📚 Common Research Methods in Psychology Explained | 心理学常见研究方法解析
Psychology is a science, and the methods psychologists use to gather evidence determine whether their conclusions can be trusted. From tightly controlled laboratory experiments to rich, detailed case studies, every approach has its own logic, strengths, and limitations. This guide explains the most common research methods you will encounter in your psychology revision.
心理学是一门科学，心理学家收集证据的方式决定了结论是否可信。从严格受控的实验室实验到丰富细致的个案研究，每种方法都有其内在逻辑、优点和局限。本指南将为你解析心理学复习中最常遇到的研究方法。
1. The Scientific Method | 科学方法
Psychology is an empirical discipline: knowledge should be based on observation and measurement rather than on intuition or authority. Researchers begin with a theory, derive a hypothesis from it, and design a study to test that hypothesis. A hypothesis must be operationalised, meaning that the variables are given clear definitions so they can be measured objectively.
心理学是一门经验学科：知识应基于观察和测量，而非直觉或权威。研究者从理论出发，从中推导出假设，并设计研究来检验该假设。假设必须被操作化，即变量需要明确的操作定义，以便被客观地测量。
A key principle is falsifiability. A useful hypothesis is one that can, in principle, be disproved. If the results do not match the predictions, the theory may be revised or rejected. This cycle of testing and revision is what makes psychology scientific.
一个关键原则是可证伪性。一个有用的假设在原则上必须有可能被证伪。如果结果与预测不符，理论就可以被修正或推翻。这种检验与修订的循环正是心理学的科学性所在。
2. Laboratory Experiments | 实验室实验
The laboratory experiment is the classic “scientific” method in psychology. The researcher deliberately manipulates the independent variable (IV) and measures its effect on the dependent variable (DV). Extraneous variables are controlled through standardised procedures and the random allocation of participants to different conditions.
实验室实验是心理学中经典的”科学”方法。研究者有意操纵自变量（IV），并测量其对因变量（DV）的影响。额外变量通过标准化程序以及将参与者随机分配到不同条件来控制。
For example, Milgram (1963) studied obedience by manipulating the proximity of the authority figure in a controlled lab setting. The high level of control allowed him to infer that situational factors, rather than participant personality, caused the obedience shown.
例如，米尔格拉姆（1963）在受控的实验室环境中通过操纵权威人物的接近程度来研究服从。高度的控制使他能够推断：导致服从行为的是情境因素，而非参与者的个性。
The main strength is high internal validity, because confounding variables are minimised and cause-effect relationships can be inferred. The main limitations are that artificial tasks and settings reduce ecological validity, and participants may display demand characteristics if they guess the research aim.
其主要优点是内部效度高，因为混杂变量被减至最少，可以推断因果关系。主要局限是人为的任务和环境降低了生态效度，而且参与者如果猜到研究目的，可能表现出需求特征。
3. Field and Natural Experiments | 现场实验与自然实验
Field experiments keep the researcher’s manipulation of the IV but take place in natural settings. Participants may not know they are being studied, which reduces demand characteristics. For instance, a researcher could test helping behaviour by dropping papers in a busy street and observing who stops to help.
现场实验保留研究者对自变量的操纵，但在自然环境中进行。参与者可能不知道自己正被研究，从而减少了需求特征。例如，研究者可以在繁忙的街道上掉下纸张，观察谁会停下来帮忙，以此测试助人行为。
Natural experiments, by contrast, involve no manipulation at all. The researcher measures the effect of an event that happens naturally. For example, comparing children’s screen time before and after a new law cannot ethically be manipulated, so a natural experiment is the only feasible option.
相比之下，自然实验完全不涉及操纵。研究者测量一个自然发生的事件所带来的影响。例如，比较新法律颁布前后儿童的屏幕时间，这在伦理上无法操纵，因此自然实验是唯一可行的选择。
Field experiments have high ecological validity but lose control over extraneous variables. Natural experiments share this limitation and also lack random allocation, so claims about cause and effect are weaker.
现场实验生态效度高，但对额外变量的控制较弱。自然实验同样有此局限，而且没有随机分配，因此关于因果关系的论断较弱。
4. Quasi-Experiments | 准实验
Quasi-experiments resemble true experiments, but the IV is a pre-existing difference such as age, gender, or a psychological disorder. The researcher compares groups that already exist rather than creating them through random allocation.
准实验与真正的实验相似，但自变量是既有差异，如年龄、性别或心理障碍。研究者比较已经存在的组别，而不是通过随机分配来创建组别。
This design is essential when random assignment is impossible or unethical, for example when comparing the memory of amnesic patients with that of healthy controls. However, the groups may differ in many confounding ways, so we cannot confidently claim that the IV caused the observed difference.
当随机分配不可能或不道德时，这类设计极为必要。例如，比较失忆症患者与健康对照组的记忆。然而，这些组别可能在许多混杂因素上存在差异，因此我们无法自信地声称自变量导致了所观察到的差异。
5. Observational Methods | 观察法
Observational methods record behaviour directly rather than asking people about it. Observations may be structured, using a coding scheme for predefined behaviours, or unstructured, recording everything that seems relevant. Time sampling and event sampling are used to manage what is recorded.
观察法直接记录行为，而不是询问人们。观察可以是结构化的，即使用编码方案记录预先定义的行为；也可以是非结构化的，即记录一切看似相关的内容。时间取样和事件取样用于管理记录内容。
Observation can be overt, where participants know they are being watched, or covert, where they do not. Covert observation avoids demand characteristics but raises ethical concerns about consent and privacy. Researchers should calculate inter-observer reliability to ensure that different observers record behaviour consistently.
观察可以是公开的，即参与者知道被观察；也可以是隐蔽的，即他们不知道。隐蔽观察避免了需求特征，但带来知情同意与隐私方面的伦理问题。研究者应计算观察者间信度，以确保不同观察者的记录具有一致性。
6. Self-Report Methods: Questionnaires | 自我报告法：问卷
Questionnaires are a widely used self-report method. Closed questions, such as Likert scales (1 = strongly disagree, 5 = strongly agree), produce quantitative data that is easy to analyse. Open questions produce qualitative data that is richer but harder to summarise.
问卷是广泛使用的自我报告法。封闭式问题，如李克特量表（1 = 非常不同意，5 = 非常同意），产生易于分析的定量数据。开放式问题产生更丰富但较难概括的定性数据。
The main weaknesses are social desirability bias, where participants answer in ways that make them look good, and response bias, such as always choosing the middle option. Leading questions and ambiguous wording also reduce validity. Researchers should pilot the questionnaire and use standardised instructions.
主要缺点是社会赞许偏差，即参与者以让自己看起来更好的方式作答，以及反应偏差，如总是选择中间选项。引导性问题和模糊措辞也会降低效度。研究者应试测问卷并使用标准化的指导语。
7. Self-Report Methods: Interviews | 自我报告法：访谈
Interviews allow face-to-face or remote exploration of participants’ experiences. A structured interview uses a fixed list of questions, ensuring consistency but limiting flexibility. An unstructured interview is more like a conversation, guided by the participant’s answers.
访谈允许面对面或远程地探索参与者的经验。结构化访谈使用固定的问题列表，确保一致性但限制了灵活性。非结构化访谈更像对话，由参与者的回答来引导进程。
Semi-structured interviews balance both approaches, using a topic guide while leaving room for probing. Interviews can access rich, subjective meaning, but transcription is time-consuming. The interviewer’s tone, phrasing, or body language may bias responses, producing interviewer bias.
半结构化访谈兼顾两者，使用主题指南，同时留下追问空间。访谈能获得丰富的主观意义，但转录非常耗时。访谈者的语气、措辞或肢体语言可能使回答产生偏差，即访谈者偏差。
8. Correlational Studies | 相关研究
Correlational studies measure the strength and direction of the relationship between two co-variables without manipulating either one. A positive correlation means the two variables increase together; a negative correlation means one increases while the other decreases.
相关研究在不操纵任何变量的情况下，测量两个共变量之间关系的强度与方向。正相关意味着两个变量共同增加；负相关意味着一个增加而另一个减少。
Interpreting correlation coefficients | 相关系数的解读
| r value | r 值 | Interpretation | 解释 |
|---|---|
| +0.80 to +1.00 / −0.80 to −1.00 | Strong correlation | 强相关 |
| +0.40 to +0.79 / −0.40 to −0.79 | Moderate correlation | 中等相关 |
更多咨询请联系16621398022（同微信）
CommentsMore posts |
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