📚 CAIE A-Level Psychology: Mastering Experimental & Practical Assessment | CAIE A-Level心理学：掌握实验与实践考核
In CAIE A-Level Psychology (9990), assessment of experimental and practical skills is woven into the examination papers, most notably Paper 2 (AS) and Paper 4 (A2) Section A. For Year 13 students, the ability to design, evaluate, and analyse psychological investigations is essential. Success requires a firm grasp of research methodology, from formulating hypotheses to selecting appropriate inferential statistics. This guide breaks down the key components and common pitfalls, equipping you to write high-scoring responses in the practical-based questions.
在 CAIE A-Level 心理学 (9990) 课程中，实验与实践技能的考核贯穿于试卷中，主要体现在试卷二（AS）和试卷四（A2）A 部分。对 Year 13 的学生而言，设计、评估和分析心理学研究的能力至关重要。取得高分需要牢牢掌握研究方法，从提出假设到选择合适的推断统计量。本指南拆解了关键组成部分和常见误区，帮助你在实践类题目中写出高分答案。
1. The Role of Practical Assessment in CAIE Psychology | 实践考核在CAIE心理学中的角色
Paper 4 Section A presents a novel research scenario and asks you to design a complete study. This section carries 30 marks and explicitly tests your ability to apply the research-methods toolkit you have built throughout the course. The exam expects precise operationalisation, justified design choices, and accurate data analysis planning.
试卷四的 A 部分会给出一个全新的研究情境，要求你设计一项完整的研究。该部分占 30 分，明确考察你运用整个课程所建立的研究方法工具包的能力。考试期望你给出精确的操作性定义、有充分理由的设计选择以及准确的数据分析规划。
2. Types of Experiments You Must Distinguish | 必须区分的实验类型
You should not only recognise but also be ready to justify laboratory, field, natural, and quasi-experiments. Each type brings a different trade-off between internal validity and ecological validity. A lab experiment offers high control but low mundane realism; a field experiment captures more natural behaviour at the cost of some control.
你不仅要能识别实验室实验、现场实验、自然实验和准实验，还要能为它们提供理由。每种类型都在内部效度和生态效度之间作出不同的权衡。实验室实验控制程度高，但世俗真实性低；现场实验能捕捉更自然的行为，但会牺牲一定控制。
| Type | IV Manipulation | Control Over EV | Ecological Validity |
|---|---|---|---|
| Lab | Direct | High | Low |
| Field | Direct | Moderate | High |
| Natural | No manipulation | Low | Very high |
| Quasi | Based on existing difference | Low | Variable |
表格：四类实验在自变量操控、额外变量控制和生态效度方面的对比。
3. Variables and the Art of Operationalisation | 变量与操作性定义的艺术
Every high-mark answer begins with a clearly identified independent variable (IV) and dependent variable (DV). The IV is the factor you manipulate or compare, while the DV is the measurable outcome. Operationalisation means turning abstract concepts into specific, replicable procedures or metrics. For example, ‘aggression’ could be operationalised as the number of times a participant delivers a loud noise burst to a confederate.
每一份高分答案都始于清晰识别的自变量（IV）和因变量（DV）。自变量是你操控或比较的因素，因变量是可测量的结果。操作性定义意味着将抽象概念转化为具体、可重复的程序或度量。例如，“攻击性”可以被操作定义为参与者向一名同谋者发送高分贝噪声的次数。
Do not forget to state at least two levels of the IV (e.g., caffeine vs. placebo) and explain exactly how the DV will be recorded. This precision signals strong methodological understanding to the examiner.
不要忘记说明自变量的至少两个水平（例如咖啡因组与安慰剂组），并准确解释因变量将如何记录。这种精准度向考官显示了你扎实的方法论理解。
4. Writing Directional and Non-directional Hypotheses | 撰写方向性与非方向性假设
Your hypothesis is the foundation of the study. An experimental/alternative hypothesis (H₁) predicts the relationship between IV and DV. It can be directional (one-tailed), stating the expected direction of difference, or non-directional (two-tailed), simply stating there will be a difference. A null hypothesis (H₀) must also be stated, predicting no significant difference or relationship.
你的假设是研究的基础。实验/备择假设（H₁）预测自变量与因变量之间的关系。它可以是有方向的（单尾），指明预期的差异方向，也可以是无方向的（双尾），只说明存在差异。还必须陈述零假设（H₀），预测不存在显著差异或关系。
Examiners reward the use of the exact phrasing ‘there will be a significant difference in [DV] between participants who [Level 1 of IV] and those who [Level 2 of IV]’. Always include the population and both IV levels.
考官会奖励使用准确措辞的答案，如“在[IV水平1]的参与者和[IV水平2]的参与者之间，[因变量]将存在显著差异”。始终要包含研究对象群体和IV的两个水平。
5. Choosing and Justifying an Experimental Design | 选择并论证实验设计
Independent groups, repeated measures, and matched pairs each carry unique strengths and limitations. Independent groups eliminate order effects but risk participant variables. Repeated measures require fewer participants and control individual differences, yet demand counterbalancing to manage order effects. Matched pairs offer a middle ground but are time-consuming.
独立组、重复测量和配对组各有独特的优势和局限。独立组消除了顺序效应，但面临参与者变量的风险。重复测量需要较少参与者并能控制个体差异，但必须通过平衡法来控制顺序效应。配对组提供了一种折中方案，但耗时较长。
When the question specifies a design, you must follow it. When free to choose, justify your decision by linking it to the nature of the IV and practical constraints. For instance, if studying a skill that improves with practice, an independent groups design avoids practice effects.
当题目指定了设计时，你必须遵循。当可以自由选择时，要通过联系IV的性质和实际限制来论证你的决定。例如，如果研究一项会随练习而提高的技能，独立组设计可以避免练习效应。
6. Sampling Methods and Representativeness | 抽样方法与代表性
In your design, you are expected to identify the target population and describe the sampling technique clearly. Opportunity sampling is quick but prone to researcher bias; volunteer (self-selected) sampling gathers motivated participants but may lack representativeness; random sampling is the gold standard but often impractical; stratified sampling ensures subgroups are proportionally represented.
在你的设计中，你要明确目标总体并清晰地描述抽样方法。机会抽样速度快但容易受到研究者偏差影响；自愿（自我选择）抽样能招募到有动力的参与者但可能缺乏代表性；随机抽样是黄金标准但常常不切实际；分层抽样确保各亚群按比例被代表。
Your choice must be realistic for the given scenario. A school-based study is likely to use opportunity sampling from sixth-form students, and you should acknowledge the limitation, e.g., ‘the sample may be ethnocentric and lack generalisability to older populations’.
你的选择必须符合所给情境的现实情况。一项在校研究很可能从六年级学生中进行机会抽样，你应当指出其局限性，例如“样本可能存在种族中心主义，无法推广到更年长的人群”。
7. Controlling Extraneous Variables to Protect Validity | 控制额外变量以保护效度
Internal validity refers to whether the IV truly caused the change in DV. To protect it, you must identify key extraneous variables (participant variables, situational variables, demand characteristics) and propose effective controls. Standardised instructions, random allocation, counterbalancing, and single-blind or double-blind procedures are your go-to tools.
内部效度指是否真正由自变量引起了因变量的变化。为保护内部效度，你必须识别关键的额外变量（参与者变量、情境变量、需求特征）并提出有效的控制措施。标准化指导语、随机分配、平衡法以及单盲或双盲程序是你的必备工具。
For example, if participants might guess the aim, use a double-blind technique where neither the participant nor the investigator interacting with them knows the condition. This minimises both demand characteristics and experimenter effects.
例如，如果参与者可能猜到研究目的，就使用双盲技术，使得参与者和与其交流的研究者都不知道分组情况。这能同时将需求特征和实验者效应降至最低。
8. Embedding Ethics Throughout Your Study | 将伦理贯穿于你的研究
Ethical considerations are not an afterthought. You must explicitly describe how you will obtain informed consent (or parental consent for under 16s), maintain confidentiality (using codes instead of names), offer the right to withdraw, protect participants from psychological and physical harm, and provide a full debrief at the end.
伦理考量并非事后补充。你必须明确描述如何获取知情同意（或对16岁以下者取得家长同意）、维护保密性（使用代码代替姓名）、提供退出权、保护参与者免受心理和生理伤害，并在结束时进行充分的事后解释。
When deception is unavoidable, explain why it is necessary and state that a debrief will reveal the true purpose of the study and offer the opportunity to withdraw data. Always connect your ethical safeguards to the BPS guidelines that CAIE advocates.
当欺骗不可避免时，解释为何有必要，并说明事后解释将揭示研究的真实目的，并提供撤回数据的机会。始终将你的伦理保障措施与CAIE推崇的BPS准则联系起来。
9. Selecting Descriptive and Inferential Statistics | 选择描述统计与推断统计
Your answer must include a planned analysis section. Start with descriptive statistics: measures of central tendency (mean, median, mode) and dispersion (range, standard deviation). A bar chart showing mean DV scores for each condition is often appropriate. Then, choose an inferential test matched to your design and data type.
你的答案必须包含计划分析部分。从描述统计开始：集中趋势量数（平均数、中位数、众数）和离散量数（全距、标准差）。一个展示各条件因变量平均数的柱状图通常是合适的。然后，选择一个与你的设计和数据类型相匹配的推断检验。
| Design & Data | Parametric Test | Non-parametric Test |
|---|---|---|
| Independent groups, interval data | Unrelated t-test | Mann-Whitney U |
| Repeated measures, interval data | Related t-test | Wilcoxon signed-rank |
| Correlation, at least ordinal data | Pearson’s r | Spearman’s rₛ |
| Nominal data, independent groups | – | Chi-squared (χ²) |
表格：常见研究设计对应的参数检验和非参数检验。记住 CAIE 不要求你计算公式，但要能给出检验名称和理由。
You must also state the chosen level of significance, typically p ≤ 0.05, and explain how the obtained value will be compared with a critical value to determine whether to reject H₀. Mention the degrees of freedom (df) where relevant.
你还必须说明所选的显著性水平，通常为 p ≤ 0.05，并解释如何将计算所得值与临界值比较，以决定是否拒绝 H₀。适当时，提及自由度 (df)。
10. Structuring a Research Report in Your Answer | 在你的答案中构建研究报告
Even when the question does not request a full report, structuring your design along standard sections reveals methodical thinking. Reference the essential components: abstract, introduction, method (design, participants, materials, procedure), results, discussion, and references. In the procedure, use a step-by-step numbered format for clarity.
即使题目不要求一份完整报告，按照标准部分来组织你的设计也能体现条理清晰的思维。参考基本框架：摘要、引言、方法（设计、参与者、材料、程序）、结果、讨论和参考文献。在程序部分，使用逐步编号的格式以求清晰。
For the results section in a design question, do not invent data. Instead, explain what descriptive statistics would be calculated, which inferential test would be applied, and what a significant result would mean. This approach satisfies the ‘analysis’ criterion.
对于设计题中的结果部分，不要捏造数据。相反，要解释将计算哪些描述统计量，将应用哪个推断检验，以及一个显著性结果意味着什么。这种方法能够满足“分析”的评分标准。
11. Common Pitfalls in Research Design Questions | 研究设计题常见错误
Many candidates lose marks due to vague operationalisation, such as ‘memory will be measured by a test’ without specifying the type of test and scoring. Another frequent error is mismatching the hypothesis to the design—for instance, using a directional hypothesis without justification from previous research. Failing to describe counterbalancing for repeated measures is a recurring omission.
许多考生因模糊的操作性定义而失分，例如“记忆将用一个测试来测量”，却没有说明测试类型和评分方式。另一个常见错误是假设与设计不匹配——例如，在没有先前研究支撑的情况下使用方向性假设。未能为重复测量描述平衡法则是一个反复出现的疏漏。
Additionally, candidates sometimes propose a t-test for nominal data or neglect to mention how ethical approval would be obtained. Ensure every element of your design is explicitly stated and logically connected to the aim.
此外，考生有时会对称名数据提议使用 t 检验，或者忘记提及如何获得伦理批准。确保你的设计中每个元素都明确陈述，并与研究目的有逻辑关联。
12. Exam Tactics for Maximum Marks | 获取最高分的考试策略
Read the scenario stem at least twice, underlining key terms such as the proposed IV, target group, and setting. Allocate roughly 25–30 minutes for Section A in Paper 4. Use the mark allocation as a guide—a 4-mark question on ethics expects four distinct, developed points. Use precise psychological terminology throughout; for example, state ‘counterbalancing’ rather than ‘mixing up the order’.
至少将情境题干阅读两遍，并在关键术语下划线，例如提议的IV、目标群体和背景。为试卷四的A部分留出大约25–30分钟。利用分值分配作为指引——一道4分的伦理题期望四个明确且展开的要点。全文使用精准的心理学专业术语；例如，表述为“平衡法”而不是“打乱顺序”。
Finally, always relate your decisions back to the aim of the investigation. This holism demonstrates that you are not merely reciting a template but thinking like a psychologist, which is exactly what the highest band descriptors demand.
最后，始终将你的决定与研究目的联系起来。这种整体性表明你不仅仅是在套用模板，而是在像心理学家一样思考，而这正是最高评分等级的描述符所要求的。
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