📚 AS CAIE Psychology: Experimental Practical Assessment Key Points | AS CAIE 心理学：实验/实践考核要点
Mastering the experimental method is fundamental for success in AS CAIE Psychology, especially in Paper 2 Research Methods. This guide distils the core assessment points for experimental design, control, ethics, data analysis, and evaluation. By understanding these practical essentials, you will be well-equipped to tackle exam questions that require you to design, critique, or interpret psychological experiments.
掌握实验方法是 AS CAIE 心理学（尤其是试卷二“研究方法”）成功的基础。本指南提炼了实验设计、控制、伦理、数据分析与评价等核心评估要点。通过理解这些实践精要，你将能从容应对要求设计、评论或解释心理实验的考题。
1. Key Experimental Variables | 关键实验变量
In any experiment, the independent variable (IV) is the condition manipulated by the researcher, while the dependent variable (DV) is the outcome measured to assess the effect of the IV. Extraneous variables (EVs) are any other factors that might influence the DV, and if not controlled they can become confounding variables that systematically distort the findings.
在任何实验中，自变量（IV）是研究者操控的条件，因变量（DV）是为评估 IV 效应而测量的结果。额外变量（EVs）是可能影响 DV 的任何其他因素，若不加以控制，它们会变成混淆变量，系统性地扭曲研究结果。
Operationalisation is the process of defining variables precisely so they can be manipulated or measured in a clear, replicable manner. For instance, 'memory performance' might be operationalised as the number of words correctly recalled from a list, while 'anxiety' could be operationalised as heart rate measured in beats per minute.
操作化是精确定义变量，以便以清晰、可复制的方式操控或测量它们的过程。例如，“记忆表现”可操作化为正确回忆单词的数量；“焦虑”可操作化为以每分钟心跳次数衡量的心率。
2. Types of Experimental Designs | 实验设计类型
Experimental design refers to how participants are allocated to the conditions of the IV. The three main types are independent groups, repeated measures, and matched pairs. Each has strengths and limitations that affect the validity and practicality of the experiment.
实验设计指的是参与者如何被分配到 IV 的各条件。主要有三种：独立组设计、重复测量设计和配对设计。每种设计都有影响实验效度与可行性的优点和局限。
| Design | How It Works | Key Advantage | Key Limitation |
|---|---|---|---|
| Independent groups 独立组设计 | Different participants in each condition 每种条件使用不同参与者 | No order effects (e.g., practice or fatigue) 无顺序效应（如练习或疲劳） | Participant variables may differ between groups 组间参与者变量可能不同 |
| Repeated measures 重复测量设计 | The same participants take part in all conditions 同一批参与者接受所有条件 | Participant variables are controlled because the same people are used 因使用相同人群，参与者变量得到控制 | Order effects are likely; counterbalancing is needed 可能出现顺序效应，需要平衡处理 |
| Matched pairs 配对设计 | Participants are matched on key traits and then split into conditions 参与者在关键特质上被配对，然后分入不同条件 | Reduces participant variables without order effects 减少参与者变量，无顺序效应 | Time-consuming and difficult to match perfectly 耗时且难以完美配对 |
Choosing the right design depends on the nature of the task and the resources available. In the exam, you may be asked to justify a particular design or suggest how to overcome its weaknesses, for example by using random allocation in independent groups or counterbalancing in repeated measures.
选择正确设计取决于任务性质与可用资源。考试中可能要求你说明为何选用某种设计，或提出克服其弱点的方法，例如在独立组中使用随机分配，或在重复测量中实施平衡法。
3. Controlling Extraneous Variables | 控制额外变量
To ensure the IV alone is responsible for changes in the DV, researchers must minimise extraneous variables. Standardisation of procedure means all participants experience the same environment, instructions, and materials. This can involve using a scripted briefing and debriefing, controlling room temperature, or ensuring timing is identical across conditions.
为确保只有 IV 引起 DV 的变化，研究者必须将额外变量降至最低。程序标准化意味着所有参与者经历相同的环境、指导语和材料。这包括使用脚本化的任务说明与任务后解释、控制室温或确保各条件计时一致。
Random allocation is a key technique in independent groups design: each participant has an equal chance of being placed in any condition, which distributes participant variables evenly. In repeated measures, counterbalancing (e.g., ABBA design) is used to control order effects, where half the participants experience condition A first and half experience condition B first.
随机分配是独立组设计的关键技术：每位参与者有同等机会被分入任一条件，从而均匀分布参与者变量。在重复测量中，采用平衡法（如 ABBA 设计）控制顺序效应，即一半参与者先接受条件 A，另一半先接受条件 B。
Single-blind and double-blind procedures can reduce demand characteristics and experimenter bias. In a single-blind study, participants do not know which condition they are in; in a double-blind study, neither participants nor the researcher interacting with them knows the condition allocation until after data collection.
单盲与双盲程序可降低需求特征和实验者偏见。单盲研究中，参与者不知自己处于何种条件；双盲研究中，参与者和与之互动的研究者均不知道条件分配，直至数据收集结束。
4. Sampling Techniques | 抽样技术
The sample is the group of people who take part in an experiment; they should represent the target population so that findings can be generalised. Opportunity sampling involves recruiting people who are readily available, such as students in a common room. It is quick and convenient, but the sample may be unrepresentative and biased.
样本是参与实验的一组人；他们应能代表目标人群，以便研究结果可推广。机会抽样是指招募方便找到的人，如休息室里的学生。此法迅速便捷，但样本可能不具代表性且有偏差。
Random sampling gives every member of the target population an equal chance of being selected, for example by drawing names from a hat or using a random number generator. This helps reduce bias, although it is often difficult to obtain a truly random sample in practice. Volunteer sampling relies on individuals responding to advertisements; it can attract motivated participants but often suffers from volunteer bias, where the sample differs systematically from the general population.
随机抽样使目标人群中每位成员都有均等被选中的机会，例如从帽子中抽签或使用随机数生成器。这有助于减少偏差，尽管在实践中往往难以获得真正随机样本。志愿者抽样依赖个人回应广告；它能吸引动机较强的参与者，但常带有志愿者偏差，即样本系统性地有别于总体。
Stratified sampling ensures subgroups within the population (such as age bands or socioeconomic classes) are proportionately represented. This enhances representativeness but requires detailed population knowledge and can be time-consuming. In the exam, you may be asked to recommend a sampling method and explain its impact on generalisability.
分层抽样确保人群中的子群体（如年龄段或社会经济阶层）按比例得到代表。这能提升代表性，但需要详尽的人口信息且耗时。考试中可能要求你推荐一种抽样方法并解释其对推广性的影响。
5. Ethical Guidelines in Experiments | 实验中的伦理准则
Psychological experiments must adhere to strict ethical standards to protect participants' wellbeing. Informed consent means participants should be given full information about the nature of the study (where possible) and must freely agree to take part. When minor deception is necessary, researchers must obtain presumptive consent or prior general consent and debrief participants afterward.
心理实验必须遵守严格的伦理标准以保护参与者的福祉。知情同意意味着参与者应（尽可能）获得研究性质的完整信息并自愿同意参与。当有必要进行轻微欺骗时，研究者必须获得推定同意或事先一般同意，并在事后进行解释。
Participants must be told they have the right to withdraw at any stage without penalty. Their data should be destroyed if they request it. Protection from harm is paramount: both physical and psychological harm must be avoided. If distress is possible, researchers must monitor participants and offer support or withdrawal immediately.
参与者必须被告知其有权在任何阶段退出且不受惩罚。如他们要求，数据应被销毁。保护免受伤害至关重要：必须避免生理和心理伤害。若可能出现痛苦，研究者须监测参与者并立即提供支持或退出选项。
Confidentiality means that participants' personal details are kept anonymous, often by using codes instead of names. Debriefing after the experiment informs participants of the true aims and addresses any misconceptions. These ethical points are frequently examined through scenario-based questions demanding practical solutions.
保密性指参与者个人信息保持匿名，常用代码代替姓名。实验后的任务后解释告知参与者真实目的并澄清任何误解。这些伦理要点常通过情景题考核，要求提出可行方案。
6. Procedure and Materials | 程序与材料
A well-designed experiment requires a detailed, standardised procedure that can be replicated. The materials might include stimulus lists, questionnaires, computer tasks, or physiological measurement devices. Everything should be prepared in advance to minimise variation. The procedure should include a clear consent form, standardised instructions read verbatim to each participant, and a debriefing sheet.
一个精心设计的实验需要详细、标准化的可复制程序。材料可能包括刺激清单、问卷、计算机任务或生理测量设备，均须提前准备以减少差异。程序应包括清晰的同意书、向每位参与者逐字宣读的标准化指导语以及任务后解释说明书。
Pilot studies are small-scale trial runs conducted before the main experiment. They help identify ambiguous instructions, check equipment, and ensure the DV is sensitive enough. Pilot studies improve the internal validity of the final study and are a key mark-scoring point in design questions.
试点研究是在主实验前进行的小规模试运行。它们帮助识别含糊的指导语、检查设备并确保 DV 足够灵敏。试点研究提升最终研究的内部效度，是设计题中重要的得分点。
7. Data Collection Methods within Experiments | 实验中的数据收集方法
Experiments often employ multiple data collection techniques. Behavioural measures include recording accuracy, reaction time, or frequency of a specific action. These are objective and easy to quantify. Self-report measures such as questionnaires or rating scales can capture subjective experiences like mood or confidence levels, but they are prone to social desirability bias.
实验常采用多种数据收集技术。行为测量包括记录正确率、反应时或特定动作的频率。这些客观且易于量化。自我报告测量如问卷或评定量表可以捕捉主观体验，如情绪或信心水平，但易受社会期许偏差影响。
Physiological measures, such as heart rate, skin conductance, or cortisol levels, provide objective data about bodily arousal linked to psychological states. Combining several measures (triangulation) increases the richness and validity of the findings, though it may require more complex data handling.
生理测量如心率、皮肤电导或皮质醇水平，提供与心理状态相关的客观生理唤起数据。结合多种测量（三角互证）可增加研究结果的丰富性与效度，尽管这可能需要更复杂的数据处理。
8. Descriptive and Inferential Statistics | 描述性与推断性统计
Descriptive statistics summarise the data collected. Measures of central tendency (mean, median, mode) indicate a typical value; measures of dispersion (range, standard deviation) show how spread out the data are. For example, a low standard deviation suggests scores are clustered closely around the mean.
描述性统计总结所收集的数据。集中趋势量数（平均数、中位数、众数）指示典型值；离散程度量数（全距、标准差）显示数据分布广度。例如，低标准差表明分数紧密围绕平均数。
Data can be displayed in bar charts, histograms, or scattergrams. Inferential statistics allow researchers to determine whether the results are likely due to the manipulation of the IV rather than chance. The p-value indicates the probability that the observed effect occurred if the null hypothesis were true. Typically, if p < 0.05, the result is considered statistically significant. You do not need to calculate tests, but you must interpret simple findings and understand the logic of significance.
数据可用条形图、直方图或散点图展示。推断性统计让研究者能判断结果是否很可能是由 IV 的操纵而非偶发因素所致。p 值指示若零假设为真，观察到该效应的概率。通常，若 p < 0.05，则结果被认为具有统计显著性。你不必计算检验，但必须解释简单的发现并理解显著性的逻辑。
9. Reliability and Validity in Experimental Research | 实验研究中的信度与效度
Reliability refers to consistency. Internal reliability concerns whether the items within a test or the procedure are consistent, for instance all trials measuring the same thing. External reliability addresses whether the experiment produces similar results if repeated under the same conditions (test-retest). Standardised procedures and careful training of researchers improve reliability.
信度指一致性。内部信度关注测试内部项目或程序是否一致，例如所有试验是否测量同一事物。外部信度关注若在相同条件下重复实验，是否产生相似结果（重测信度）。标准化程序和严格的研究者训练可提升信度。
Validity is about whether the experiment genuinely measures what it claims to measure. Internal validity asks if the IV truly caused the change in the DV, free from confounding variables. External validity concerns generalisability beyond the experimental setting: ecological validity refers to how well findings apply to real-life situations, while population validity refers to the extent results can be generalised to other groups. High internal validity sometimes comes at the expense of ecological validity in tightly controlled lab experiments.
效度关乎实验是否真正测量了其声称测量的内容。内部效度探问 IV 是否真正引起了 DV 的变化，不受混淆变量影响。外部效度关乎超越实验背景的推广性：生态效度指研究结果多大程度适用于真实情境；人群效度指结果可推广至其他群体的程度。在严格控制的实验室实验中，高内部效度有时以牺牲生态效度为代价。
10. Evaluating and Improving Experiments | 评价与改进实验
When evaluating an experiment, start by identifying the type of experimental design and how well it addressed potential confounds. Consider the sample: was it large enough and representative? Discuss the ethical safeguards and whether any ethical issues might have affected participants' behaviour or well-being.
评价实验时，先识别实验设计类型及其处理潜在混淆的程度。考量样本：是否足够大且有代表性？讨论伦理保障措施，以及是否存在可能影响参与者行为或福祉的伦理问题。
Assess the reliability and validity of the study. Could demand characteristics or researcher bias have influenced results? Suggest practical improvements, such as using a double-blind procedure, random assignment, a matched-pairs design, or increasing the sample size. Also consider whether the task is realistic and whether the DV reflects real-world behaviour. Offering specific, justified improvements is a skill regularly tested in AS exam questions.
评估研究的信度与效度。需求特征或研究者偏见是否可能影响了结果？提出实际改进，例如采用双盲程序、随机分配、配对设计或扩大样本量。同时考虑任务是否真实，DV 是否反映现实行为。提出具体、有据的改进措施是 AS 考试中常常考查的技能。
The ability to critically evaluate methodology and propose well-reasoned modifications demonstrates thorough understanding of experimental practical assessment. Use concrete details from the study description and link every suggestion to a clear methodological benefit.
批判性评价方法并提出论证充分的修改建议，能展示你对实验实践考核的透彻理解。运用研究描述中的具体细节，并将每条建议与明确的方法论益处挂钩。
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