📚 Year 12 Cambridge Psychology: Experiments & Practical Assessment Essentials | 剑桥12年级心理学：实验与实践考核要点
Experiments lie at the heart of the Cambridge AS Psychology syllabus, forming the core of Paper 2: Research Methods. In both written exams and practical assessments, you must demonstrate the ability to design, conduct, analyse, and critically evaluate experimental investigations. The key to success is mastering the essential components: from formulating testable hypotheses and identifying variables to selecting appropriate designs, adhering to ethical guidelines, and applying statistical reasoning. This guide breaks down the must-know topics for your Year 12 experimental and practical assessments, helping you approach exam questions and lab-based tasks with confidence.
实验是剑桥 AS 心理学教学大纲的核心，构成了试卷 2（研究方法）的基础。无论是笔试还是实践考核，你都需要展示设计、实施、分析和批判性评估实验研究的能力。成功的关键在于掌握这些基本要素：从提出可检验的假设、识别变量，到选择恰当的设计、遵守伦理准则以及运用统计推理。这份指南将为你拆解 12 年级实验与实践考核的必知要点，帮助你自信地应对考试题目和实验室任务。
1. Understanding Experiments in Psychology | 理解心理学中的实验
An experiment is the only research method that allows psychologists to establish cause-and-effect relationships between variables. In an experiment, the researcher deliberately manipulates one factor (the independent variable) while holding all other conditions constant, and then measures the effect on another factor (the dependent variable). This level of control enables researchers to infer that changes in the IV directly cause changes in the DV, provided extraneous variables are properly managed.
实验是唯一能够让心理学家确立变量间因果关系的研究方法。在实验中，研究者刻意操纵一个因素（自变量），同时保持所有其他条件不变，然后测量它对另一个因素（因变量）的影响。这种控制水平使得研究者能够推断，只要额外变量得到妥善管理，自变量的变化就会直接导致因变量的变化。
However, not all independent variables can be directly manipulated – for instance, a participant's gender or natural disaster experience. In such cases, quasi-experiments are used, where the IV is not manipulated by the researcher but still allows comparison between pre-existing groups. Understanding the degree of control available is fundamental to choosing the right experimental approach.
然而，并非所有自变量都能被直接操纵——例如被试的性别或自然灾害经历。在这种情况下，就会采用准实验，此时自变量不由研究者操纵，但仍可对现有组间进行比较。理解可用的控制程度是选择正确实验方法的基础。
2. Formulating Hypotheses | 提出假设
Every experiment begins with a clear, testable hypothesis. A hypothesis is a precise, testable statement predicting the expected outcome of an investigation. Cambridge examinations differentiate between a directional (one-tailed) hypothesis, which predicts the specific direction of the effect (e.g., 'Participants who consume 200 mg of caffeine will recall more words than those who consume 0 mg'), and a non-directional (two-tailed) hypothesis, which merely states that a difference will exist without specifying its direction (e.g., 'There will be a difference in word recall between participants who consume caffeine and those who do not').
每个实验都始于一个清晰、可检验的假设。假设是一种精确、可检验的陈述，预测研究的预期结果。剑桥考试区分了定向（单尾）假设和非定向（双尾）假设：前者预测效应的具体方向（例如，“摄入 200 毫克咖啡因的被试会比摄入 0 毫克的被试回忆出更多的单词”），后者仅陈述存在差异而不指明方向（例如，“摄入咖啡因与不摄入咖啡因的被试在单词回忆数量上存在差异”）。
For inferential statistics you will also work with the null hypothesis, which states that any observed effect is due to chance and that no true relationship exists between the IV and DV. In your practical tasks, you must be able to write operationalised directional and null hypotheses that clearly define how variables are measured.
在进行推断统计时，你还会用到零假设，它陈述任何观察到的效应都是由偶然因素造成的，自变量与因变量之间不存在真正的关联。在实践任务中，你必须能够写出操作化的定向假设和零假设，明确定义变量的测量方式。
3. Variables: IV, DV, and Control | 变量：自变量、因变量与控制变量
A well-designed experiment rests on the precise operationalisation of its variables. Use the table below to review the key variable types that appear in every examination scenario.
一个设计良好的实验依赖于变量的精确操作化。请使用下表复习每个考试情境中都会出现的关键变量类型。
| Term / 术语 | Definition & Example / 定义与示例 |
|---|---|
| Independent Variable (IV) / 自变量 | The factor manipulated by the researcher to create different conditions. E.g., noise level (silence vs. 80 dB). 研究者操纵以形成不同条件的因素。例如，噪音水平（安静 vs. 80 分贝）。 |
| Dependent Variable (DV) / 因变量 | The factor measured to assess the effect of the IV. E.g., number of words correctly recalled. 为评估自变量效应而测量的因素。例如，正确回忆的单词数量。 |
| Confounding Variable / 混淆变量 | A variable that varies systematically with the IV and provides an alternative explanation for the results. E.g., time of day could confound a memory experiment if all caffeine group tested in the morning. 与自变量发生系统性共变并为结果提供替代解释的变量。例如，若所有咖啡因组都在上午测试，一天中的时间段就会混淆记忆实验的结果。 |
| Control Variable / 控制变量 | Any factor kept constant to prevent it from influencing the DV. E.g., same word list, same room temperature. 为阻止其对因变量造成影响而保持恒定的任何因素。例如，相同的单词表、相同的室温。 |
When evaluating an experiment described in a stimulus material, always identify whether the researcher has successfully controlled for confounding variables and standardised the procedure. This directly links to the study's internal validity.
在评估题干材料中描述的实验时，务必判断研究者是否成功控制了混淆变量并标准化了流程。这一点直接关联到研究的内部效度。
4. Experimental Designs: Independent Groups, Repeated Measures, Matched Pairs | 实验设计：独立组、重复测量与匹配对
Once your variables are defined, you must decide how to allocate participants to conditions. Cambridge Psychology expects you to know the three classic experimental designs, their strengths, and their limitations. The table below summarises the essentials.
定义变量之后，你需要决定如何将参与者分配到各个条件。剑桥心理学要求你了解三种经典的实验设计及其优缺点。下表总结了关键信息。
| Design / 设计 | Description / 描述 | Key Strength / 主要优势 | Key Limitation / 主要局限 |
|---|---|---|---|
| Independent Groups / 独立组 | Different participants in each condition. 每个条件使用不同的被试。 | No order effects; participants are naive to the other condition. 无顺序效应；被试对另一条件不知情。 | Participant variables (individual differences) may bias results; requires more participants. 被试变量（个体差异）可能使结果产生偏差；需要更多被试。 |
| Repeated Measures / 重复测量 | Same participants take part in all conditions. 相同的被试参与所有条件。 | No participant variables; fewer participants needed. 无被试变量；需要较少被试。 | Order effects (practice, fatigue) can distort performance; counterbalancing is required. 顺序效应（练习、疲劳）可能扭曲表现；需要平衡顺序。 |
| Matched Pairs / 匹配对 | Participants matched on key traits, then split into conditions. 被试按关键特质配对，再分配到不同条件。 | Reduces participant variables without order effects. 减少被试变量，且无顺序效应。 | Time-consuming; perfect matching is impossible; can lose participants if one drops out. 耗时；无法完美匹配；若一人退出可能损失数据。 |
In your experimental write-up, you must justify your choice of design and explicitly state how you minimised its weaknesses – for instance, by using counterbalancing with repeated measures or random allocation with independent groups.
在你的实验报告中，你必须对你选择的设计给出理由，并明确说明你是如何弱化其缺陷的——例如，在重复测量中使用平衡抵消法，或在独立组中使用随机分配。
5. Types of Experiments: Lab, Field, Natural, Quasi | 实验类型：实验室、现场、自然与准实验
Not all experiments take place in a sterile laboratory. Cambridge candidates must be able to distinguish between laboratory, field, natural, and quasi-experiments, and evaluate how the setting and control affect the reliability and applicability of findings.
并非所有实验都在无菌的实验室中进行。剑桥考生必须能够区分实验室实验、现场实验、自然实验和准实验，并评价实验环境和控制水平如何影响研究结果的可靠性和应用性。
Laboratory experiments are conducted in a highly controlled, artificial environment. They offer excellent internal validity because the researcher can manipulate the IV precisely and hold extraneous variables constant. However, the artificial setting may lead to low ecological validity – behaviour may not reflect real life.
实验室实验在高度受控的人造环境中进行。它们具有出色的内部效度，因为研究者可以精确操纵自变量并控制额外变量。然而，人造环境可能导致生态效度低——行为可能无法反映真实生活。
Field experiments take place in a natural setting, but the IV is still deliberately manipulated. This provides higher ecological validity while maintaining a degree of control. However, researchers have less control over extraneous variables and ethical issues (e.g., lack of informed consent) may arise.
现场实验在自然环境中进行，但自变量仍然被刻意操纵。这提供了更高的生态效度，同时保持了一定的控制。但研究者对额外变量的控制较弱，且可能产生伦理问题（例如缺乏知情同意）。
Natural experiments naturally occurring IV (e.g., a flood) where the researcher simply measures the DV on exposed vs. non-exposed groups. This allows the study of variables that cannot be ethically manipulated, but random allocation is impossible, so cause-and-effect conclusions are weaker.
自然实验的情况下，自变量是自然发生的（例如洪水），研究者仅测量暴露组和非暴露组的因变量。这允许研究那些无法进行伦理操纵的变量，但无法实施随机分配，因此因果推断较弱。
Quasi-experiments contain a naturally varying IV that is a characteristic of the participant (e.g., age, diagnosis). These are common in clinical and developmental psychology. As with natural experiments, lack of random assignment weakens internal validity.
准实验包含一个自然变化且属于被试个人特征的自变量（如年龄、诊断）。这在临床和发展心理学中很常见。与自然实验一样，缺少随机分配削弱了内部效度。
6. Sampling Methods | 取样方法
The participants (or sample) you select directly influence how far your results can be generalised. For Cambridge practical assessments, you must be able to describe and evaluate opportunity, random, stratified, systematic, and volunteer sampling.
你所选择的被试（样本）直接影响你的结果在多大程度上可以推广。在剑桥实践考核中，你必须能够描述并评价机会取样、随机取样、分层取样、系统取样和志愿者取样。
Opportunity sampling uses people who are readily available, such as students in a common room. It is quick and convenient, but often unrepresentative because it may draw from a single, narrow demographic. Random sampling, where every member of the target population has an equal chance of selection, is the ideal for minimising bias, yet it is difficult to achieve in school-based studies without a complete population list.
机会取样使用那些容易接触到的人，例如休息室里的学生。它快速且方便，但通常缺乏代表性，因为它可能只来自单一、狭隘的人口群体。随机取样是指目标人群中的每一个成员都有平等的被选中的机会，它能够最大程度地减少偏见，但在学校研究中，如果没有完整的人口名单，就很难实现。
Stratified sampling involves identifying subgroups in the population and then sampling proportionally from each stratum. This yields a representative sample, but it requires detailed knowledge of the population structure. Systematic sampling uses a fixed interval (e.g., every 5th name on a list) and can be efficient, though bias can occur if the list order contains a hidden pattern. Volunteer sampling relies on participants approaching the researcher; it can attract motivated individuals but often results in a biased, self-selected sample.
分层取样需要先识别出人群中的亚群，然后从每个层次中按比例取样。这可以得到具有代表性的样本，但需要详细了解人口结构。系统取样采用固定的间隔（例如，名单上每第 5 个人），效率很高，但如果名单顺序中隐藏着某种模式，就可能产生偏差。志愿者取样依赖于被试主动联系研究者；它可以吸引到积极性高的个体，但常常导致一个有偏差的、自我选择的样本。
7. Ethical Considerations | 伦理考量
Ethics are non-negotiable in psychological research. Every practical investigation you conduct or evaluate must adhere to the principles laid out by the British Psychological Society (BPS): respect, competence, responsibility, and integrity. In Year 12, you are expected to embed ethical thinking into your research design from the very start.
伦理在心理学研究中是不可妥协的。你所开展或评估的每一项实践调查都必须遵循英国心理学会（BPS）制定的原则：尊重、胜任、责任与正直。在 12 年级，你从一开始就要将伦理思维融入你的研究设计之中。
The core ethical issues you must address include obtaining informed consent from participants (or parental consent for those under 16), safeguarding their right to withdraw at any time, protecting them from physical and psychological harm, ensuring confidentiality of data, and providing a thorough debrief after the study. When it is impossible to gain full informed consent at the start – for instance, in a field experiment where participants are observed in public – you must justify why deception was minimal and explain how you will debrief participants afterwards.
你必须处理的核心伦理问题包括：获得被试的知情同意（16 岁以下需获得家长同意），保障他们随时退出的权利，保护他们免受身体和心理伤害，确保数据的保密性，以及在研究后提供全面的事后解释。当无法在研究一开始就获得完全知情同意时——例如，在公共场所观察被试的现场实验中——你必须说明欺骗为何是微乎其微的，并解释你事后将如何向被试作汇报解释。
In your exam answers, always suggest concrete actions: using signed consent forms, offering the right to withdraw data even after the study, screening for distress with a post-study questionnaire, replacing names with code numbers, and sharing the true aims during debrief. Showing this precise, actionable knowledge earns maximum marks.
在考试答案中，始终要提出具体的行动：使用签名的知情同意书，即使在研究后也允许被试撤回数据的权利，用研究后问卷筛查其不安情绪，用代码代替姓名，以及在事后解释中告知真实目的。展示这种精确、可操作的知识能为你赢得最高分。
8. Data Collection Techniques | 数据收集技术
While the experiment defines the overall method, the way you collect the DV data requires equal attention. Common techniques include direct observation (using behavioural categories and coding frames), self-report measures (questionnaires with Likert scales, structured interviews), and physiological recordings (e.g., heart rate monitor). Each technique must be operationalised into quantifiable data.
虽然实验确定了整体方法，但你收集因变量数据的方式同样需要关注。常见的技术包括直接观察（使用行为类别和编码框架）、自我报告测量（带里克特量表的问卷、结构化访谈）以及生理记录（例如心率监测）。每一项技术都必须被操作化，转化为可量化的数据。
When using observation, create clearly defined behavioural categories that are mutually exclusive – for instance, 'smiling', 'frowning', 'neutral expression'. This allows for reliable inter-rater reliability checks. For questionnaires, closed questions produce quantitative data that is easy to analyse, whereas open questions yield rich qualitative data but are harder to quantify. In your practical design, justify your choice of data collection technique by linking it directly to your hypothesis and the type of data you need (e.g., interval/ratio data for parametric tests).
在使用观察法时，要创建明确且互斥的行为类别——例如，“微笑”、“皱眉”、“中性表情”。这便于进行可靠的评分者间信度检验。对于问卷，封闭式问题能产生易于分析的定量数据，而开放式问题则可以获得丰富的定性数据，但量化难度更高。在你的实践设计中，要通过直接联系你的假设和所需的数据类型（例如，用于参数检验的等距/比率数据）来证明你选择该数据收集技术的合理性。
9. Descriptive and Inferential Statistics | 描述与推断统计
Numerical data is only meaningful when properly summarised and tested for significance. Descriptive statistics give you a snapshot of your data set: measures of central tendency (mean, median, mode) tell you where the centre lies, while measures of dispersion (range, standard deviation) indicate how spread out the scores are.
只有经过恰当的概括和显著性检验，数字数据才有意义。描述统计为你提供了一幅数据集的快照：集中趋势测量（平均数、中位数、众数）告诉你中心在哪里，而离散测量（全距、标准差）则显示了分数的分散程度。
For the Cambridge exam, you must be able to calculate basic descriptive statistics by hand and interpret the output. Standard deviation is particularly valued because it gives a precise measure of variability that feeds into many inferential tests. In a practical assessment, presenting your data in a well-labelled bar chart or histogram and commenting on the mean and standard deviation is a fundamental skill.
在剑桥考试中，你必须能够手工计算基本的描述统计量，并解释计算结果。标准差尤其受重视，因为它提供了精确的变异度测量，许多推断检验都是以此为基础的。在实践评估中，将你的数据用标注清晰的条形图或直方图呈现出来，并对平均数和标准差进行评论，这是一项基本技能。
ide whether your results are genuine or simply due to chance. In Year 12, you will learn to select an appropriate test – Chi-squared for nominal data, Spearman' l/correlational data, or the Mann-Whitney U and Wilcoxon tests for ordinal difference data – ignificance. Never forget to relate your statistical conclusion back to the original hypothesis: did you reject the null hypothesis, and what does that mean psychologically?
推断统计帮助你判断你的结果是真实的还是仅仅出于偶然。在 12 年级，你将学习如何选择恰当的检验方法——名义数据用卡方检验，顺序/相关数据用斯皮尔曼等级相关系数，顺序差异数据用曼-惠特尼 U 检验或威尔科克森检验——并且会解释 p ≤ 0.05 的显著性水平。切勿忘记将你的统计结论联系回最初的假设：你是否拒绝了零假设？这在心理学上意味着什么？
10. Evaluating Experiments: Validity, Reliability, and Generalisability | 评估实验：效度、信度与推广性
Critical evaluation is the highest-order skill assessed in Cambridge Psychology. When you appraise an experiment – whether your own design or a published study – you must systematically consider its internal validity, external validity (including ecological and population validity), reliability, and generalisability.
批判性评价是剑桥心理学评估的最高阶技能。当你评估一个实验时——无论是你自己设计的还是已发表的研究——你必须系统地考量它的内部效度、外部效度（包括生态效度和人群效度）、信度和推广性。
Internal validity asks whether the IV truly caused the change in the DV. Threats include confounding variables, demand characteristics, and researcher bias. Reliability refers to consistency: if the experiment were replicated, would you obtain similar results? Standardised procedures and clear operational definitions greatly improve reliability. External validity concerns the extent to which findings can be generalised beyond the experimental context. A lab experiment may lack ecological validity, while a narrow sample of university students limits population validity.
内部效度询问的是，自变量的改变是否真正导致了因变量的变化。潜在威胁包括混淆变量、需求特征和研究者偏见。信度指的是结果的一致性：如果复制这个实验，你会得到相似的结果吗？标准化的流程和清晰的操作化定义能够极大地提升信度。外部效度关注的是研究结果在多大程度上可以推广到实验情境之外。实验室实验可能缺乏生态效度，而仅针对大学新生的狭窄样本则会限制人群效度。
In your exam or practical report, never just list these terms. Provide specific, evidence-based recommendations: 'To improve ecological validity, the experiment could be repeated as a field experiment in a busy café,' or 'To reduce demand characteristics, a single-blind procedure could be used, and the true hypothesis only revealed during debrief.' Such thoughtful, applied evaluation earns top marks.
在考试或实践报告中，绝不要只是罗列这些术语。要提供基于证据的具体建议：“为了提高生态效度，这项实验可以作为一个现场实验在一间热闹的咖啡馆中进行复证。”或者“为减少需求特征，可以采用单盲程序，仅在事后解释时揭示真实假设。”这种深思熟虑、注重应用的评价才能帮你斩获高分。
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