为单位进行测量。因变量必须被操作化——即以精确、可观察和可测量的方式定义，例如”视觉刺激出现后按下按钮所需的时间”。
Operationalisation is critical because it ensures that abstract concepts are translated into concrete, measurable terms. Without clear operational definitions, replication becomes impossible and the validity of the research is undermined.
操作化至关重要，因为它确保抽象概念被转化为具体、可测量的术语。没有清晰的操作定义，复制研究将变得不可能，研究的效度也会受到损害。
3. Experimental and Control Groups | 实验组与对照组
In a typical experiment, participants are divided into at least two groups. The experimental group receives the treatment or is exposed to the manipulation of the IV, while the control group does not receive the treatment or receives a placebo. The control group serves as a baseline against which the effects of the IV can be compared.
在典型的实验中，参与者被分为至少两组。实验组接受处理或暴露于自变量的操纵中，而对照组不接受处理或接受安慰剂。对照组作为基线，用于与自变量的效果进行比较。
For example, in Milgram’s famous obedience studies, participants in the experimental condition were instructed to administer increasingly severe shocks to a “learner,” while control conditions varied the proximity or authority of the experimenter. The comparison between conditions allowed Milgram to identify situational factors that influenced obedience.
例如，在米尔格拉姆著名的服从研究中，实验条件下的参与者被指示对”学习者”施加越来越严重的电击，而对照条件则改变了实验者的接近程度或权威性。条件之间的比较使米尔格拉姆能够识别影响服从的情境因素。
The inclusion of a control group is essential for ruling out alternative explanations. Without a control group, it is impossible to know whether any observed change in the DV is truly due to the IV or to other factors such as the mere passage of time, participant expectations, or natural fluctuations in behaviour.
纳入对照组对于排除替代性解释至关重要。没有对照组，就不可能知道因变量的任何观察到的变化究竟是由于自变量还是由于其他因素，如时间流逝、参与者期望或行为的自然波动。
4. Random Allocation | 随机分配
Random allocation is the process of assigning participants to different conditions in such a way that each participant has an equal chance of being placed in any group. This is a cornerstone of experimental design because it helps to distribute participant differences — such as age, intelligence, personality, or motivation — evenly across conditions.
随机分配是指将参与者分配到不同条件的过程，使每位参与者被分到任何一组的机会均等。这是实验设计的基石，因为它有助于将参与者的个体差异——如年龄、智力、性格或动机——均匀地分布到各条件中。
When random allocation is used, any pre-existing individual differences are unlikely to systematically bias the results. For instance, if a researcher is testing the effect of a new teaching method on memory, random allocation ensures that one group is not inadvertently filled with stronger students who would skew the results regardless of the teaching method.
当使用随机分配时，任何预先存在的个体差异都不太可能系统地偏倚结果。例如，如果研究者正在测试一种新教学方法对记忆的影响，随机分配可确保某一组不会不经意间全是成绩较好的学生，因为无论教学方法如何，这些学生都会使结果产生偏差。
However, random allocation is not the same as random sampling. Random sampling refers to how participants are selected from the wider population, whereas random allocation refers to how selected participants are assigned to conditions. A study can have random allocation but still use a convenience sample, which limits the generalisability of the findings.
然而，随机分配并不等同于随机抽样。随机抽样是指如何从更广泛的人群中选择参与者，而随机分配是指如何将已选定的参与者分配到各条件中。一项研究可以具有随机分配，但仍然使用方便样本，这限制了研究结果的推广性。
5. Controlling Extraneous Variables | 控制额外变量
Extraneous variables are any variables other than the IV that could affect the DV. If they are left uncontrolled, they become confounding variables, offering an alternative explanation for the findings. For example, in a study on noise levels and concentration, the time of day, participant fatigue, or room temperature could all confound the results if not controlled.
额外变量是除自变量之外可能影响因变量的任何变量。如果它们未被控制，就会成为混淆变量，为研究结果提供替代解释。例如，在一项关于噪音水平与专注力的研究中，一天中的时间、参与者的疲劳程度或室温如果不加控制，都可能混淆结果。
There are several methods for controlling extraneous variables. Standardisation ensures that all participants experience the same procedures, instructions, and environmental conditions. The use of a double-blind procedure prevents both the researcher and the participants from knowing which condition is being administered, eliminating demand characteristics and experimenter bias.
控制额外变量有几种方法。标准化确保所有参与者经历相同的程序、指令和环境条件。使用双盲程序可防止研究者和参与者知道正在实施的是哪个条件，从而消除需求特征和实验者偏差。
Another key technique is counterbalancing, used in repeated-measures designs to control for order effects such as practice or fatigue. By presenting conditions in a different order for different participants, any effects due to the sequence of presentations are evenly distributed, allowing the researcher to isolate the true effect of the IV.
另一个关键技术是平衡设计（counterbalancing），用于被试内设计以控制练习效应或疲劳等顺序效应。通过为不同参与者以不同顺序呈现条件，任何由于呈现顺序而产生的效应都被均匀分布，从而使研究者能够分离出自变量的真实效应。
6. Types of Experimental Design | 实验设计的类型
Three main types of experimental design are commonly used in psychology: independent measures (between-subjects), repeated measures (within-subjects), and matched pairs. Each has distinct strengths and limitations that researchers must weigh when planning a study.
心理学中常用三种主要的实验设计类型：独立测量设计（被试间设计）、重复测量设计（被试内设计）和匹配配对设计。每种类型都有其独特的优势和局限，研究者在规划研究时必须加以权衡。
Independent measures design involves different participants in each condition. It avoids order effects and demand characteristics, but requires more participants and is vulnerable to participant variability. Repeated measures design uses the same participants in all conditions, eliminating participant variability and requiring fewer participants, but it introduces order effects and may lead to fatigue or practice effects.
独立测量设计是在每个条件下使用不同的参与者。它避免了顺序效应和需求特征，但需要更多参与者，且易受参与者变异的影响。重复测量设计在所有条件下使用相同的参与者，消除了参与者变异且所需参与者较少，但它引入了顺序效应，并可能导致疲劳或练习效应。
The matched pairs design seeks a middle ground: participants are matched on key characteristics such as age, IQ, or gender, and then one member of each pair is allocated to each condition. This reduces participant variability while avoiding order effects, although matching is difficult and time-consuming in practice.
匹配配对设计则寻求一种折中方案：参与者在年龄、智商或性别等关键特征上进行匹配，然后将每对中的一名成员分配到每个条件中。这减少了参与者变异，同时避免了顺序效应，尽管在实际操作中匹配是困难且耗时的。
7. Internal and External Validity | 内部效度与外部效度
Internal validity refers to the degree to which a study can confidently attribute changes in the DV to the IV, rather than to other factors. A study with high internal validity allows the researcher to make causal claims. Threats to internal validity include confounding variables, demand characteristics, experimenter bias, and attrition — when participants drop out of the study.
内部效度是指一项研究能够有信心地将因变量的变化归因于自变量而非其他因素的程度。具有高内部效度的研究使研究者能够做出因果性断言。内部效度的威胁包括混淆变量、需求特征、实验者偏差以及参与者流失。
External validity refers to the extent to which the findings of a study can be generalised beyond the specific sample, setting, and time of the research. A study conducted with a small, unrepresentative sample in a laboratory setting may have high internal validity but low external validity, meaning its findings may not apply to real-world populations or situations.
外部效度是指研究结果能够在多大程度上推广到特定样本、环境和时间之外。在实验室环境中使用小型、不具代表性的样本进行的研究可能具有较高的内部效度，但外部效度较低，这意味着其结果可能不适用于现实世界的人群或情境。
There is often a trade-off between internal and external validity. Tightly controlled laboratory experiments tend to maximise internal validity but may be artificial and lacking in ecological validity. Field experiments, conducted in real-world settings, often have higher external validity but are more difficult to control, reducing internal validity. Researchers must prioritise based on their research question and the stage of the research programme.
内部效度与外部效度之间往往存在权衡。严格控制变量的实验室实验倾向于最大化内部效度，但可能显得人为化且缺乏生态效度。在现实环境中进行的现场实验通常具有较高的外部效度，但更难以控制，从而降低了内部效度。研究者必须根据其研究问题和研究项目的阶段来确定优先顺序。
8. Experimenter Bias and Demand Characteristics | 实验者偏差与需求特征
Experimenter bias occurs when the researcher’s expectations or beliefs inadvertently influence the outcome of the study. This can happen through subtle cues — such as tone of voice, facial expressions, or body language — that signal to participants how they are expected to behave. Experimenter bias can also affect the interpretation of data, leading the researcher to see what they expect to see.
实验者偏差是指研究者的期望或信念在无意中影响研究结果的情况。这可以通过微妙的线索——如语调、面部表情或肢体语言——发生，向参与者暗示他们被期望如何表现。实验者偏差还可能影响数据的解释，使研究者看到他们期望看到的结果。
Demand characteristics are cues in the research setting that reveal the purpose of the study to the participants, who may then alter their behaviour accordingly. Participants might try to please the researcher by acting as they believe is expected, or they might deliberately behave in ways that they think will undermine the study — a phenomenon known as the “screw-you effect.”
需求特征是研究情境中揭示研究目的的线索，参与者可能会据此改变他们的行为。参与者可能试图取悦研究者，按照他们认为被期望的方式行事，或者他们可能故意以他们认为会破坏研究的方式行事——这种现象被称为”去你的效应”（screw-you effect）。
Several strategies can mitigate these threats. A double-blind procedure, where neither the participant nor the researcher knows the condition assignment, is the most powerful technique. Standardised scripts, automated data collection, and debriefing — where participants are informed about the true purpose after the study — also help to reduce bias and maintain ethical standards.
有几种策略可以减轻这些威胁。双盲程序是其中最强有力的技术，即参与者和研究者都不知道条件分配。标准化脚本、自动化数据收集以及在研究结束后告知参与者真实目的的事后说明（debriefing），也有助于减少偏差并维持伦理标准。
9. Ethical Considerations | 伦理考量
Ethical principles are fundamental to psychological research. Researchers must obtain informed consent from participants, ensuring that they understand the nature of the study and their right to withdraw at any time. Deception may be used only when strictly necessary and when the potential benefits outweigh the risks, but participants must be debriefed and given the opportunity to withdraw their data.
伦理原则是心理学研究的基础。研究者必须获得参与者的知情同意，确保他们了解研究的性质以及随时退出研究的权利。只有在严格必要且潜在收益大于风险的情况下才允许使用欺骗，但必须对参与者进行事后说明，并给予他们撤回自己数据的机会。
Researchers must also protect participants from physical and psychological harm. This includes ensuring that procedures are not distressing, that confidential information is protected, and that data are anonymised. Ethical review boards, such as institutional review boards (IRBs) in the US or university ethics committees in the UK, must approve any study involving human participants before it begins.
研究者还必须保护参与者免受身体和心理伤害。这包括确保程序不会引起痛苦、保护机密信息以及使数据匿名化。伦理审查委员会——如美国的研究机构审查委员会（IRB）或英国的大学伦理委员会——必须在涉及人类参与者的任何研究开始之前予以批准。
Ethical considerations are not merely bureaucratic formalities; they are integral to the scientific integrity of the research. If participants cannot trust that researchers will protect their welfare, the entire enterprise of psychological research is jeopardised. Moreover, ethical research is often better science — it is more transparent, more replicable, and more respectful of the individuals who make research possible.
伦理考量不仅仅是行政程序上的形式；它们与研究科学完整性密不可分。如果参与者不能信任研究者会保护他们的福祉，那么整个心理学研究事业都将受到损害。此外，符合伦理的研究往往是更好的科学——它更透明、更可复制，也更尊重那些使研究成为可能的个体。
10. Statistical Significance and Drawing Conclusions | 统计显著性与得出结论
Once the data are collected, the researcher must determine whether the results are statistically significant — that is, whether the observed differences between conditions are unlikely to have occurred by chance. Significance is typically assessed using a p-value; a result is considered significant if p < .05, meaning there is less than a 5% probability that the finding is due to random variation.
数据收集完成后，研究者必须确定结果是否具有统计显著性——即观察到的条件间差异是否不太可能由偶然因素造成。显著性通常使用p值进行评估；当 p < .05 时，结果被认为是显著的，这意味着该发现由随机变异造成的概率小于5%。
It is important to distinguish between statistical significance and practical significance. A result can be statistically significant but have a trivial effect size, meaning the effect is real but so small that it has little real-world impact. Researchers should report effect sizes, such as Cohen’s d or eta-squared, alongside p-values to give a fuller picture of their findings.
区分统计显著性和实际显著性非常重要。一个结果可能具有统计显著性，但效应量很小，这意味着效应是真实存在的，但小到在现实世界中几乎没有影响。研究者应在报告p值的同时报告效应量，如Cohen’s d或eta²，以更全面地展示其发现。
Finally, a single experiment is never sufficient to establish a scientific fact. Replication — the repetition of a study to see if the same results are obtained — is the gold standard of scientific evidence. Only when findings are consistently replicated across different samples, settings, and researchers can we have confidence in their robustness. This is why researchers must report their methods in sufficient detail to allow others to replicate their work.
最后，单次实验永远不足以确立一个科学事实。复制——重复研究以检验是否获得相同结果——是科学证据的黄金标准。只有当研究结果在不同样本、不同环境和不同研究者之间被一致地复制时，我们才能对其稳健性抱有信心。这就是为什么研究者必须足够详细地报告其方法，以便他人能够复制他们的工作。
In conclusion, experimental design is the rigorous framework that allows psychologists to move beyond description and towards causal explanation. At its core lie key elements: careful manipulation of the IV, precise measurement of the DV, control of extraneous variables, random allocation, and ethical guardianship throughout. Mastery of these principles enables students not only to evaluate the quality of published research but also to design studies of their own that contribute meaningfully to our understanding of human behaviour.
总之，实验设计是使心理学家能够超越描述、迈向因果解释的严格框架。其核心要素包括：对自变量的谨慎操纵、对因变量的精确测量、对额外变量的控制、随机分配以及贯穿始终的伦理守护。掌握这些原则使学生不仅能够评估已发表研究的质量，还能够设计自己的研究，为我们理解人类行为做出有意义的贡献。
Published by TutorHao | Psychology 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