📚 Psychology Research Methods: Key Approaches and Applications | 心理学考点：主要研究方法及其应用
Research methods are the backbone of psychological science. Every theory, every finding and every clinical guideline in psychology ultimately rests on the way data were collected, measured and interpreted. For A-level and IB students, mastering the main research methods is not just an exam requirement; it is the foundation for critical thinking about any psychological claim you encounter.
研究方法是心理科学的基石。心理学中的每一个理论、每一项发现、每一条临床指南，最终都取决于数据的收集、测量与解读方式。对于 A-level 和 IB 考生而言，掌握主要研究方法不仅是考试要求，更是你批判性思考任何心理学论断的基础。
1. Experimental Methods | 实验法
Experiments are the most rigorous way to establish cause and effect. In a true experiment, the researcher manipulates the independent variable (IV) while controlling extraneous variables, then measures the effect on the dependent variable (DV). For example, to test whether caffeine improves memory, you might give one group a caffeinated drink and the other a placebo, then compare their recall scores on a standardised test.
实验法是确立因果关系最严谨的方法。在真实验中，研究者操纵自变量（IV），同时控制额外变量，然后测量对因变量（DV）产生的影响。例如，要测试咖啡因是否能改善记忆，你可以给一组被试含咖啡因的饮品，另一组给安慰剂，然后比较两组在标准化测验中的回忆成绩。
Laboratory experiments offer high internal validity because conditions are tightly controlled. However, the artificial setting can trigger demand characteristics, where participants guess the aim and change their behaviour, thereby reducing ecological validity. Field experiments trade some control for realism, while natural experiments use existing conditions, meaning the researcher cannot randomly allocate participants to conditions.
实验室实验由于条件受到严格控制，具有较高的内部效度。然而，人为环境可能引发需求特征——被试猜测实验目的后改变行为——从而降低生态效度。现场实验以部分控制换取现实性，而自然实验利用既有条件，研究者无法将被试随机分配到不同条件中。
- Strengths: high control, replicable, clear cause-effect conclusions | 优点：控制严格、可复现、结论因果关系明确
- Limitations: artificial, demand characteristics, ethical limits on manipulation | 局限：人为性强、存在需求特征、操纵变量受伦理限制
2. Correlational Studies | 相关研究
Correlational studies examine the strength and direction of the relationship between two co-variables. The result is expressed as a correlation coefficient, r, which ranges from -1.00 through 0 to +1.00. A positive value means both variables increase together, while a negative value means one increases as the other decreases. A value near 0 indicates little or no relationship.
相关研究考察两个共变量之间关系的强度与方向。结果用相关系数 r 表示，取值范围从 -1.00 经过 0 到 +1.00。正值意味着两个变量同增同减，负值意味着一个变量增大时另一个减小，趋于 0 则表示几乎没有关系。
Correlation does not imply causation. Two variables may be correlated because a third variable influences both, a problem called the third-variable problem. For example, ice cream sales and drowning incidents both rise in summer; the hidden variable is warm weather, not ice cream causing drownings. In exams, you must be able to identify this flaw and suggest the possible third variable.
相关并不等于因果。两个变量之所以相关，可能是因为第三个变量同时影响着它们，这就是第三变量问题。例如，冰淇淋销量和溺水事件都在夏季上升；背后的隐藏变量是炎热天气，而不是冰淇淋导致溺水。在考试中，你必须能识别这一缺陷并推测可能的第三变量。
r = +1.00 perfect positive | r = -1.00 perfect negative | r = 0 no correlation
3. Observational Methods | 观察法
Observation involves watching and recording behaviour without intervening. In naturalistic observation, behaviour is studied in the participants’ everyday environment, such as observing playground aggression at a school. In controlled observation, the researcher sets up a situation in a laboratory or structured setting to elicit specific behaviours. Both can be conducted with or without the participants’ knowledge of being watched.
观察法是在不干预的情况下观看并记录行为。在自然观察中，研究在被试的日常环境中进行，例如在学校操场上观察攻击行为。在控制观察中，研究者在实验室或结构化情境中设置场景以引发特定行为。两种方式都可以在被试知情或不知情的情况下进行。
A key decision is whether to use participant or non-participant observation. Participant observers join the group being studied, which increases depth but risks losing objectivity. Non-participant observers remain apart, which preserves objectivity but may miss subtle interactions. Behavioural categories must be clearly defined before recording, and inter-observer reliability should be checked to ensure consistency.
一个关键决策是采用参与式还是非参与式观察。参与式观察者加入被研究群体，能加深理解，但可能丧失客观性。非参与式观察者保持距离，能维护客观性，但可能错过微妙的互动。记录前必须明确定义行为类别，并检查观察者间信度以保证一致性。
4. Surveys and Questionnaires | 调查与问卷法
Surveys collect data from large samples using questionnaires or polls. Questions can be closed, offering fixed response options such as Likert scales (strongly agree to strongly disagree), or open, allowing participants to answer freely. Closed questions produce quantitative data that is easy to analyse statistically; open questions yield rich qualitative data but require coding and thematic analysis.
调查法通过问卷或民意测验从大样本中收集数据。问题可以是封闭式的，提供固定选项，如李克特量表（从非常同意到非常不同意）；也可以是开放式的，允许被试自由作答。封闭式问题产生易于统计分析的定量数据；开放式问题产生丰富的定性数据，但需要编码和主题分析。
Questionnaires are efficient and can reach thousands of participants at low cost. However, they rely on self-report, which is vulnerable to social desirability bias—participants answering in ways they believe are socially acceptable rather than truthful. Response bias, such as always selecting the midpoint, can also distort results. Careful question wording and anonymous administration help reduce these problems.
问卷效率高，能以低成本触达成千上万的被试。然而，问卷依赖自我报告，容易受到社会赞许偏差的影响——被试倾向于按社会认可的方式作答而非如实回答。反应偏差，比如总选中间值，也会扭曲结果。仔细措辞和匿名施测有助于减少这些问题。
5. Case Studies | 个案研究
A case study is an in-depth investigation of a single individual, a small group or a specific institution. Researchers typically combine interviews, observations, medical records and psychological tests to build a detailed picture. Classic examples include Phineas Gage, whose personality changed after a metal rod damaged his frontal lobes, and Clive Wearing, who lost the ability to form new long-term memories.
个案研究是对单个个体、小群体或特定机构进行的深入调查。研究者通常结合访谈、观察、病历和心理测验来构建完整图景。经典案例包括菲尼亚斯·盖奇——铁棍损伤额叶后性格大变；以及克莱夫·韦尔林——他失去了形成新长时记忆的能力。
Case studies adopt an idiographic approach, focusing on what is unique about the person rather than generalising across populations. Their strength lies in generating hypotheses and providing rich detail that nomothetic studies may miss. However, results cannot easily be generalised, and the researcher’s subjective interpretation can bias conclusions. Replication is rarely possible, limiting reliability.
个案研究采用个别化方法（idiographic approach），关注个体的独特性而非跨人群概括。其优势在于生成假设，并提供大规模研究可能遗漏的丰富细节。然而，结果难以推广，研究者的主观解释也可能影响结论。复现几乎不可能，限制了信度。
6. Interviews | 访谈法
Interviews allow researchers to gather detailed verbal data through direct interaction. A structured interview uses a fixed set of identical questions for every participant, ensuring consistency and making comparisons easy. An unstructured interview starts with a general topic and follows the participant’s responses freely, producing deep but hard-to-compare data. A semi-structured interview sits in between, with preplanned core questions plus room for follow-up probes.
访谈法通过直接互动收集详细的口头数据。结构化访谈对每位被试使用一套完全相同的问题，保证一致性并便于比较。非结构化访谈以一个宽泛主题开始，自由跟随被试的回答，能产生深入但难以比较的数据。半结构化访谈介于两者之间：有预先设计的核心问题，也留有追问的空间。
Interviews are especially useful for studying sensitive topics such as trauma or sexuality, where a skilled interviewer can build rapport and encourage honest disclosure. The main drawbacks are interviewer bias—subtle cues from tone or body language—and the heavy time cost of transcribing and coding verbal data. Training interviewers and piloting the schedule improve quality.
访谈特别适合研究创伤、性行为等敏感话题，训练有素的访谈者可以建立融洽关系并鼓励坦诚作答。主要缺点是访谈者偏差——语气或肢体语言的微妙暗示——以及转录和编码言语数据所需的大量时间成本。培训访谈者并进行预测试有助于提高质量。
7. Reliability and Validity | 信度与效度
Reliability refers to consistency in measurement. Test-retest reliability checks whether the same test gives the same scores on two separate occasions; split-half reliability checks whether two halves of a test produce similar results; inter-rater reliability measures the agreement between different observers. If reliability is low, the data are unstable and conclusions become questionable. The correlation coefficient is often used to report reliability, with values above +0.80 considered acceptable in many A-level specifications.
信度指测量的一致性。重测信度检验同一测验在两个不同时间点是否给出相同分数；分半信度检验测验两半是否产生相似结果；评分者间信度衡量不同观察者之间的一致性。如果信度低，数据就不稳定，结论也就值得怀疑。信度常用相关系数报告，在多数 A-level 考纲中，高于 +0.80 的值被视为可接受。
Validity concerns whether the research measures what it claims to measure. Internal validity asks whether the IV really caused the change in the DV; external validity asks whether findings generalise beyond the study context. Ecological validity, a subtype of external validity, asks whether results apply to real-life settings. Population validity asks whether results apply to other groups of people. To improve validity, researchers use standardised procedures, counterbalancing, single-blind or double-blind designs, and pilot studies.
效度关乎研究是否测量了它声称要测量的内容。内部效度问的是自变量是否真的引起了因变量的变化；外部效度问的是结果能否推广到研究情境之外。生态效度是外部效度的一种，追问结果是否适用于真实生活场景；总体效度追问结果是否适用于其他人群。为提升效度，研究者会采用标准化程序、抵消平衡、单盲或双盲设计以及预测试。
8. Sampling Methods | 抽样方法
Since psychologists cannot study entire populations, they select samples. Random sampling gives every member of the population an equal chance of selection, which minimises bias, but is often impractical. Opportunity sampling takes whoever is available, as when a researcher asks students in a corridor to participate; it is convenient but prone to bias. Stratified sampling divides the population into subgroups based on key characteristics, then selects proportionally from each, improving representativeness. Volunteer (self-selected) sampling relies on participants who respond to adverts, which attracts motivated individuals who may not reflect the wider population.
由于心理学家无法研究整个总体，他们需要选择样本。随机抽样使总体中的每个成员有均等机会被选中，能最大程度减少偏差，但往往不切实际。机会抽样选取任何可用的人，比如研究者在走廊里请学生参与；这种方式方便，但容易产生偏差。分层抽样先按关键特征将总体分为若干子群体，再从各层按比例抽取，从而提高代表性。志愿者（自我选择）抽样依赖回应广告的被试，这类被试积极性高，但可能不代表更广的人群。
Sample size matters too. Small samples increase the risk of extreme or unrepresentative results, and reduce statistical power. However, large samples are no guarantee of validity if the sampling method is biased. In exam answers, always connect the sampling technique to its consequences for generalising the findings.
样本量也很重要。小样本增加极端或不具代表性结果的风险，并降低统计检验力。然而，大样本并不能保证效度——如果抽样方法本身有偏。在考试作答中，始终要把抽样技术与结果推广性的后果联系起来。
9. Ethical Considerations | 伦理考量
Psychological research must protect participants from harm. The British Psychological Society (BPS) guidelines require informed consent, meaning participants agree to take part after being told about the study’s nature. Deception must be avoided unless fully justified and followed by a thorough debriefing. Participants have the right to withdraw at any time and to have their data kept confidential and anonymous. Researchers must also protect participants from physical and psychological distress during and after the study.
心理学研究必须保护被试免受伤害。英国心理学会（BPS）指南要求知情同意，即被试在了解研究性质后同意参与。除非有充分理由，否则必须避免欺骗，且事后要进行完整的汇报说明。被试有权随时退出，并有权要求数据保密和匿名。研究者还必须在研究过程中及之后保护被试免受身心痛苦。
In the exam, you may be asked to evaluate a study ethically or to suggest how a flawed design could be improved. A good answer identifies the specific ethical issue, names the relevant guideline, and proposes a practical remedy—such as replacing deception with a cover story, adding a debrief session, or offering counselling for participants in stressful studies.
考试中，你可能会被要求从伦理角度评价一项研究，或提出改进有缺陷设计的建议。好的答案应指明具体的伦理问题，说出相关指南名称，并提出切实可行的补救措施——例如用经过伦理审批的掩饰故事替代欺骗、增加事后说明环节，或为压力大的研究提供心理辅导。
10. Quantitative vs Qualitative Data | 定量与定性数据
Quantitative data are numerical and can be summarised with descriptive statistics such as the mean, median, mode, range and standard deviation. They allow inferential statistical tests to determine whether results are significant or due to chance, for example whether p < 0.05. Quantitative data are objective and easy to compare, but they can oversimplify complex behaviours and lose contextual meaning.
定量数据是数值型的，可用描述统计进行概括，如平均数、中位数、众数、全距和标准差。它们允许通过推断统计检验判断结果是显著还是由偶然因素造成，例如 p < 0.05。定量数据客观且易于比较，但可能过度简化复杂行为，丢失情境意义。
Qualitative data are verbal or visual, such as interview transcripts, diaries and open-ended responses. They are analysed using thematic analysis or content analysis, which identify recurring patterns and categories. Qualitative data offer depth and insight into individual experience, but they are subjective, harder to summarise, and prone to researcher bias during interpretation. In practice, many studies use a mixed-methods design to combine the strengths of both.
定性数据是言语或视觉形式的，如访谈转录稿、日记和开放性回答。它们常通过主题分析或内容分析来识别反复出现的模式和类别。定性数据能提供深度以及对个体经验的洞察，但具有主观性，难以概括，解读时也容易受到研究者偏差的影响。实践中，许多研究采用混合方法设计，结合两者的优势。
11. Descriptive and Inferential Statistics | 描述统计与推断统计
Descriptive statistics summarise the data collected. The mean is the arithmetic average, sensitive to extreme scores; the median is the middle value, robust to outliers; the mode is the most frequent score. Measures of dispersion include the range, which is simple but easily distorted by one extreme value, and the standard deviation, which reflects how spread out scores are around the mean. A smaller standard deviation means more consistent performance.
描述统计用于概括收集到的数据。平均数是算术平均值，对极端分数敏感；中位数是中间值，不受离群值影响；众数是最频繁出现的分数。离散程度的度量包括全距——简单但易被一个极端值扭曲；以及标准差——反映分数围绕平均数的分散程度。标准差越小，说明成绩越稳定。
Inferential statistics help researchers decide whether an observed difference is significant or simply due to chance. Significance tests generate a p-value; when p < 0.05, psychologists accept that there is less than a 5% probability the result occurred by chance, and therefore reject the null hypothesis. Choosing the correct statistical test depends on the experimental design, the level of measurement and whether the data meet parametric assumptions—a core skill in the research methods section of the exam.
推断统计帮助研究者判断观察到的差异是显著还是仅仅出于偶然。显著性检验产生 p 值；当 p < 0.05 时，心理学家认为结果由偶然因素造成的概率小于 5%，因此拒绝零假设。选择正确的统计检验取决于实验设计、测量水平以及数据是否满足参数假设——这是考试研究方法部分的核心技能。
12. Choosing the Right Method | 如何选择合适的研究方法
Selecting a method depends on the research question, the ethical constraints and practical resources. If you want to establish cause and effect, an experiment is the best choice. If the topic is sensitive or little is known, a qualitative interview or case study may be more appropriate. If the goal is to measure the strength of a relationship, correlation is suitable. If the goal is to describe behaviour in real life, naturalistic observation is preferred.
选择哪种方法取决于研究问题、伦理限制和实际资源。如果你想确立因果关系，实验法是最佳选择。如果话题敏感或知之甚少，访谈或个案研究可能更合适。如果目标是衡量两个变量关系的强度，适合用相关研究。如果目标是描述现实生活中的行为，自然观察是首选。
| Method | 方法 | Best for | 适用情境 | Key weakness | 主要弱点 |
| Experiment | 实验法 | Cause-effect claims | 因果关系 | Artificiality | 人为性 |
| Correlation | 相关研究 | Measuring relationships | 测量关系 | Cannot infer causation | 不能推断因果 |
| Observation | 观察法 | Natural behaviour | 自然行为 | Observer bias | 观察者偏差 |
| Survey | 问卷法 | Large samples, attitudes | 大样本、态度 | Social desirability bias | 社会赞许偏差 |
| Case study | 个案研究 | Rare conditions, depth | 罕见情况、深度 | Poor generalisation | 推广性差 |
In the exam, when asked to evaluate a method, always consider both strengths and limitations, and anchor each point in the specific context of the study described. Remember: method choice is a trade-off. No single method is perfect, and the best psychologists combine approaches or acknowledge the limitations of their chosen design with humility and rigour.
考试中，当被要求评价某种方法时，务必同时考虑优点和局限，并将每个要点锚定在所描述研究的具体情境中。记住：方法选择是一种权衡。没有任何单一方法是完美的，最优秀的心理学家会结合多种方法，或带着谦逊和严谨的态度承认所选设计的局限性。
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