📚 Year 13 AQA Psychology: Key Points for Experimental/Practical Assessment | Year 13 AQA 心理学：实验/实践考核要点
The AQA A-level Psychology specification places strong emphasis on research methods and the ability to design, conduct, analyse, and evaluate psychological investigations. In Year 13, you are expected to demonstrate competencies in experimental design, hypothesis testing, data handling, and the critical appraisal of research evidence. This article distils the essential points for the experimental/practical assessment, covering everything from variables and controls to inferential statistics and report writing.
AQA A-level 心理学大纲高度重视研究方法以及设计、实施、分析和评价心理学研究的能力。在 Year 13 阶段，你需要展示实验设计、假设检验、数据处理以及对研究证据进行批判性评估的能力。本文提炼了实验/实践考核的核心要点，涵盖变量与控制、推断统计和报告撰写等方面。
1. Understanding Variables and Operationalisation | 变量及其操作化
Every experiment must clearly define the independent variable (IV) and the dependent variable (DV). The IV is the factor manipulated by the researcher, while the DV is the factor measured. Operationalisation means stating precisely how each variable will be manipulated or measured to allow replication and validity. For example, 'memory' could be operationalised as the number of words correctly recalled from a list of 20.
每个实验都必须清晰地界定自变量（IV）和因变量（DV）。自变量是研究者操纵的因素，因变量是被测量的因素。操作化是指精确说明每个变量如何被操纵或测量，以便重复实验并确保效度。例如，“记忆力”可以操作化为从 20 个单词中正确回忆出的单词数量。
You should also consider extraneous variables that might confound results, such as participant variables, situational variables, and experimenter effects. Confounding variables vary systematically with the IV and can invalidate an experiment.
你还应该考虑可能混淆结果的额外变量，如被试变量、情境变量和实验者效应。混淆变量与自变量系统性地共同变化，可能会使实验失效。
2. Hypotheses Formulation | 假设构建
A clear, testable hypothesis is the foundation of any experiment. You need to distinguish between a directional (one-tailed) hypothesis, which predicts the direction of the effect (e.g., 'participants who drink 200 mg of caffeine will recall more words than those who drink a placebo'), and a non-directional (two-tailed) hypothesis, which simply predicts a difference without stating direction. The null hypothesis states that any observed difference or relationship is due to chance alone.
一个清晰、可检验的假设是任何实验的基础。你需要区分方向性（单尾）假设，它预测效应的方向（如“摄入 200 毫克咖啡因的被试会比服用安慰剂的被试回忆更多单词”），以及非方向性（双尾）假设，它只预测存在差异但不说明方向。零假设则表明任何观察到的差异或关系完全是由偶然因素造成的。
3. Experimental Designs | 实验设计
AQA expects you to know three main experimental designs: independent groups design (different participants in each condition), repeated measures design (same participants in all conditions), and matched pairs design (participants paired on a relevant variable, then split between conditions). Each design has strengths and limitations related to participant variables, order effects, and demand characteristics.
AQA 要求你掌握三种主要的实验设计：独立组设计（不同被试在不同条件下）、重复测量设计（相同被试经历所有条件）和配对组设计（被试根据相关变量配对后分配到不同条件）。每种设计在控制被试变量、顺序效应和要求特征方面各有优缺点。
For example, repeated measures eliminates individual differences but introduces order effects, which can be controlled through counterbalancing. Independent groups avoids order effects but requires more participants and can be confounded by participant variables.
例如，重复测量设计消除了个体差异，但引入了顺序效应，可以通过平衡法来控制。独立组设计避免了顺序效应，但需要更多被试，并且可能受被试变量混淆。
4. Control Over Extraneous Variables | 控制外来变量
To increase internal validity, you must control extraneous variables. Common control techniques include random allocation (for independent groups), counterbalancing (for repeated measures), standardisation of procedures and instructions, and single- or double-blind procedures to control for demand characteristics and experimenter bias.
为了提高内部效度，你必须控制外来变量。常用的控制技术包括随机分配（用于独立组）、平衡法（用于重复测量）、程序和指导语的标准化，以及单盲或双盲程序来控制要求特征和实验者偏见。
Pilot studies help identify potential confounding variables and refine the procedure before the main experiment. A pilot study uses a small number of participants to test the design and materials.
试测研究有助于在主实验前识别潜在的混淆变量并完善程序。试测研究使用少量被试来检验设计和材料。
5. Sampling Techniques | 取样技术
Your choice of sampling method affects the generalisability of findings. You should be familiar with random sampling (every member of the target population has an equal chance), stratified sampling (proportional representation from subgroups), opportunity sampling (using those readily available), volunteer sampling (self-selected), and systematic sampling (selecting every nth person). Discuss the representativeness and potential biases of each method.
取样方法的选择会影响研究结果的可推广性。你应该熟悉随机取样（目标总体中每个成员都有同等机会被选中）、分层取样（从子群中按比例抽取）、机会取样（使用最方便获取的被试）、志愿者取样（自我选择）和系统取样（每隔若干人选一个）。要讨论每种方法的代表性和潜在偏差。
6. Ethical Guidelines | 伦理准则
All psychological research in the UK must adhere to the British Psychological Society (BPS) Code of Ethics. Key ethical principles include obtaining informed consent (or parental consent for under-16s), avoiding deception unless absolutely necessary and justified, providing the right to withdraw, ensuring confidentiality, protecting from physical and psychological harm, and conducting a full debrief at the end.
英国所有的心理学研究都必须遵守英国心理学会（BPS）伦理准则。关键伦理原则包括获取知情同意（或对未满 16 岁者获取父母同意）、避免除非必要且经论证的欺骗、提供退出权、确保保密、避免身心伤害，以及在研究结束后进行充分的事后解释。
If deception is used, participants must be told the true nature of the study during debrief and be given the opportunity to withhold their data. Addressing ethical issues strengthens the credibility of your practical work.
如果使用了欺骗，必须通过事后解释告知参与者研究的真实性质，并给予他们撤回数据的机会。处理好伦理问题可以增强实践作业的可信度。
7. Data Collection and Levels of Measurement | 数据收集与测量层次
You may collect data through observation, questionnaires, or direct measurement in an experiment. Crucially, you must recognise the level of measurement: nominal data (categories, e.g., yes/no), ordinal data (ranked but intervals not equal, e.g., Likert-scale ratings), and interval/ratio data (equal intervals, true zero, e.g., reaction time in milliseconds). The level of measurement determines which statistical test is appropriate.
你可以通过观察、问卷或实验中的直接测量来收集数据。关键是要识别测量的层次：称名数据（分类，例如是/否）、顺序数据（排序但间隔不等，例如李克特量表评分），以及等距/比率数据（等距单位，有绝对零点，例如以毫秒计的反应时）。测量层次决定了适用何种统计检验。
8. Descriptive Statistics | 描述统计
Descriptive statistics summarise and present data. You should be able to calculate and interpret measures of central tendency (mean, median, mode) and measures of dispersion (range, variance, standard deviation SD). For normally distributed interval data, the mean and SD are appropriate; for skewed distributions or ordinal data, the median and interquartile range are more suitable.
描述统计用于概括和呈现数据。你应该能够计算和解释集中趋势量数（平均数、中位数、众数）和离散量数（极差、方差、标准差 SD）。对于正态分布的等距数据，平均数与标准差是合适的；对于偏态分布或顺序数据，中位数和四分位距更为恰当。
Graphical displays such as bar charts, histograms, scatterplots, and box plots should be correctly labelled and scaled. Understand that the SD formula involves squaring deviations, summing them, dividing by N-1 (sample SD), and taking the square root. A simple representation is shown below.
条形图、直方图、散点图和箱线图等图形展示应正确标注和标度。理解标准差公式包含偏差平方后求和、除以 n–1（样本标准差），再开平方。简明表示为：
SD = √[ Σ(x – x̄)² ÷ (n – 1) ]
9. Inferential Statistical Tests | 推断统计检验
Inferential tests allow you to determine whether results are statistically significant, i.e., whether the null hypothesis can be rejected. You need to select the correct test based on the design and level of measurement. The AQA specification includes these tests: Chi-squared test (χ²) for nominal data and an independent groups design; Mann-Whitney U test for ordinal data with independent groups; Wilcoxon signed-rank test for ordinal data with repeated measures or matched pairs; Related t-test for interval data with repeated measures/matched pairs; Unrelated t-test for interval data with independent groups; Spearman's rho (ρ) for correlation with ordinal data; and Pearson's r for correlation with interval data. The Sign test may also be used for nominal difference with repeated measures.
推断检验帮助你判断结果是否具有统计显著性，即能否拒绝零假设。你需要根据实验设计和测量层次选择正确的检验。AQA 大纲包括以下检验：卡方检验（χ²）适用于称名数据和独立组设计；曼-惠特尼 U 检验适用于顺序数据和独立组设计；威尔科克森符号秩检验适用于顺序数据且为重复测量或配对组设计；相关 t 检验适用于等距数据与重复测量/配对组；独立 t 检验适用于等距数据与独立组；斯皮尔曼 ρ 适用于顺序数据的相关分析；皮尔逊 r 适用于等距数据的相关分析。符号检验也可用于重复测量的称名差异。
All tests produce an observed (calculated) value that is compared against a critical value from a statistical table at a given significance level (usually p < 0.05). You must know how to determine degrees of freedom (df) for t-tests and χ², or use N for non-parametric tests. For example, for a chi-squared test of association, the formula is:
所有检验都会得出一个观测（计算）值，将其与特定显著性水平（通常 p < 0.05）下的统计表临界值进行比较。你必须知道如何确定 t 检验和卡方检验的自由度（df），或在非参数检验中使用 N。例如，对于卡方关联性检验，公式为：
χ² = Σ (O – E)² ÷ E
where O is the observed frequency and E is the expected frequency. When using the Sign test, you count the number of '+' and '-' differences, ignore ties, and the observed value is the smaller of the number of '+'s or '-'s.
其中 O 是观察频次，E 是期望频次。使用符号检验时，计算“+”“-”差异的数目，忽略平局，观测值是“+”或“-”数目中较小的一个。
10. Reliability and Validity in Experiments | 实验的信度与效度
Reliability refers to the consistency of results. You can improve reliability by standardising procedures, using operationalised definitions, and employing test-retest or inter-rater reliability checks. Validity concerns whether you are measuring what you intend to measure. Internal validity can be threatened by confounding variables, demand characteristics, and social desirability bias; external validity relates to generalisability across settings, people, and times (ecological, population, and temporal validity).
信度指结果的一致性。通过标准化程序、使用操作化定义以及进行重测信度或评分者间信度检查，可以提高信度。效度涉及你是否测量到了想要测量的内容。内部效度可能受到混淆变量、要求特征和社会期望偏差的威胁；外部效度则涉及跨情境、人群和时间的可推广性（生态效度、人口效度和历史效度）。
11. Writing the Lab Report | 撰写实验报告
Your practical report should follow the standard sections: Abstract (summary of aims, method, results, conclusion), Introduction (literature review and rationale with aims/hypotheses), Method (design, participants, apparatus/materials, procedure, ethics), Results (descriptive and inferential statistics, graphs), Discussion (explanation of findings, relationship to previous research, limitations, modifications), and References. Appendices may include raw data, calculation sheets, and consent forms.
你的实验报告应包含以下标准部分：摘要（目标、方法、结果、结论的总结）、引言（文献综述和理论依据，包含目标/假设）、方法（设计、被试、仪器/材料、程序、伦理）、结果（描述性与推断统计、图表）、讨论（对发现的
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