📚 High-Frequency Topics and Common Mistakes in CAIE A-Level Psychology (Year 13) | 高频考点与易错题深度解析
For Year 13 students sitting the CAIE A-Level Psychology (9990) exam, the step up from AS to A2 involves not only deeper theoretical content but also a more sophisticated application of research methodology, evaluation skills, and statistical reasoning. This article identifies the topics that appear most frequently across past papers, highlights where candidates routinely lose marks, and provides targeted guidance to strengthen your exam responses.
对于参加 CAIE A-Level 心理学 (9990) 考试的 Year 13 同学来说，从 AS 到 A2 的提升不仅意味着更深入的理论内容，还要求更精细地运用研究方法、评估技巧和统计推理。本文梳理了历年真题中出现频率最高的考点，指出考生普遍失分的地方，并给出针对性的提升建议，帮助你强化答题表现。
1. Navigating Research Methods: Experiments vs. Non‑Experimental Designs | 研究方法导航：实验与非实验设计对比
A classic high‑stakes distinction tested in both short‑answer and essay questions is between laboratory, field, natural, and quasi‑experiments. Many candidates mistakenly label a natural experiment as a field experiment simply because it takes place outside a lab. The defining feature is the manipulation of the independent variable (IV): in a natural experiment the IV changes naturally (e.g., a policy reform), whereas in a field experiment the researcher deliberately alters the IV in a real‑world setting.
无论是在简答还是论述题中，区分实验室实验、现场实验、自然实验和准实验都是一项高频且关键的考点。很多考生仅因研究发生在实验室之外就误将其标为现场实验。区分的核心在于自变量的操纵方式：自然实验中自变量是自然变化的（如政策改革），而现场实验中研究者在真实环境中主动操纵自变量。
Another frequent error is conflating correlational designs with quasi‑experiments. A quasi‑experiment involves a pre‑existing participant variable (e.g., gender, clinical diagnosis) and attempts to measure a cause‑effect relationship, but it lacks full random allocation. A correlational design, however, measures the relationship between two co‑varying variables without an IV–DV framework. Students often credit a quasi‑experiment with establishing causality, which cannot be done due to participant variables acting as confounds.
另一个常见错误是将相关设计与准实验混淆。准实验包含一个既存的被试变量（如性别、临床诊断）并试图测量因果关系，但缺少完全随机分配。相关设计则测量两个共同变化的变量之间的关系，没有明确的 IV‑DV 框架。考生常误认为准实验可以确立因果关系，但由于被试变量本身就是混淆变量，实际上无法得出因果结论。
2. Validity and Reliability in Assessment Questions | 评估题中的效度与信度
When asked to “assess the validity” of a study or measurement, candidates often recite definitions without applying them to the scenario. Internal validity concerns whether the IV truly caused the change in the DV, requiring you to identify specific confounds (e.g., demand characteristics, investigator effects). External validity asks how far findings generalise – you must name the type (population, ecological, temporal) and explain why the sample or setting limits generalisation.
当题目要求“评估研究的效度”时，考生常常只背诵定义而没有结合具体情境应用。内部效度关注自变量是否确实导致了因变量的变化，需要你指出具体的混淆变量（如需求特征、研究者效应）。外部效度考察结果的推广程度，你必须指明类型（群体效度、生态效度、时间效度）并解释样本或环境为何限制了推广性。
Reliability is frequently conflated with validity. In the context of structured interviews or behavioural categories, inter‑rater reliability requires a clear, operationalised coding scheme. When evaluating self‑report measures, test‑retest or split‑half reliability methods should be discussed with reference to the time interval or item consistency. A common mistake is simply stating “the study had high reliability” without referencing how it was established or why it matters for the specific measure.
信度常常与效度混淆。在结构化访谈或行为类别的情境中，评分者间信度需要一个明确、可操作化的编码方案。评估自陈报告工具时，应结合时间间隔或题目一致性讨论重测信度或分半信度。常见的错误是仅仅说“该研究信度高”，而不提及该信度如何确立或它对该具体测量为何重要。
3. Ethical Considerations in A2 Evaluations | A2评价中的伦理考量
At A2, ethical evaluation must go beyond the obvious “informed consent was not obtained.” High‑scoring answers apply the cost–benefit model: weigh the potential harm to participants against the societal or scientific benefits, and then justify whether the study would pass an ethics committee review today. Candidates often overlook the right to withdraw in naturalistic observations where participants may be unaware they are being watched.
在 A2 阶段，伦理评价必须超越“未获得知情同意”这样肤浅的表述。高分答案会运用成本‑收益模型：权衡对被试的潜在伤害与社会或科学收益，然后论证该研究在今天能否通过伦理委员会审查。考生常忽视在自然观察中参与者的退出权利，因为被试可能根本不知道自己正在被观察。
Another pitfall is failing to distinguish between deception and lack of full disclosure. Deception involves actively misleading participants, whereas many studies merely withhold the full aim to avoid demand characteristics – this partial disclosure is often justified if a debrief is provided. When citing specific core studies, such as Bandura’s aggression research, you must link the ethical issue (e.g., psychological harm from observing aggression) to the 2020 BPS Code of Ethics and suggest realistic improvements.
另一个易错点是未能区分欺骗与不完全披露。欺骗涉及主动误导被试，而许多研究只是隐去完整目的以避免需求特征——只要提供了事后解释，这种部分披露往往是可接受的。在引用具体核心研究（如 Bandura 的攻击行为研究）时，你必须将伦理问题（如观察攻击行为造成的心理伤害）与 2020 年 BPS 伦理规范联系起来，并提出切实可行的改进建议。
4. Mastering Core Studies: Baron‑Cohen et al. (1997, 2001) | 掌握核心研究：Baron‑Cohen 等人的眼神测试
Both the original Eyes Task (1997) and its revised version (2001) are examined repeatedly. A high‑frequency error is failing to recall the exact participant groups: the 1997 study compared adults with high‑functioning autism/AS, adults with Tourette syndrome, and neurotypical controls. The 2001 revision used a much larger, more diverse control group and narrowed the forced‑choice options to two per item to reduce chance performance.
原版眼神任务 (1997) 和修订版 (2001) 反复出现在考题中。一个高频错误是记错被试组别：1997 年的研究比较了高功能自闭症/AS 成人、妥瑞氏症成人和神经典型对照组。2001 年修订版采用了规模更大、更多样化的对照组，并将每道题的选项缩减为两个，以降低猜测概率。
Candidates often discuss the ceiling effect in the Happé Strange Stories as a weakness but mislabel the Eyes Task itself. The revised Eyes Task deliberately reduced the number of basic emotion items to avoid a ceiling effect and included more complex mental states. When evaluating validity, note that the task measures a cognitive component of empathy (theory of mind), not affective empathy – a distinction frequently missed in essays that critique the task as lacking ecological validity because real‑life emotion recognition relies on dynamic facial expressions.
考生常将 Happé 奇怪故事任务中的天花板效应作为弱点讨论，却错误地套用在眼神任务上。修订版眼神任务刻意减少了基本情绪条目以避免天花板效应，并增加了更复杂的心理状态。在评估效度时需要注意，该任务测量的是认知共情（心理理论），而非情感共情——这一区别在批评任务缺乏生态效度（因为真实情绪识别依赖动态面部表情）的论述中经常被忽略。
5. Bandura’s Social Learning Theory and Aggression Studies | Bandura 社会学习理论与攻击研究
The 1961 and 1963 Bobo doll studies are staples, but Year 13 candidates must go beyond simple description. You need to articulate the pre‑operationalised behavioural categories: physical aggression (e.g., punching, kicking Bobo), verbal aggression (e.g., specific phrases like “Sock him in the nose”), and non‑aggression. A common mistake is claiming the studies proved that observing aggression causes aggressive behaviour. Bandura’s classic experiments show that children imitate modelled behaviour; they do not directly prove the cause of real‑world violence because the Bobo doll is designed to be hit.
1961 年和 1963 年的波波玩偶实验是必考内容，但 Year 13 考生必须超越简单描述。你需要清楚阐述预先操作化的行为类别：身体攻击（如拳打、脚踢 Bobo）、言语攻击（如“打他的鼻子”等特定话语）和非攻击行为。一个常见错误是声称这些研究证明了观察攻击行为会导致攻击行为。Bandura 的经典实验表明儿童会模仿示范行为，但它们并没有直接证明真实世界暴力行为的成因，因为波波玩偶本身就是设计来被击打的。
When evaluating, address the tension between experimental control and demand characteristics – children might have believed they were expected to hit the doll. Also, the 1963 study demonstrated that filmed and cartoon‑based models produced similar imitative aggression, supporting the generalisability of social learning beyond live models. Many candidates fail to link these findings to the broader theory’s mediating processes: attention, retention, reproduction, and motivation.
在评估时，要讨论实验控制与需求特征之间的张力——孩子们可能以为他们被期望去打玩偶。此外，1963 年的研究证明，电影和卡通形式的示范也能引发类似的模仿攻击行为，支持社会学习超越真人示范的推广性。许多考生未能将这些发现与更广泛理论的中介过程（注意、保持、动作再现和动机）联系起来。
6. Schizophrenia: Classification, Symptoms and Diagnostic Issues | 精神分裂症：分类、症状与诊断争议
Schizophrenia is one of the most popular A2 options and features heavily in essay questions. Candidates must differentiate between positive symptoms (hallucinations, delusions, disorganised speech) and negative symptoms (avolition, affective flattening, alogia). A table comparing the DSM‑5 and ICD‑11 criteria helps clarify diagnostic requirements and the duration of symptoms needed.
精神分裂症是最热门的 A2 选项之一，在论述题中出现频率很高。考生必须区分阳性症状（幻觉、妄想、言语紊乱）和阴性症状（意志缺乏、情感淡漠、言语贫乏）。通过表格对比 DSM‑5 和 ICD‑11 的标准，有助于厘清诊断要求以及症状持续时间的规定。
| Feature | DSM‑5 | ICD‑11 |
|---|---|---|
| Core symptom requirement | At least two of: delusions, hallucinations, disorganised speech, grossly disorganised/catatonic behaviour, negative symptoms (one must be delusions, hallucinations, or disorganised speech) | At least two of: persistent delusions, persistent hallucinations, disorganised thinking, experiences of influence/passivity/control, negative symptoms, grossly disorganised behaviour |
| Duration | Continuous signs for at least 6 months, including ≥1 month of active symptoms | Symptoms present for most of the time over ≥1 month |
| Subtypes | Subtypes removed in DSM‑5 due to poor reliability | Multiple subtypes retained (e.g., paranoid, hebephrenic) |
A classic mistake is arguing that diagnostic manuals are unreliable because “clinicians disagree.” You must cite specific evidence, such as Rosenhan (1973) on labelling or studies showing low inter‑rater reliability for certain subtypes, and then link this to the presence of symptom overlap (e.g., bipolar disorder can present with psychotic features). Simply stating “schizophrenia has low validity” without mentioning comorbidity, cultural bias, or gender bias will cap your marks at the mid‑range.
一个经典错误是泛泛地说诊断手册不可靠，因为“临床医生意见不一致”。你必须引用具体证据，如 Rosenhan (1973) 关于贴标签的研究，或表明某些亚型评分者间信度较低的研究，并进一步联系到症状重叠（例如双相情感障碍也可能出现精神病性症状）。如果只是简单地说“精神分裂症效度低”，却没有提到共病、文化偏差或性别偏差，得分就会被限制在中等水平。
7. Explanations of Schizophrenia: Integrating Biological and Psychological Perspectives | 精神分裂症的解释：整合生物与心理视角
The dopamine hypothesis remains a core biological explanation, but the original version (excess dopamine) has been refined to account for the role of dopamine in the mesolimbic and mesocortical pathways. A high‑frequency mistake is to treat the hypothesis as solely hyper‑dopaminergic; modern accounts recognise hypodopaminergia in the prefrontal cortex contributing to negative symptoms. You should reference supporting research such as the effectiveness of typical (e.g., chlorpromazine) versus atypical antipsychotics (e.g., clozapine).
多巴胺假说依然是核心的生物学解释，但最初的版本（多巴胺过剩）已经过修正，以反映多巴胺在中脑边缘和中脑皮层通路中的作用。一个高频错误是将该假说视作单一的多巴胺功能亢进；现代解释认识到前额叶皮层多巴胺功能低下会导致阴性症状。你应当引用支持性研究，如典型抗精神病药（如氯丙嗪）与非典型抗精神病药（如氯氮平）的效果对比。
On the psychological side, Frith’s cognitive model of faulty metarepresentation is frequently tested. Candidates describe the model but forget to connect it to specific symptoms: delusions of control arise from an inability to monitor self‑generated actions, whereas auditory hallucinations stem from misattributing inner speech to an external source. The family dysfunction explanation (double‑bind theory, expressed emotion) must be evaluated as a risk factor rather than a direct cause, because it has difficulty explaining the biological onset and cannot account for all cases.
在心理学解释方面，Frith 的元表征缺陷认知模型常被考查。考生能描述模型，却忘记将其与具体症状联系起来：被控制妄想源于无法监控自己产生的动作，而幻听则是因为将内部言语错误归因于外部来源。家庭功能失调解释（双重束缚理论、情绪表达）必须被评估为风险因素而非直接原因，因为它难以解释生物学的起病过程，也无法涵盖所有案例。
8. Treatments for Schizophrenia: Evaluating Drug Therapy and CBT | 精神分裂症治疗：评估药物治疗与认知行为疗法
When evaluating antipsychotic medication, side effects are a necessary but insufficient point. You must distinguish between extrapyramidal symptoms (e.g., tardive dyskinesia) typical of first‑generation drugs and metabolic side effects (weight gain, diabetes) linked to second‑generation drugs. An analysis of effectiveness should use specific data, such as response rates from large‑scale trials like CATIE, and discuss the revolving‑door phenomenon linked to non‑adherence.
在评估抗精神病药物时，副作用是必要但不充分的角度。你必须区分第一代药物典型的锥体外系症状（如迟发性运动障碍）和第二代药物引发的代谢副作用（体重增加、糖尿病）。分析有效性时应引用具体数据，如 CATIE 等大规模试验中的应答率，并讨论与不依从相关的“旋转门”现象。
Cognitive Behavioural Therapy (CBT) for psychosis is often evaluated superficially as “effective” without describing what the therapist actually does: normalising the voice‑hearing experience, challenging delusional beliefs through Socratic questioning, and developing coping strategies. A common error is to present CBT and medication as mutually exclusive rather than complementary – the NICE guidelines recommend both in combination. Marks are lost when candidates do not mention that CBT works best for medication‑resistant positive symptoms and shows limited effect on negative symptoms.
针对精神病性症状的认知行为疗法 (CBT) 经常被肤浅地评价为“有效”，却未描述治疗师实际做了什么：将幻听体验正常化、通过苏格拉底式提问挑战妄想信念、以及制定应对策略。一个常见错误是将 CBT 与药物治疗视作互相排斥，而事实上 NICE 指南推荐两者结合使用。如果考生未提及 CBT 对药物抵抗的阳性症状效果最佳，而对阴性症状效果有限，就会失分。
9. Patient Adherence to Medical Advice: Measurement and Improvement Pitfalls | 患者对医嘱的遵从：测量与改善的易错陷阱
Questions on adherence often ask you to contrast subjective measures (self‑report, interviews) with objective measures (pill counts, electronic monitoring, blood tests). Students regularly lose marks by claiming self‑report is “biased” without explaining the direction of bias – social desirability tends to inflate reported adherence rates. Objective measures overcome this but introduce their own limitations: pill counts do not confirm ingestion, and electronic monitors may alter behaviour (Hawthorne effect).
关于遵从性的题目常要求对比主观测量（自陈报告、访谈）与客观测量（药片计数、电子监测、血液检测）。考生常因断言自陈“有偏差”却未解释偏差方向而失分——社会称许性通常会虚报遵从率。客观测量能克服这一点，但也带来其自身的局限：药片计数无法证实药物确实被服用，电子监测则可能改变行为（霍桑效应）。
When proposing strategies to improve adherence, a common mistake is listing techniques without linking them to a theoretical rationale. For example, simplifying the dosing regimen relates to the Health Belief Model’s perceived barriers; using motivational interviewing targets the Stages of Change model. Without these links, the answer remains descriptive rather than evaluative. Furthermore, candidates sometimes overlook that no single strategy works for all patients – a multi‑component intervention tailored to individual barriers yields the strongest evidence.
在提出改善遵从性的策略时，常见错误是罗列技巧却不与理论依据相联系。例如，简化用药方案与健康信念模型中的感知障碍相关；运用动机性访谈则针对行为改变阶段模型。缺少这些联系，答案就停留在描述层面而非评估层面。此外，考生有时会忽略，没有一种策略对所有患者都有效——针对个体障碍量身定制的多组分干预拥有最强的证据支持。
10. Designing a Study: Operationalising Variables and Controlling Confounds | 设计研究：操作化变量与控制混淆变量
The “design a study” question is a reliable 8‑ to 10‑mark item that many candidates approach too vaguely. Operationalisation must be precise: state exactly how the DV will be measured (e.g., “number of seconds out of a 120‑second observation that the child spends within 50 cm of the same‑sex model”) rather than a generic “time spent near model.” Every design decision – independent measures vs. repeated measures, sampling technique – must be justified with reference to the specific research aim.
“设计一个研究”是固定的 8 到 10 分题目，许多考生回答得过于笼统。操作化必须精确：要说明如何具体测量因变量（例如，“在 120 秒的观察中，儿童待在同性示范者 50 厘米范围内的秒数”），而不是泛泛的“待在示范者附近的时间”。每个设计决策——独立组还是重复测量设计、抽样技术——都必须结合具体研究目的进行论证。
The most damaging error is ignoring potential confounds. You must identify at least one participant variable (e.g., pre‑existing attitudes), one situational variable (e.g., time of day, noise), and explain how you would control each. For an observation, inter‑rater reliability must be addressed by giving an example of a behavioural code and a procedure (e.g., two observers independently code 20% of sessions and you compute Cohen’s κ). Many candidates state they will “use a pilot study” but never say what they would look for or refine.
最具破坏性的错误是忽视了潜在的混淆变量。你必须至少指出一个被试变量（如既存态度）、一个情境变量（如时间段、噪音），并解释如何控制每一个变量。对于观察研究，必须通过给出行为编码示例和流程（如两名观察者独立编码 20% 的观察时段并计算 Cohen’s κ）来说明评分者间信度。许多考生声称会“进行试点研究”，但从未说明要寻找什么或如何修正。
11. Interpreting Inferential Statistics: Choosing Tests and Understanding p‑Values | 解读推断统计：选择检验与理解 p 值
Year 13 candidates are expected to select an appropriate inferential test based on the research design and level of measurement, and then justify that choice. The table below summarises the tests most commonly required.
Year 13 考生需要根据研究设计和测量层次选择合适的推断性检验，并给出选择理由。下表总结了最常考到的检验。
| Design | Level of Measurement | Parametric Test | Non‑Parametric Equivalent |
|---|---|---|---|
| Independent measures | Interval/Ratio | Unrelated t‑test | Mann‑Whitney U |
| Repeated measures / Matched pairs | Interval/Ratio | Related t‑test | Wilcoxon signed‑rank |
| Independent measures | Nominal (frequency data) | Chi‑squared (χ²) test of association / independence | |
| Correlational design | Interval/Ratio (at least one variable) | Pearson’s r | Spearman’s rho |
A widespread misinterpretation is that a non‑significant result (p > 0.05) proves the null hypothesis or that the IV has no effect. The correct interpretation is that there is insufficient evidence to reject the null hypothesis – a Type II error may have occurred. Similarly, stating that p < 0.05 means “the results prove the experimental hypothesis” is incorrect; it signifies a low probability (less than 5%) of obtaining the observed result if the null hypothesis were true. Always link the decision about significance to the actual calculated and critical values in your answer.
一个普遍的错误解读是，不显著的结果 (p > 0.05) 证明了虚无假设，或说明自变量没有效果。正确的解释是，没有足够证据拒绝虚无假设——可能发生了第二类错误。同样，声称 p < 0.05 表示“结果证明了实验假设”也是错误的；它意味着在虚无假设正确的前提下，获得当前结果或更极端结果的概率低于 5%。你的答案中必须将显著性的判断与实际的统计值和临界值联系起来。
12. Exam Technique: Structuring Evaluation and Discussion Responses | 考试技巧：构建评估与讨论题答案
In 8‑mark “assess” and 12‑mark “discuss” questions, a point‑by‑point argument structure is essential. Start with a clear statement of the theory, study, or claim, then immediately follow with a strength or weakness supported by evidence. Each evaluation paragraph should contain a signpost (e.g., “A strength of the dopamine hypothesis is…”), the evidence‑based point itself, a brief explanation of why it matters, and a link back to the question. Avoid launching into descriptive paragraphs under the guise of evaluation.
在 8 分的“评估”题和 12 分的“讨论”题中，逐点论证的结构至关重要。先明确提出理论、研究或观点，然后立即用有证据支持的优点或缺点进行跟进。每个评估段落都应包含一个方向指引（如“多巴胺假说的一个优点在于……”）、基于证据的论点本身、简短的重要性解释，以及回扣题干的联系。避免在评估的名义下展开描述性段落。
A recurring mistake is the failure to present a balanced, holistic conclusion. For the highest band, the examiner expects a reasoned judgement that weighs the evidence and considers the overall credibility, application, and ethical dimensions. For instance, when discussing the use of antipsychotic medication, conclude by acknowledging its symptomatic efficacy while recognising the limitations of purely biological approaches, thus supporting a biopsychosocial model of treatment. Balance does not mean giving equal space to strengths and weaknesses but proportionally reflecting the weight of evidence.
反复出现的错误是未能给出平衡、全面的结论。要拿到最高分数段，考官期待一个经过推理的判断，能够权衡证据并考虑整体可信度、应用性和伦理维度。例如，在讨论抗精神病药物的使用时，结论应承认其在缓解症状方面的有效性，同时认识到纯生物学方法的局限性，从而支持治疗的生物‑心理‑社会模型。平衡并不意味着优缺点在篇幅上平分秋色，而是要按证据的分量来相应地分配空间。
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