📚 Experimental Research Methods in Psychology | 心理学实验研究方法解析
Experimental research is the cornerstone of scientific psychology. It allows researchers to establish cause-and-effect relationships by manipulating variables under controlled conditions. Understanding experimental methods is essential for any psychology student preparing for exams, as questions on research methods appear in nearly every major syllabus.
实验研究是科学心理学的基石。它通过在被控制的条件下操纵变量，使研究者能够建立因果关系。对任何备考心理学考试的学生来说，理解实验方法都至关重要，因为几乎所有主流考纲都会涉及研究方法相关题目。
1. What Is an Experiment? | 什么是实验
An experiment is a research method in which the researcher deliberately manipulates one variable (the independent variable) to observe its effect on another variable (the dependent variable), while controlling extraneous factors. The key feature that distinguishes an experiment from other research methods — such as observation or correlation — is the element of manipulation and control.
实验是一种研究方法，研究者有目的地操纵一个变量（自变量）以观察它对另一个变量（因变量）的影响，同时控制额外因素。将实验与观察法、相关法等其他研究方法区分开来的关键特征，是操纵与控制这两个要素。
For an experiment to establish causation, three conditions must be met: (1) the cause must precede the effect; (2) the variables must be correlated; and (3) alternative explanations must be ruled out. Experiments uniquely satisfy all three conditions because they involve temporal sequencing, systematic measurement, and random allocation to conditions.
实验要确立因果关系，必须满足三个条件：（1）因在前、果在后；（2）变量之间存在相关；（3）排除替代性解释。实验之所以独特地满足这三个条件，是因为它包含时间序列安排、系统测量以及随机分配到不同条件。
2. Independent and Dependent Variables | 自变量与因变量
The independent variable (IV) is the factor that the researcher manipulates or changes. It is the presumed cause in a cause-and-effect relationship. The dependent variable (DV) is the factor that the researcher measures; it is the presumed effect. For example, in a study on sleep and memory, the IV might be hours of sleep (4 hours vs. 8 hours), and the DV would be the number of words correctly recalled on a memory test.
自变量（IV）是研究者操纵或改变的因素，是因果关系中被假设为因的一方。因变量（DV）是研究者测量的因素，是被假设为果的一方。例如，在一项关于睡眠与记忆的研究中，自变量可能是睡眠时长（4小时对8小时），因变量则是记忆测试中正确回忆的单词数量。
Operationalisation is the process of defining how a variable will be measured or manipulated in concrete terms. A good operational definition is specific, replicable, and measurable. For instance, 'aggression' might be operationalised as the number of times a participant delivers a loud noise blast to another participant in a competitive task. Exam answers that clearly state operationalised variables earn higher marks.
操作化是将变量以具体方式定义如何测量或操纵的过程。好的操作性定义应当具体、可复制、可测量。例如，"攻击性"可以被操作化为参与者在竞争任务中向对方发出响亮噪音的次数。考试答案中清楚说明操作化变量的作答往往能获得更高分数。
IV: manipulated by researcher → DV: measured by researcher
3. Laboratory Experiments | 实验室实验
A laboratory experiment is conducted in a controlled environment where the researcher has maximum control over extraneous variables. Participants are fully aware they are taking part in research. The classic Milgram obedience study and Asch conformity study are both laboratory experiments that produced powerful findings under artificial conditions.
实验室实验在受控环境中进行，研究者对额外变量拥有最大程度的控制。参与者清楚知道自己正在参与研究。经典的米尔格拉姆服从研究和阿施从众研究，都是在人为条件下产生有力发现的实验室实验。
The primary strength of laboratory experiments is high internal validity: because confounding variables are tightly controlled, the researcher can be confident that changes in the DV are caused by the IV. However, the artificial nature of the setting can reduce external validity — findings may not generalise to real-world situations. Demand characteristics and evaluation apprehension may also arise when participants know they are being studied.
实验室实验的主要优势是内部效度高：由于混淆变量受到严格控制，研究者可以确信因变量的变化是由自变量引起的。然而，环境的人为性可能降低外部效度——研究结果未必能推广到现实情境。当参与者知道自己正在被研究时，还可能产生需求特征和评价焦虑。
4. Field Experiments | 现场实验
A field experiment is conducted in a real-world setting rather than a laboratory. The researcher still manipulates the IV, but participants are often unaware that an experiment is taking place. For example, Piliavin's 1969 'subway experiment' on bystander intervention involved staged collapses on a New York subway train; the IV was the apparent illness of the victim, and the DV was the speed and frequency of helping responses.
现场实验在真实世界情境中进行，而非在实验室里。研究者仍然操纵自变量，但参与者通常不知道自己正在参与实验。例如，皮利亚文1969年关于旁观者干预的"地铁实验"，在纽约地铁上安排假装晕倒；自变量是受害者的表面病症，因变量是帮助行为的速度和频率。
Field experiments offer higher mundane realism and external validity because behaviour is observed in natural contexts. They also reduce demand characteristics because participants are unaware of being studied. However, the researcher has less control over extraneous variables, making replication more difficult, and ethical issues arise because participants cannot give informed consent.
现场实验具有更高的生活真实性和外部效度，因为行为是在自然情境中被观察的。由于参与者不知道自己在被研究，需求特征也会减少。然而，研究者对额外变量的控制力较弱，复现研究更加困难；而且参与者无法给予知情同意，因此会引发伦理问题。
5. Natural Experiments | 自然实验
A natural experiment takes advantage of naturally occurring events that create conditions resembling an experimental manipulation. The researcher does not actively manipulate the IV; instead, they study the effect of an event that happens independently of the researcher, such as a natural disaster, a policy change, or an adoption into a different family environment.
自然实验利用自然发生的事件——这些事件构成了类似于实验操纵的条件。研究者并不主动操纵自变量，而是研究与研究无关的自然事件的影响，例如自然灾害、政策变化或被收养到不同的家庭环境。
Natural experiments are valuable when it is unethical or impractical to manipulate the IV directly. For example, studying the effects of institutionalisation on child development would be impossible as a lab experiment, but natural experiments have been conducted with Romanian orphans adopted into UK families. The drawback is that participants cannot be randomly assigned, so causal conclusions are weaker; many alternative explanations remain plausible.
当直接操纵自变量不道德或不现实时，自然实验非常有价值。例如，研究机构养育对儿童发展的影响不可能通过实验室实验完成，但针对被英国家庭收养的罗马尼亚孤儿已经开展了自然实验。其缺点是参与者无法被随机分配，因此因果结论较弱；许多替代性解释仍然合理存在。
6. Experimental Designs | 实验设计
Experimental design refers to how participants are allocated to the different conditions of the experiment. There are three major designs: independent groups, repeated measures, and matched pairs. Each has distinct strengths and limitations that examiners frequently test.
实验设计指参与者如何被分配到实验的不同条件中。主要有三种设计：独立组设计、重复测量设计和匹配组设计。每种设计都有不同的优点和局限，考官经常就此出题。
| Design | 设计 | Procedure | 程序 | Strength | 优势 | Limitation | 局限 |
| Independent Groups 独立组 |
Different participants in each condition | No order effects; participants less likely to guess aim | Participant variables may confound results; requires more participants |
| Repeated Measures 重复测量 |
Same participants in all conditions | No participant variables; fewer participants needed | Order effects and demand characteristics; practice effects can distort results |
| Matched Pairs 匹配组 |
Different but similar participants matched on key traits | Controls participant variables; no order effects | Time-consuming to match; cannot match all relevant traits |
Order effects in repeated measures designs can be controlled using counterbalancing, in which half the participants complete condition A first and half complete condition B first. This ensures that any fatigue or practice effects are evenly distributed across conditions rather than systematically attached to one condition.
重复测量设计中的顺序效应可以通过平衡法加以控制：一半参与者先完成条件A，另一半先完成条件B。这样可以确保疲劳或练习效应均匀分布在各个条件之间，而不是系统性地贴合到某一个条件上。
7. Controlling Extraneous Variables | 额外变量的控制
Extraneous variables are any variables other than the IV that could affect the DV. When an extraneous variable systematically varies with the IV, it becomes a confounding variable, threatening the internal validity of the experiment. Common extraneous variables include participant factors (age, IQ, motivation), situational factors (noise, lighting, time of day), and procedural factors (instructions, equipment).
额外变量是除自变量之外可能影响因变量的任何变量。当额外变量随自变量系统性变化时，它就成为混淆变量，威胁实验的内部效度。常见的额外变量包括参与者因素（年龄、智商、动机）、情境因素（噪音、照明、一天中的时间段）和程序因素（指导语、设备）。
Several techniques are used to control extraneous variables: standardisation ensures that all participants receive identical instructions and procedures; random allocation distributes participant variables across conditions; counterbalancing controls for order effects; single-blind designs prevent participant expectations; and double-blind designs prevent both participant and researcher bias. For situational variables, researchers may use randomisation of trial order and automated equipment to ensure consistency.
控制额外变量有若干技术：标准化确保所有参与者接受相同的指导和程序；随机分配使参与者变量分布到各个条件；平衡法控制顺序效应；单盲设计防止参与者的预期偏差；双盲设计则同时防止参与者和研究者的偏见。对于情境变量，研究者可以使用试验顺序随机化以及自动化设备来确保一致性。
8. Reliability and Validity | 信度与效度
Reliability refers to the consistency of a measure — whether the same results would be obtained if the study were repeated. Internal reliability is assessed using split-half or Cronbach's alpha techniques; external reliability is assessed using test-retest or inter-rater methods. An unreliable experiment cannot produce trustworthy conclusions regardless of its design.
信度指测量的一致性——如果重复研究是否能够得到相同的结果。内部信度通过折半法或克隆巴赫α系数评估；外部信度通过重测法或评分者间一致性评估。无论设计如何，不可靠的实验都无法产生值得信赖的结论。
Validity refers to whether the experiment actually measures what it claims to measure. Internal validity concerns the extent to which the IV truly caused the observed change in the DV. External validity concerns generalisability: population validity (to other people), ecological validity (to other settings), and temporal validity (to other time periods). A well-designed experiment must balance both — a tightly controlled lab study may have high internal validity but low ecological validity, whereas a field study often shows the reverse pattern.
效度指实验是否真正测量了它声称要测量的内容。内部效度关注自变量是否真正导致了因变量的观察变化。外部效度关注可推广性：总体效度（推广到其他人）、生态效度（推广到其他情境）以及时间效度（推广到其他时期）。设计良好的实验必须兼顾两者——严格控制下的实验室研究可能内部效度高但生态效度低，而现场研究往往呈现相反模式。
Reliability (consistency) × Validity (accuracy) = Trustworthy findings
9. Ethics in Experimental Research | 实验研究中的伦理问题
Ethical guidelines — such as those published by the British Psychological Society (BPS) and the American Psychological Association (APA) — require that psychological research protects participants from harm. Key ethical principles include informed consent, the right to withdraw, deception, debriefing, confidentiality, and protection from physical and psychological harm.
伦理准则——例如英国心理学会（BPS）和美国心理学会（APA）发布的指南——要求心理学研究保护参与者免受伤害。关键伦理原则包括知情同意、退出权、欺骗问题、事后说明、保密以及身心伤害防护。
In practice, ethical obligations can conflict with methodological needs. Full informed consent would ruin a study on obedience or bystander intervention because demand characteristics would emerge. The solution is usually to obtain presumptive consent or a full debriefing after the study, and to ensure that if deception is used, the research has high scientific value and participants are not harmed. Examiners expect students to evaluate experiments on ethical grounds as well as methodological ones.
在实践中，伦理义务有时会与方法需求产生冲突。如果完全知情同意，会毁掉一项关于服从或旁观者干预的研究，因为需求特征会浮现。通常的解决方案是获取假设性同意，或在研究结束后进行全面的事后说明；如果使用了欺骗手段，需要确保研究具有高度的科学价值，且参与者未受到伤害。考官期望学生同时从伦理角度和方法角度评价实验。
10. Strengths and Limitations of Experimental Methods | 实验方法的优势与局限
The overall strengths of experimental methods include their ability to establish cause-and-effect relationships, their precision through quantitative measurement, their replicability when properly standardised, and their compatibility with statistical testing. These features give experimental findings a credibility that other methods often lack.
实验方法的总体优势包括：能够建立因果关系、通过量化测量获得精确性、在标准化得当的情况下可以复现，以及适合进行统计检验。这些特征使实验发现具有其他方法往往不具备的可信度。
The overall limitations include artificiality (low mundane realism), ethical constraints on what can be manipulated, problems of demand characteristics and researcher bias, and the difficulty of studying complex, multi-causal human behaviours in isolation. Moreover, experimental results describe what happens on average in groups, but may obscure meaningful individual differences.
总体局限包括：人为性（生活真实性低）、伦理上不能操纵某些变量、需求特征和研究者偏见的问题，以及难以在孤立条件下研究复杂的、多因性的人类行为。此外，实验结果描述的是群体平均水平上发生了什么，可能会掩盖有意义的个体差异。
When writing essay answers, a balanced evaluation should address both sides and then arrive at a reasoned conclusion. For example: 'Although laboratory experiments lack ecological validity, their rigorous control of confounding variables makes them the gold standard for establishing causality in cognitive psychology, where artificial tasks often still tap fundamental processes.'
撰写论文式答案时，平衡的评价应兼顾两方面，并最终得出有推理依据的结论。例如："尽管实验室实验缺乏生态效度，但其对混淆变量的严格控制使其成为认知心理学中确立因果关系的黄金标准——在该领域，人工任务往往仍然触及基本心理过程。"
11. Exam Focus Points | 考试要点提示
To score highly on experimental method questions, remember these principles: always identify the IV and DV explicitly; use the correct terminology (confounding variable, demand characteristics, counterbalancing); compare designs rather than merely describing them; and support every evaluation point with a concrete example from a classic study. Marks are often awarded for the specificity and range of your evaluation points, not just their quantity.
要在实验方法题目上拿高分，请记住这些原则：总是明确指出自变量和因变量；使用正确的术语（混淆变量、需求特征、平衡法）；比较不同设计而不只是做描述；为每个评价点都配上经典研究的具体例子。评分往往取决于评价点的具体性和覆盖面，而不只是数量。
A practical revision strategy is to test yourself on the following questions: Which design would you choose to investigate a memory improvement technique, and why? How would you control for participant variables in a study of exam anxiety? What ethical issues arise in field experiments on helping behaviour, and how could they be resolved? Being able to answer these fluently will prepare you for the majority of research methods exam questions.
一个实用的复习策略是自我测验以下问题：如果要研究一种记忆提升技术，你会选择哪种设计，为什么？在考试焦虑研究中，你会如何控制参与者变量？关于助人行为的现场实验会引发哪些伦理问题，如何解决？能够流利回答这些问题，就能在大多数研究方法考题中做好准备。
Finally, remember that psychology is a science built on evidence. Every famous theory — from attachment to memory models — rests on experimental evidence. A deep understanding of experimental methods will not only help you pass exams but will also enable you to critically evaluate psychological claims throughout your further studies.
最后请记住，心理学是建立在证据之上的科学。每一个著名理论——从依恋理论到记忆模型——都依赖于实验证据。深入理解实验方法不仅能帮助你通过考试，还能让你在后续学习中批判性地评估各种心理学主张。
Published by TutorHao | Psychology 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