📚 Experimental and Practical Assessment Essentials for Year 13 Cambridge Psychology | 剑桥心理学13年级实验/实践考核要点
The experimental and practical component of Year 13 Cambridge Psychology is critical for success in Papers 2 and 4, which together assess research methods and data analysis skills. This guide outlines essential knowledge, from formulating hypotheses to conducting inferential statistics, and highlights common mistakes to avoid.
剑桥13年级心理学的实验和实践部分是卷二和卷四成功的关键，这两张试卷共同考察研究方法和数据分析技能。本指南概述了从提出假设到进行推论统计的重要知识，并强调了要避免的常见错误。
1. Crafting Research Questions and Hypotheses | 提出研究问题与假设
A research question must be clear, focused, and testable. It guides the entire study. From this, you formulate an alternative hypothesis (H₁), which predicts a significant effect or relationship, and a null hypothesis (H₀), which states there will be no effect or relationship. For directional predictions, use a one‑tailed hypothesis; for non‑directional, use a two‑tailed hypothesis. Always operationalise the independent and dependent variables—define exactly how they will be manipulated or measured.
研究问题必须清晰、聚焦且可检验。它指导整个研究。由此，你提出备择假设（H₁），预测有显著效果或关系，以及零假设（H₀），陈述将没有效果或关系。对于有方向的预测，使用单尾假设；对于无方向的预测，使用双尾假设。始终对自变量和因变量进行操作化定义——明确它们如何被操纵或测量。
2. Experimental Methods and Designs | 实验方法与设计
Laboratory experiments offer high control over extraneous variables but often lack ecological validity. Field experiments are conducted in natural settings with some manipulation, balancing control and realism. Natural experiments exploit naturally occurring changes in the independent variable; the researcher does not manipulate it directly. Quasi‑experiments compare groups based on pre‑existing characteristics (e.g., gender) without random assignment.
实验室实验对外部变量控制高，但常缺乏生态效度。现场实验在自然环境中进行一定的操纵，平衡了控制与真实性。自然实验利用自变量的自然发生变化；研究者不直接操纵它。准实验根据预先存在的特征（如性别）比较组别，没有随机分配。
For design, independent groups use different participants in each condition, avoiding order effects but requiring more participants. Repeated measures use the same participants in all conditions, reducing individual differences but risking order effects. Matched pairs match participants on key traits, controlling individual differences without order effects but demanding extensive pre‑testing.
在实验设计方面，独立组设计在每个条件下使用不同的参与者，避免了顺序效应，但需要更多被试。重复测量设计在所有条件下使用相同的参与者，减少个体差异，但存在顺序效应的风险。匹配对设计根据关键特征匹配参与者，控制了无需顺序效应的个体差异，但需要大量的预测试。
3. Controlling Variables and Extraneous Influences | 控制变量与无关影响
Confounding variables change systematically with the independent variable and can ruin an experiment's internal validity. Extraneous variables (situational, participant, researcher effects) must be minimised. Standardise instructions and procedures to ensure consistency. Randomly allocate participants to conditions to avoid selection bias. Use control groups to provide a baseline for comparison. Single‑blind designs keep participants unaware of the condition, reducing demand characteristics. Double‑blind designs ensure neither participants nor the experimenter knows the allocation, eliminating both demand characteristics and experimenter bias.
混淆变量与自变量系统性地变化，可能破坏实验的内部效度。无关变量（情境、参与者、研究者效应）必须最小化。标准化指导语和程序以确保一致性。随机分配参与者到各条件以避免选择偏差。使用控制组提供比较的基线。单盲设计不让参与者知道所处的条件，减少需求特征。双盲设计确保参与者和实验者都不知道分组，消除需求特征和实验者偏差。
4. Sampling Techniques and Participant Selection | 抽样技术与被试选择
The sample must represent the target population for generalisable findings. Opportunity sampling selects readily available participants—quick but often biased. Volunteer (self‑selected) sampling attracts motivated individuals but may produce an unrepresentative sample. Random sampling gives every member of the population an equal chance of being chosen, reducing bias but not guaranteeing representativeness. Stratified sampling identifies subgroups (strata) and draws randomly from each in proportion to their population share; this ensures representativeness but requires knowledge of population characteristics and can be time‑consuming.
样本必须代表目标人群才能得出可推广的结论。机会抽样选择现成的参与者——快速但经常有偏差。志愿者（自选）抽样吸引有动机的个体，但可能产生缺乏代表性的样本。随机抽样让总体中的每个成员被选中的机会均等，减少偏见，但不保证代表性。分层抽样识别亚组（层），并按其在总体中所占比例随机抽取；这确保了代表性，但需要了解总体特征且可能费时。
5. Data Collection and Levels of Measurement | 数据收集与测量层级
Common data‑collection methods include observations (structured/unstructured, participant/non‑participant), self‑report (questionnaires, interviews), and psychological tests. Data are classified into four levels of measurement. Nominal data consist of categories (e.g., favourite colour). Ordinal data have a meaningful order but unequal intervals (e.g., class rank). Interval data have equal intervals but no true zero (e.g., IQ scores). Ratio data have equal intervals and a true zero, allowing ratio comparisons (e.g., reaction time in milliseconds). Identifying the correct level is crucial for choosing appropriate descriptive and inferential statistics.
常见的数据收集方法包括观察（结构化/非结构化、参与/非参与）、自陈报告（问卷、访谈）和心理测验。数据分为四个测量层级。名义数据由类别组成（如最喜欢的颜色）。顺序数据有有意义的顺序但间隔不等（如班级排名）。间隔数据具有等距间隔但没有绝对的零点（如智商分数）。比率数据有等距间隔和绝对零点，允许比率比较（如反应时毫秒）。正确识别层级对于选择合适的描述和推论统计至关重要。
6. Descriptive Statistics | 描述统计
Descriptive statistics summarise data patterns. Measures of central tendency include the mean (average), which is sensitive to outliers; the median (middle value), robust for skewed distributions; and the mode (most frequent), useful for nominal data. Measures of dispersion include the range (highest minus lowest) and the standard deviation, which quantifies the average distance of scores from the mean. For interval/ratio data, report mean and standard deviation; for ordinal data, report median and range. Bar charts are used for discrete categories, while histograms and scattergraphs display continuous data and correlations.
描述统计概括数据模式。集中趋势的测量包括均值（平均数），对异常值敏感；中位数（中间值），对于偏态分布稳健；众数（最常见的值），对名义数据有用。离散量的测量包括全距（最大值减最小值）和标准差，量化了分数与均值之间的平均距离。对于间隔/比率数据，报告均值和标准差；对于顺序数据，报告中位数和全距。条形图用于离散类别，而直方图和散点图显示连续数据与相关关系。
7. Inferential Statistics: Selecting and Justifying Tests | 推论统计：选择与论证检验
Inferential statistics allow you to draw conclusions about a population from sample data. The selection of a test depends on three criteria: (1) the type of hypothesis—testing a difference or a correlation; (2) the experimental design—independent groups, repeated measures, or matched pairs; and (3) the level of measurement. For nominal data, use the chi‑square (χ²) test of association or goodness of fit. For ordinal data in a repeated measures (or matched pairs) test of difference, use the Wilcoxon signed‑rank test; for independent groups, use the Mann‑Whitney U test. For ordinal correlation, use Spearman's rho (rₛ). For interval/ratio data with independent groups, use the unrelated t‑test; with repeated measures, use the related t‑test (though these are rarely required to calculate by hand in Cambridge Psychology). For a single‑sample sign test, use the binomial sign test. Always justify your choice by linking to the design and data type.
推论统计使你能够从样本数据中得出关于总体的结论。检验的选择取决于三个标准：（1）假设类型——检验差异还是相关；（2）实验设计——独立组、重复测量或匹配对；（3）测量层级。对于名义数据，使用卡方（χ²）关联检验或拟合优度检验。对于顺序数据，在重复测量（或匹配对）的差异检验中使用威尔科克森符号秩检验；对于独立组，使用曼-惠特尼U检验。对于顺序数据的相关，使用斯皮尔曼等级相关（rₛ）。对于间隔/比率数据，独立组使用无关t检验，重复测量使用相关t检验（尽管剑桥心理学很少要求手工计算这些）。对于单一样本符号检验，使用二项符号检验。始终通过联系实验设计和数据类型来论证你的选择。
χ² = Σ (O − E)² / E, where O = observed frequency, E = expected frequency.
U = n₁ n₂ + n₁(n₁+1)/2 − R₁, where R₁ is the sum of ranks for group 1.
8. Ethical Considerations in Practical Investigations | 实践调查中的伦理考量
All practical work must adhere to ethical guidelines, such as those from the British Psychological Society (BPS). Obtain informed consent from
Published by TutorHao | Year 13 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