A-Level CAIE Psychology: Key Points for Experimental/Practical Assessments | A-Level CAIE心理学：实验与实践考核要点

📚 A-Level CAIE Psychology: Key Points for Experimental/Practical Assessments | A-Level CAIE心理学：实验与实践考核要点

In Cambridge International A-Level Psychology (9990), practical and experimental skills are not just a small part of the syllabus; they form the core of Papers 2 and 4. Paper 2 (Research Methods) tests your ability to design, evaluate, and critique psychological investigations, while Paper 4 (Personal Investigation) requires you to plan, carry out, and write up your own original study. Mastering these practical assessment points is essential for achieving top marks. This article breaks down the key areas you must understand, from formulating a hypothesis to analysing data and discussing ethical implications. By internalising these principles, you will be able to approach any research method scenario with confidence and precision.

在剑桥国际A-Level心理学（9990）课程中，实验与实践技能并不仅仅是课程的一小部分，而是Paper 2和Paper 4的核心。Paper 2（研究方法）考查你设计、评估和批判心理学研究的能力，而Paper 4（个人研究）则要求你规划、实施并撰写自己的原创研究报告。掌握这些实践考核要点是取得高分的关键。本文将逐一拆解你必须理解的核心领域，从提出假设到分析数据，再到讨论伦理意义。内化这些原则后，你将能够自信而精准地应对任何研究方法的情境题。

1. Understanding Research Aims and Hypotheses | 理解研究目的与假设

Every investigation begins with a clear research aim, which is a general statement of what the study intends to investigate. From this aim, you derive a hypothesis — a precise, testable prediction about the relationship between variables. In CAIE Psychology, you must be able to differentiate between one-tailed (directional) and two-tailed (non-directional) hypotheses, and state them in operationalised terms. For example, instead of saying ‘memory will be affected by noise’, you would state: ‘Participants who learn a word list in silence will recall significantly more words than those who learn in 80 dB white noise.’ This operationalisation ensures that the variables are measurable and the study can be replicated.

每一项研究都始于一个明确的研究目的，这是对研究意图的一般性陈述。基于这个目的，你会提出一个假设——关于变量之间关系的精确、可检验的预测。在CAIE心理学中，你必须能够区分单尾（方向性）和双尾（非方向性）假设，并用操作化的方式表述。例如，不要说“噪音会影响记忆”，而应陈述：“在安静环境中学习单词表的参与者，其回忆单词的数量将显著多于在80分贝白噪音下学习的参与者。”这种操作化确保了变量可测量且研究可重复。

2. Variables: Independent, Dependent and Control | 变量：自变量、因变量与控制变量

The independent variable (IV) is the factor you manipulate or compare, and the dependent variable (DV) is the factor you measure. A common pitfall is failing to operationalise the DV clearly — ‘aggression’ could be measured by the number of times a child hits a Bobo doll, or by a teacher’s rating scale. You must specify exactly how the DV will be quantified. Equally critical are control variables; these are all the other potential influences that must be held constant to prevent confounding results. If you are testing the effect of caffeine on reaction time, you must control for participants’ prior caffeine intake, sleep quality, and age. Uncontrolled variables lower internal validity and make it impossible to draw a firm conclusion about cause and effect.

自变量（IV）是你操纵或比较的因素，因变量（DV）是你测量的因素。一个常见的错误是未能清楚地操作化因变量——“攻击性”可以通过儿童击打波波娃娃的次数来测量，也可以通过教师的评分量表来衡量。你必须确切说明如何量化因变量。同样至关重要的是控制变量；所有其他可能的影响因素都必须保持恒定，以防止混淆结果。如果你在测试咖啡因对反应时间的影响，就必须控制参与者之前的咖啡因摄入量、睡眠质量和年龄。未控制的变量会降低内部效度，使得关于因果关系的结论无法确立。

3. Experimental Designs: Independent Measures, Repeated Measures and Matched Pairs | 实验设计：独立组设计、重复测量设计与配对组设计

Choosing the right experimental design is a fundamental decision that affects how you allocate participants to conditions and how you handle individual differences. Use the following table to compare the three main designs tested in CAIE:

选择合适的实验设计是一项基本决策，它影响着你如何将参与者分配到不同条件以及如何处理个体差异。利用下表比较CAIE考查的三种主要设计：

Design	Description	Strengths	Limitations
Independent Measures	Different participants in each condition	No order effects; less demand characteristics	Participant variables may differ; needs more participants
Repeated Measures	Same participants in all conditions	Controls participant variables; fewer participants needed	Order effects (practice, fatigue); demand characteristics
Matched Pairs	Different but similar participants matched on key traits	Reduces participant variables; no order effects	Difficult to match perfectly; time-consuming

中文对照：独立组设计（不同参与者在各条件）、重复测量设计（相同参与者经历所有条件）、配对组设计（不同但按关键特质配对的参与者）。独立组无顺序效应但受参与者变量影响；重复测量控制参与者变量但有练习/疲劳效应；配对组减少参与者变量但难以完美匹配。

In a CAIE exam, you may be asked to suggest an appropriate design for a given scenario and justify your choice. Always consider how to deal with the limitations — for repeated measures, randomisation or counterbalancing can reduce order effects; for independent measures, random allocation to conditions helps distribute participant variables evenly.

在CAIE考试中，你可能会被要求为给定情境建议一种合适的设计并说明理由。始终考虑如何应对局限性——对于重复测量，随机化或平衡设计可以减少顺序效应；对于独立组，将参与者随机分配到不同条件有助于均匀分布参与者变量。

4. Sampling Methods and Their Implications | 抽样方法及其影响

The way you recruit participants directly impacts the generalisability of your findings. The four main sampling methods you need to know are: opportunity sampling (using whoever is available), volunteer/self-selected sampling (participants choose to take part), random sampling (every member of the target population has an equal chance of being selected), and stratified sampling (recruiting proportionally from subgroups). Opportunity sampling is quick and cheap but often produces a biased, unrepresentative sample. Random sampling theoretically gives the most representative sample but is rarely practical in school-based investigations. When writing up your own study, always discuss the sampling method used, explain its limitations, and suggest how a more representative sample could be obtained in future research.

你招募参与者的方式直接影响研究结果的推广性。你需要了解的四种主要抽样方法为：便利抽样（使用可找到的任何人）、志愿者/自选抽样（参与者主动参与）、随机抽样（目标人群中的每个成员都有相等机会被选出）和分层抽样（按比例从各亚群体中招募）。便利抽样快捷廉价，但往往会产生一个有偏差、不具代表性的样本。随机抽样理论上能提供最具代表性的样本，但在校内研究中很少可行。在撰写你自己的研究时，始终要讨论所使用的抽样方法，解释其局限性，并建议未来研究如何获得更有代表性的样本。

5. Controlling Extraneous Variables: Standardisation and Randomisation | 控制外源变量：标准化与随机化

High internal validity requires rigorous control over extraneous variables. Standardisation means keeping the procedure exactly the same for every participant — the same instructions (read verbatim), the same environment, the same materials. This reduces experimenter bias and situational variability. Randomisation is the technique of leaving some choices to chance, for instance, randomly assigning participants to conditions or randomising the order of word lists. This prevents systematic bias from creeping into the allocation process. In the Personal Investigation, you must document exactly how you standardised your procedure and used randomisation where possible, as examiners look for evidence of such controls.

高内部效度要求严格地控制外源变量。标准化意味着对每位参与者保持完全相同的程序——相同的指导语（逐字朗读）、相同的环境、相同的材料。这降低了实验者偏差和情境变异性。随机化是一种让某些选择由机会决定的技术，例如，将参与者随机分配到不同条件或随机化单词表顺序。这能防止系统偏差悄悄进入分配过程。在个人研究中，你必须准确记录你如何标准化你的程序并在可能的地方使用了随机化，因为考官会寻找这些控制的证据。

6. Ethics in Psychological Research | 心理学研究中的伦理

Ethical considerations are not an afterthought; they are integral to planning any study. The British Psychological Society (BPS) guidelines require you to obtain informed consent from participants (or parental consent for those under 16), offer the right to withdraw at any time without penalty, protect participants from physical and psychological harm, debrief them fully after the study, and maintain confidentiality of personal data. In CAIE Paper 2, you might be asked to identify ethical issues in a proposed study and suggest solutions. For example, if a study involves mild deception, you must justify why it is necessary and explain how you will debrief participants, giving them the opportunity to withhold their data.

伦理考量并非事后补充，而是规划任何研究的组成部分。英国心理学会（BPS）指南要求你获取参与者的知情同意（16岁以下需家长同意），给予参与者在任何时候无代价退出的权利，保护参与者免受身心伤害，在实验后对他们进行充分的知情澄清，并维护个人数据的机密性。在CAIE Paper 2中，你可能会被要求指出拟议研究中的伦理问题并提出解决方案。例如，如果研究涉及轻度欺骗，你必须论证其必要性，并说明你将如何向参与者澄清，给予他们撤回数据的机会。

7. Data Collection: Types of Data and Tools | 数据收集：数据类型与工具

Psychological investigations generate either quantitative data (numerical, e.g. scores on a memory test) or qualitative data (non-numerical, e.g. transcripts of interviews), or both. Quantitative data are easier to analyse statistically and present in graphs, but they may lack depth. Qualitative data provide rich detail but can be difficult to summarise and are more prone to researcher bias. Common tools include questionnaires with Likert scales, structured observation checklists, and standardised tests. When planning your own investigation, consider mixing methods — for example, record the number of words recalled (quantitative) and ask a follow-up interview question about strategies used (qualitative). This methodological pluralism can enrich your discussion in Paper 4.

心理学研究产生的数据要么是定量数据（数字，如记忆测试的得分），要么是定性数据（非数字，如访谈转录稿），或者两者兼有。定量数据更易于统计分析并用图表呈现，但可能缺乏深度。定性数据提供了丰富的细节，但难以归纳且更易受研究者偏差影响。常用工具包括带有李克特量表的问卷、结构化观察检查表和标准化测试。在规划你自己的研究时，考虑混合方法——例如，记录回忆的单词数量（定量）并追问一个关于所用策略的后续访谈问题（定性）。这种方法多元主义能丰富你在Paper 4中的讨论。

8. Descriptive Statistics: Measures of Central Tendency and Spread | 描述统计：集中量数与离中量数

You are expected to calculate and interpret the mean, median, mode, range, and standard deviation where appropriate. The mean is the arithmetic average, sensitive to extreme scores; the median is the middle value, less affected by outliers; the mode is the most frequent score. The range gives a crude measure of spread (highest minus lowest), while the standard deviation quantifies how much scores deviate from the mean. In a CAIE exam, you may be asked to justify why the median is more suitable than the mean for a skewed distribution, or to explain what a larger standard deviation tells you about the data. Always include units and round appropriately.

你需要学会计算并在适当的情境下解释平均数、中位数、众数、全距和标准差。平均数是算术均值，易受极端值影响；中位数是中间值，较少受异常值影响；众数是最频繁出现的分数。全距提供了粗略的离散程度（最高值减最低值），而标准差量化了分数偏离平均数的程度。在CAIE考试中，你可能会被要求论证为什么对于偏态分布中位数比平均数更合适，或者解释较大的标准差说明了数据的什么特点。记着要写单位并适当取整。

9. Graphical Representation and Data Interpretation | 图形呈现与数据解释

Choosing the right graph is a skill that earns marks in both Paper 2 and Paper 4. Bar charts are used for discrete categories (including mean scores of different conditions), histograms for continuous frequency data, and scatter graphs to display correlations. Every graph must have a clear title, fully labelled axes (with units), and an even scale. When interpreting a graph, do not simply describe the shape — relate it back to the hypothesis. For instance, ‘The bar chart shows that the mean recall in the silent condition (12.4 words) is visibly higher than in the noise condition (7.8 words), supporting the directional hypothesis.’ Avoid making cause-and-effect claims from correlational data.

选择合适的图表是一项在Paper 2和Paper 4中都能得分的技能。条形图用于离散类别（包括不同条件的平均数），直方图用于连续频率数据，散点图用于展示相关性。每张图表都必须有清晰的标题、完全标注的坐标轴（带单位）和均匀的刻度。解释图表时，不要仅仅描述形状——要将其与假设联系起来。例如，“条形图显示安静条件下的平均回忆数（12.4个词）明显高于噪音条件（7.8个词），支持了方向性假设。”切勿从相关数据中做出因果断言。

10. Inferential Statistics: Choosing and Justifying a Test | 推断统计：选择并论证检验方法

At A-Level, you need to select from a range of inferential tests: Chi-squared (nominal data, independent measures, test of association/difference), Mann-Whitney U (ordinal data, independent measures), Wilcoxon Signed-Ranks (ordinal data, repeated measures), related t-test (interval/ratio data, repeated measures, normally distributed), unrelated t-test (interval/ratio, independent measures), and Spearman’s rank correlation (ordinal, correlation). The choice hinges on three questions: (1) Am I looking for a difference or a correlation? (2) What level of measurement is my data? (3) Which experimental design was used? Once chosen, you must calculate the observed value, compare it to a critical value from a statistical table (using the appropriate degrees of freedom or N, and a significance level typically p ≤ 0.05), and decide whether to accept or reject the null hypothesis. Remember, if the observed value is more extreme than the critical value (for most tests, equal to or greater than, but check the specific test rule), the result is significant.

在A-Level阶段，你需要从一系列推断检验中选择：卡方检验（称名数据、独立测量、检验关联/差异）、曼-惠特尼U检验（顺序数据、独立组）、威尔科克森符号秩检验（顺序数据、重复测量）、相关t检验（等距/等比数据、重复测量、正态分布）、不相关t检验（等距/等比、独立组）和斯皮尔曼等级相关（顺序数据、相关）。选择取决于三个问题：（1）我在寻找差异还是相关？（2）我的数据属于哪种测量层次？（3）使用的是哪种实验设计？选定后，你必须计算观测值，将其与统计表中的临界值进行比较（使用适当的自由度或N，显著性水平通常 p ≤ 0.05），然后决定接受或拒绝零假设。记住，如果观测值比临界值更极端（对于大多数检验，等于或大于临界值即为显著，但要检查具体检验的规则），结果就是显著的。

11. Writing the Personal Investigation (Paper 4) | 撰写个人研究（Paper 4）

The Personal Investigation is a coursework component where you conduct a small-scale, original research study. The report must follow a formal structure: Abstract, Introduction (with background literature and a justified hypothesis), Method (subsections: Design, Sample, Apparatus/Materials, Procedure), Results (descriptive and inferential statistics, plus a graph), Discussion (linking findings to literature, evaluating methodology and ethics), and References. The CAIE mark scheme rewards precise operationalisation, thorough control of variables, appropriate statistical analysis, and a reflective evaluation that identifies improvements, such as using a larger, more diverse sample or a double-blind procedure to reduce demand characteristics. Pay special attention to the ethical checklist and ensure all raw data and consent forms are submitted as an appendix.

个人研究是一项课程作业，要求你进行一项小规模的原创研究。报告必须遵循正式结构：摘要、引言（包含背景文献和经过论证的假设）、方法（分小节：设计、样本、仪器/材料、程序）、结果（描述和推断统计，外加图表）、讨论（将发现与文献相联系，评价方法和伦理）以及参考文献。CAIE评分方案奖励精确的操作化、对变量的全面控制、恰当的统计分析以及反思性评价，该评价应指出改进措施，例如使用更大更多样的样本或采用双盲程序以减少需求特征。特别留意伦理检查清单，并确保所有原始数据和同意书作为附录提交。

12. Common Mistakes and How to Avoid Them | 常见错误与如何避免

Even well-prepared candidates lose marks by confusing directional and non-directional hypotheses, forgetting to operationalise the DV, overlooking order effects in repeated measures designs, or misapplying the criteria for a robust inferential test. Another frequent error is including unjustified causal language in the discussion of correlational studies. To avoid these pitfalls, create a checklist for every exam question: Is my hypothesis operationalised? Have I stated the null hypothesis? Did I identify the design and mention a relevant control? Is the graph correctly labelled? Am I using ‘significantly’ only if p ≤ 0.05? Practising past papers under timed conditions and marking them yourself with the official mark scheme is one of the most effective ways to internalise these requirements.

即使是准备充分的考生也会因混淆方向性与非方向性假设、忘记操作化因变量、忽视重复测量设计中的顺序效应，或误用稳健推断检验的准则而失分。另一个常见错误是在讨论相关研究时使用不恰当的因果推论语言。为避免这些陷阱，你可以为每一道试题制作一个清单：我的假设操作化了吗？我陈述了零假设吗？我是否识别了设计并提到了相关的控制？图表标注正确吗？我是否仅在 p ≤ 0.05 时才使用“显著”一词？在限时条件下练习历年真题并用官方评分方案自行批改，是内化这些要求最有效的方式之一。

Published by TutorHao | Psychology Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）