📚 CIE IGCSE Psychology: Key Points for Experimental & Practical Assessment | CIE IGCSE 心理学：实验与实践考核要点
In the CIE IGCSE Psychology course (Year 10), the experimental and practical assessment component tests your ability to design, conduct, analyse, and evaluate psychological investigations. Understanding the key concepts of scientific methodology — from formulating hypotheses to controlling variables and addressing ethical issues — is essential for success in the alternative-to-practical paper and for building a strong foundation in psychological research. This guide summarises the most important points you need to master, with clear English–Chinese explanations paired to reinforce your learning.
在 CIE IGCSE 心理学（Year 10）课程中，实验与实践考核部分考察设计、实施、分析和评价心理学研究的能力。从提出假设到控制变量，再到处理伦理问题，掌握科学方法论的关键概念对于在替代实践考试中取得成功、打好心理学研究基础至关重要。本指南总结了需要掌握的最重要考点，并采用英中对照讲解以强化学习。
1. Aims, Hypotheses and Variables | 目标、假设与变量
An aim is a general statement describing the purpose of the study. It is not directly testable but sets the research direction. For example: 'To investigate the effect of background noise on memory recall.'
目标是对研究目的的一般性陈述，不可直接验证，但指明研究方向。例如：“调查背景噪音对记忆回忆的影响。”
A hypothesis is a precise, testable prediction about the expected outcome. It must be operationalised, meaning the variables are clearly defined so they can be measured. A hypothesis can be directional (one‑tailed), predicting the specific direction of the effect (e.g., 'Participants exposed to loud noise will recall fewer words than those in silence'), or non‑directional (two‑tailed), predicting that a difference exists without stating its direction (e.g., 'There will be a difference in word recall between noisy and silent conditions'). The null hypothesis states that there will be no difference or relationship (e.g., 'There will be no difference in recall between the two noise conditions').
假设是对预期结果的精确且可验证的预测，必须被操作化，即明确定义变量以便测量。假设可以是方向性（单尾）的，预测效应的具体方向（例如，“暴露于嘈杂噪音的参与者比安静环境中的参与者回忆的单词更少”），也可以是非方向性（双尾）的，预测存在差异但不指明方向（例如，“噪音和安静条件下的单词回忆量存在差异”）。零假设则声明不存在差异或关系（例如，“两种噪音条件下的回忆量没有差异”）。
The independent variable (IV) is the factor that the researcher manipulates. The dependent variable (DV) is the factor that is measured. In the noise–memory study, the IV could be the presence or absence of loud noise, and the DV is the number of words correctly recalled. Extraneous variables (EVs) are any other variables that might affect the DV. If an EV is not controlled and changes systematically with the IV, it becomes a confounding variable, threatening the internal validity of the experiment. Common sources of EVs include participant variables (e.g., age, mood), situational variables (e.g., time of day, temperature), and experimenter effects.
自变量（IV）是研究者操纵的变量，因变量（DV）是被测量的变量。在噪音-记忆研究中，IV 可以是是否存在大声噪音，DV 是正确回忆的单词数量。额外变量（EV）是可能影响 DV 的任何其他变量。若一个 EV 未被控制且随 IV 系统性变化，它就会成为混淆变量，威胁实验的内部效度。常见的 EV 来源包括参与者变量（如年龄、情绪）、情景变量（如时间、温度）和实验者效应。
2. Types of Experiments: Laboratory, Field and Natural | 实验类型：实验室实验、现场实验与自然实验
A laboratory experiment is conducted in a controlled, artificial environment. The researcher manipulates the IV and strictly controls EVs. This design allows strong control over variables and high internal validity, but may lack ecological validity because the setting is unlike real life. Demand characteristics can also be a problem, as participants may guess the aim and change their behaviour.
实验室实验在受控的人工环境中进行。研究者操纵 IV 并严格控制 EV。这种设计能很好地控制变量、具有较高的内部效度，但由于环境不似真实生活，可能缺乏生态效度。需求特征也可能成为问题，因为参与者可能猜出研究目的并改变行为。
A field experiment takes place in a natural, everyday setting. The IV is still manipulated, but there is less control over EVs. Ecological validity is higher, and demand characteristics may be reduced because participants often do not know they are taking part. However, it is harder to replicate and there may be ethical issues, such as lack of informed consent.
现场实验在自然的日常环境中进行。IV 仍被操纵，但对 EV 的控制较少。生态效度更高，且由于参与者通常不知道自己在参与实验，需求特征可能减少。但实验较难重复，且可能出现伦理问题，如未获得知情同意。
A natural (or quasi) experiment occurs when the IV changes naturally — it is not directly manipulated by the researcher. The researcher takes advantage of a naturally occurring difference, such as comparing people before and after a natural disaster. This design is useful when manipulation would be unethical or impractical. It offers high ecological validity but very low control over EVs, and cause-and-effect conclusions are difficult to draw because participants are not randomly allocated to conditions.
自然实验（准实验）发生在 IV 自然变化时——研究者并未直接操纵它，而是利用自然发生的差异，例如比较自然灾害前后的人们。当直接操纵不道德或不切实际时，这种设计很有用。它具有很高的生态效度，但对 EV 的控制极低，且由于参与者未被随机分配，很难得出因果结论。
3. Experimental Designs: Independent Groups, Repeated Measures and Matched Pairs | 实验设计：独立组、重复测量与配对组
In an independent groups design, different participants are assigned to each condition of the IV. This avoids order effects such as practice or fatigue, but participant variables (individual differences) may become a confounding variable. Random allocation to conditions helps reduce this risk. This design requires more participants than repeated measures.
在独立组设计中，不同参与者被分配到 IV 的各个条件。这避免了练习效应或疲劳等顺序效应，但参与者变量（个体差异）可能成为混淆变量。随机分配有助于降低这一风险。该设计比重复测量需要更多参与者。
A repeated measures design uses the same participants in all conditions. It controls for participant variables and requires fewer participants, but it is vulnerable to order effects. Counterbalancing — where half the participants experience condition A first and the other half condition B first — can control for order effects. However, participants may guess the aim when they experience both conditions, increasing demand characteristics.
重复测量设计在所有条件下使用同一批参与者。它控制了参与者变量，需要的人数较少，但容易受到顺序效应的影响。平衡法——让一半参与者先体验条件 A，另一半先体验条件 B——可以控制顺序效应。然而，参与者经历两个条件后可能会猜到研究目的，从而增加需求特征。
Matched pairs design pairs participants on key variables (e.g., IQ, age) and then allocates one member of each pair to each condition. It reduces participant variables without the order effects of repeated measures, but matching is difficult and time‑consuming. If matching is not precise, participant effects may still confound the results.
配对组设计根据关键变量（如智商、年龄）将参与者配对，然后将每对中的一名分配到每种条件。它减少了参与者变量，且没有重复测量的顺序效应，但配对困难且耗时。如果配对不精确，参与者效应仍可能混淆结果。
4. Sampling Techniques | 抽样技术
A sample is the group of participants who take part in a study, drawn from a target population. The goal is to obtain a representative sample so findings can be generalised. Random sampling gives every member of the population an equal chance of being selected, which reduces bias, but it can still be unrepresentative by chance, especially with small samples.
样本是参加研究的参与者群体，从目标总体中抽取。目标是获得代表性样本，以便推广研究结果。随机抽样使总体中每个成员都有均等的被选中的机会，减少了偏差，但仍可能因偶然而缺乏代表性，特别是在小样本时。
Opportunity sampling involves selecting people who are easily available at the time, such as students in a classroom. It is quick and convenient, but likely to be biased because it draws from a narrow, accessible group. Volunteer (self‑selecting) sampling relies on people responding to an advert; these volunteers may be more motivated and thus not representative. Systematic sampling selects every nth person from a list, providing a structured approach that can improve representativeness if the list is random. Stratified sampling ensures subgroups (strata) within the population are proportionally represented, often producing the most representative sample.
机会抽样指选取当时容易接触到的人，如教室里的学生。这种方法快速方便，但可能因抽自狭窄的易接触群体而产生偏差。志愿者（自选）抽样依赖人们对广告的响应；这些志愿者可能更积极，因而缺乏代表性。系统抽样从名单中每隔 n 人选取一人，如果名单随机，这种结构化方法可以提高代表性。分层抽样确保总体中各亚群体（层）按比例代表，通常能产生最具代表性的样本。
5. Controlling Extraneous Variables | 控制额外变量
Standardisation means keeping the procedure identical for all participants except for the IV manipulation. This includes using the same instructions, environment, and timing. It reduces situational variables and experimenter bias.
标准化意味着除 IV 操作外，所有参与者的程序完全相同。这包括使用相同的指导语、环境和时间安排，从而减少情景变量和实验者偏差。
Randomisation involves using chance to allocate participants to conditions or to decide the order of stimuli. For example, flipping a coin helps avoid systematic bias and distributes participant variables evenly across groups. In a repeated measures design, counterbalancing is a form of randomisation that controls order effects by varying the order of conditions for different participants.
随机化指用随机方法将参与者分配到不同条件或决定呈现刺激的顺序。例如，掷硬币有助于避免系统性偏差，并使参与者变量均匀分布在各组。在重复测量设计中，平衡法是一种随机化形式，通过在不同参与者中改变条件的顺序来控制顺序效应。
Single‑blind and double‑blind procedures help control demand characteristics and experimenter effects. In a single‑blind study, participants do not know which condition they are in. In a double‑blind study, neither the participants nor the researcher interacting with them knows the condition, reducing both demand characteristics and experimenter expectation. A pilot study is a small‑scale trial run that helps identify and fix problems with the procedure before the main study, improving reliability and validity.
单盲和双盲程序有助于控制需求特征和实验者效应。在单盲研究中，参与者不知道自己在哪个条件下。在双盲研究中，参与者和直接接待他们的研究者都不知道分组情况，从而同时减少需求特征和实验者期望。预研究是小规模的试运行，有助于在正式研究前识别并解决程序问题，提高信度和效度。
6. Data Collection Methods | 数据收集方法
Psychologists use various methods to gather data, each suited to different research questions. Questionnaires are sets of written questions that can produce quantitative (closed questions) or qualitative (open questions) data. They can reach large samples quickly but may suffer from social desirability bias and low response rates.
心理学家使用多种方法收集数据，各适用于不同研究问题。问卷是一套书面问题，可产生量化（封闭式问题）或质化（开放式问题）数据。它们能快速覆盖大样本，但可能受社会赞许性偏差和低回收率的影响。
Interviews involve direct questioning, either structured (fixed questions), unstructured (free‑flowing), or semi‑structured. They can gather rich, detailed data, but interviewer effects and social desirability can distort responses. Observations allow researchers to watch and record behaviour. A naturalistic observation occurs in the participant’s own environment, while a controlled observation takes place in an arranged setting. Observations can be covert or overt, and participant or non‑participant. Behavioural categories must be clearly operationalised, and inter‑observer reliability should be checked.
访谈涉及直接提问，可分为结构化（固定问题）、非结构化（自由对话）或半结构化。访谈能收集丰富详细的数据，但访问者效应和社会赞许性可能扭曲答案。观察法让研究者观察并记录行为。自然观察发生在参与者自身环境中，而控制观察则在安排好的环境中进行。观察可以是隐蔽的或公开的，也可以是参与式或非参与式的。行为类别须清晰操作化，并应检验观察者间信度。
Case studies provide in‑depth analysis of an individual or small group, often using multiple methods. They give detailed insights but lack generalisability. Correlational research measures the relationship between two co‑variables without manipulation. A correlation coefficient (ranging from −1 to +1) indicates the direction and strength of the relationship, but correlation does not imply causation.
个案研究对个体或小群体进行深入分析，常结合多种方法。它能提供详尽见解，但缺乏可推广性。相关研究测量两个协同变量之间的关系，无需操纵。相关系数（范围从 −1 到 +1）显示关系的方向和强度，但相关并不意味着因果关系。
7. Ethical Considerations | 伦理考量
All psychological research must follow ethical guidelines to protect participants. In the UK, the British Psychological Society (BPS) sets these standards. Informed consent means participants must be told the aims, procedures, and their right to withdraw, and then voluntarily agree to take part. If deception is necessary, it should be minimised and justified, and followed by a thorough debriefing.
所有心理学研究必须遵循伦理准则以保护参与者。在英国，英国心理学会（BPS）制定了这些标准。知情同意意味着须告知参与者研究目的、程序及其退出权，并自愿同意参加。如果欺骗不可避免，应尽量减少并证明其合理性，并在事后进行充分的事后解释。
Participants must be protected from physical and psychological harm, and their privacy and confidentiality must be respected — data should be anonymised, and individuals not identifiable in reports. The right to withdraw must be made clear at the start, and participants can leave at any time without penalty. Debriefing after the study should restore participants to their original state, explain the true purpose, and offer support if distress has occurred. Special care is needed when working with vulnerable groups, such as children or those with mental health issues.
必须保护参与者免受身体和心理伤害，尊重其隐私和保密性——数据应匿名化，报告中不应识别个人。必须在开始时就明确说明退出权，参与者可随时退出而无需受罚。研究后的事后解释应让参与者恢复原状，说明真实目的，并在他们感到痛苦时提供支持。与弱势群体（如儿童或心理健康问题者）合作时需要特别谨慎。
8. Descriptive Statistics and Data Display | 描述统计与数据呈现
Descriptive statistics summarise data so they can be interpreted. Measures of central tendency include the mean (average), median (middle value), and mode (most frequent score). The mean is the most sensitive but is affected by outliers, whereas the median is unaffected by extreme scores. The mode is useful for nominal data.
描述统计概括数据以便解读。集中趋势的测量包括平均数（均值）、中位数（中间值）和众数（最常见分数）。平均数最为灵敏但受异常值影响，中位数则不受极端分数影响。众数适用于名义数据。
Measures of dispersion describe the spread of data. The range is the difference between the highest and lowest scores, simple but affected by outliers. The standard deviation (σ) shows how much scores deviate from the mean. A large standard deviation indicates widely spread scores, while a small one shows scores are clustered around the mean. It can be calculated using the formula:
离散程度测量描述数据的分散程度。全距是最高分与最低分之差，简单但受异常值影响。标准差（σ）显示分数与平均数的偏离程度。标准差大表示分数分布广，标准差小表示分数密集在平均数周围。可使用以下公式计算：
σ = √(∑(x − x̄)² / n)
Data are often presented in tables, bar charts, histograms, or scatterplots. Bar charts are used for discrete categories, while histograms present continuous data. Scatterplots show the relationship between two co‑variables, with each point representing a participant. When interpreting graphs, you must comment on the central tendency, dispersion, and any patterns or anomalies.
数据常以表格、条形图、直方图或散点图呈现。条形图用于离散类别，直方图展示连续性数据。散点图显示两个协同变量之间的关系，每个点代表一个参与者。解读图表时，必须评论集中趋势、离散程度及任何模式或异常。
9. Evaluation: Reliability, Validity and Generalisability | 评价：信度、效度与可推广性
Reliability refers to the consistency of a measure. Internal reliability checks whether items within a test are consistent (e.g., split‑half method). External reliability refers to consistency over time (test–retest) or across different observers (inter‑rater reliability). A study can be reliable but not valid — for instance, a set of weighing scales may consistently give the same wrong weight.
信度指测量的一致性。内部信度检验测试内项目是否一致（如分半法），外部信度则指跨时间的一致性（重测信度）或不同观察者间的一致性（观察者间信度）。一项研究可以有信度但无效度——例如，一台体重秤可能始终显示同一错误重量。
Validity is whether a test measures what it claims to measure. Internal validity asks whether the IV really caused the change in the DV, free from confounding variables. External validity includes ecological validity (can the findings apply to real‑life settings?), population validity (can they generalise to other groups?), and temporal validity (do the results hold over time?). Demand characteristics, social desirability, and experimenter bias all threaten internal validity.
效度是指测验是否测量了它声称要测量的东西。内部效度考察 DV 的变化是否确实由 IV 引起，而非混淆变量造成。外部效度包括生态效度（研究结果能否适用于真实生活情景？）、人群效度（能否推广到其他群体？）和时间效度（结果是否跨时间成立？）。需求特征、社会赞许性和实验者偏差都会威胁内部效度。
Generalisability refers to the extent to which findings can be applied to the wider population. It depends heavily on the sampling method and sample size. A large, representative sample using random or stratified techniques increases generalisability. Additionally, a study conducted in a natural setting (high ecological validity) is more likely to produce findings that generalise to everyday life.
可推广性指研究结果能应用到更广泛人群的程度。它高度依赖抽样方法和样本量。采用随机或分层技术的大样本、代表性样本可提高可推广性。此外，在自然环境（高生态效度）中进行的研究更易产生可推广到日常生活的结论。
10. Writing a Practical Report | 撰写实验报告
A psychological practical report follows a standard format. The abstract is a short summary of the aim, method, results, and conclusion — it should be written last. The introduction reviews relevant background theory and studies, and clearly states the aim and hypothesis (including null and alternative). The method section is divided into design (type of experiment, design, variables, controls), participants (sampling method and demographic details), apparatus/materials (list of equipment), and procedure (step‑by‑step instructions so the study can be replicated).
心理学实验报告遵循标准格式。摘要是对目标、方法、结果和结论的简短概括——应最后撰写。导言回顾相关的背景理论和研究，并清晰陈述目标和假设（含零假设和备择假设）。方法部分分为设计（实验类型、设计、变量、控制）、参与者（抽样方法和人口细节）、仪器/材料（设备清单）和程序（分步骤说明，便于复制）。
The results section presents the findings, usually with descriptive statistics (mean, median, standard deviation) and graphs. No interpretation is given here. The discussion interprets the results in relation to the hypothesis and background research, explains whether the findings support the hypothesis, considers limitations (e.g., EV control, sample bias), and suggests improvements and real‑world applications. Finally, references list all sources cited in the report, and appendices may include raw data, consent forms, or instructions.
结果部分呈现研究发现，通常包括描述统计（平均数、中位数、标准差）和图表。此处不做解读。讨论部分结合假设和背景研究解读结果，解释研究发现是否支持假设，思考局限性（如 EV 控制、样本偏差），并提出改进建议和现实应用。最后，参考文献列出报告中引用的所有来源，附录可以包含原始数据、知情同意书或指导语等。
When evaluating your own practical work, always link your comments to specific methodological concepts you have learned. For example, instead of saying 'the study was flawed', say 'the study used an opportunity sample, which limits population validity because the participants were only Year 10 students from one school'. This demonstrates deeper understanding and is essential for earning high marks in the CIE IGCSE Psychology assessment.
在评价自己的实践工作时，一定要将评述与你所学的方法论概念具体联系起来。例如，不要说“研究有缺陷”，而要说“研究采用了机会抽样，这限制了人群效度，因为参与者仅为一所学校的高一学生”。这展示了更深的理解，也是在 CIE IGCSE 心理学考核中取得高分的关键。
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