📚 AS CAIE Psychology: Common Misconceptions and Corrections | AS CAIE 心理学：常见误区与纠正方法
Many AS-level psychology students begin the course with a set of intuitive beliefs about what psychology is and how research should be conducted. While curiosity is valuable, these preconceptions often turn into stubborn misconceptions that undermine exam performance. This article identifies ten of the most persistent misunderstandings in the CAIE 9990 syllabus and explains how to correct them with accurate knowledge and appropriate terminology.
许多AS心理学学生在开始课程时都带着一套关于心理学是什么以及研究应该如何进行的直觉信念。虽然好奇心很有价值，但这些先入为主的观念往往会变成顽固的误区，不利于考试表现。本文找出CAIE 9990大纲中最常见的十个误解，并解释如何用准确的知识和恰当的术语加以纠正。
1. Psychology is Just Common Sense | 心理学只是常识
A widespread belief is that psychology simply confirms what everyone already knows. People point to proverbs like 'opposites attract' and 'birds of a feather flock together' as evidence, ignoring the fact that these sayings contradict each other. Psychological research, however, has repeatedly shown that similarity in attitudes, background and values is a far stronger predictor of attraction than complementarity (Byrne, 1971). What feels like common sense after hearing a result is actually hindsight bias — the tendency to see an outcome as predictable once you know the answer. AS students must learn to distinguish anecdotal belief from evidence supported by controlled observation and measurement.
一种普遍的看法是，心理学只是证实了每个人都已经知道的事情。人们用“异性相吸”和“物以类聚”这类谚语作为证据，却忽视了这些谚语本身相互矛盾。然而，心理学研究反复证明，态度、背景和价值观的相似性比互补性更能预测吸引力（Byrne, 1971）。在知晓结果之后觉得这是常识，实际上是一种事后聪明偏差——即一旦知道了答案，就倾向于认为结果是可预测的。AS学生必须学会将轶事信念与受控观察和测量支持的证据区分开来。
2. Correlation Implies Causation | 相关关系就是因果关系
One of the most frequent errors in exam answers is treating a correlational relationship as proof of cause and effect. For instance, a positive correlation between ice cream sales and drowning incidents does not mean that buying ice cream causes drowning. A third variable — hot weather — increases both swimming and ice cream consumption. In CAIE AS Psychology, you must describe a correlation with precise language: a higher score in one variable is associated with a higher or lower score in another, but you cannot infer causation without an experimental design. When evaluating correlational studies, always point out the lack of manipulation of the independent variable and the inability to rule out confounding variables.
在考试作答中最常见的错误之一是将相关关系视为因果关系的证明。例如，冰激凌销量与溺水事件之间的正相关并不意味着购买冰激凌会导致溺水。第三个变量——炎热的天气——同时增加了游泳和冰激凌消费。在CAIE AS心理学中，你必须用确切的语言描述相关：一个变量的较高分数与另一个变量的较高或较低分数相关联，但如果没有实验设计，你无法推断因果关系。在评价相关研究时，务必要指出缺乏对自变量的操纵以及无法排除混淆变量的问题。
3. Confusing Laboratory and Natural Experiments | 混淆实验室实验与自然实验
Many learners assume that all experiments involve a researcher actively manipulating the independent variable (IV) and randomly allocating participants to conditions. This is true for laboratory and field experiments, but a natural experiment is fundamentally different. In a natural experiment, the IV is an existing difference between people that the researcher does not control — for example, comparing children who were in a day-care setting with those raised at home, or comparing flood survivors with a control group. Because participants cannot be randomly assigned, natural experiments are quasi-experiments; they suffer from pre-existing participant variables and therefore cannot establish causality with the same confidence as a true experiment. Correcting this misconception ensures you use the right evaluation points for each design.
许多学习者以为所有实验都包含研究者主动操纵自变量（IV）并将参与者随机分配到不同条件。这对实验室实验和现场实验来说是正确的，但自然实验有着根本不同。在自然实验中，IV是人群本身存在的一种差异，研究者并不加以控制——例如，比较那些曾入读日托机构的儿童和在家抚养的儿童，或者将洪水幸存者与对照组进行比较。由于参与者无法被随机分配，自然实验是准实验；它们会受到先前存在的参与者变量影响，因此无法像真实验那样确切地建立因果关系。纠正这一误区能确保你为每种设计选用正确的评价要点。
4. Misunderstanding the Null Hypothesis | 误解零假设
A common mistake is to write the null hypothesis (H₀) as if it were the researcher' null hypothesis states that any difference or correlation observed is due to chance alone — ive (or experimental) hypothesis (H₁) predicts a real effect, whether directional (H₁: μ₁ > μ₂) or non-directional (H₁: μ₁ ≠ μ₂). Students sometimes incorrectly present a directional hypothesis as their null, which would make it impossible to test the statistical significance logically. You should also remember that you never accept the null hypothesis; you either reject it or fail to reject it. This precise understanding is essential when writing hypotheses in the exam.
一个常见错误是将零假设（H₀）写得好像它是研究者本人的预测一样。实际上，零假设陈述的是任何观察到的差异或相关纯属偶然——例如 H₀: μ₁ = μ₂。备择假设（或实验假设）（H₁）则预测存在真实的效应，可以是方向性的（H₁: μ₁ > μ₂）或非方向性的（H₁: μ₁ ≠ μ₂）。学生有时会把方向性假设错误地写成零假设，这将使得逻辑上无法检验统计显著性。你还应当记住，永远不能“接受”零假设，只能拒绝它或未能拒绝它。在考试中撰写假设时，这种准确的理解至关重要。
5. Research 'Proves' Theories | 研究“证明”理论
In everyday language, people often say a study proved something, but in science the word 'prove' is avoided. Psychological findings can support or refute a theory, but they never provide absolute proof because new evidence may later emerge, and all studies contain limitations. Karl Popper's principle of falsification reminds us that scientific claims must be testable and potentially disprovable. In CAIE AS examination responses, avoid writing 'this proves that...' and instead use terms such as 'suggests', 'supports', 'provides evidence for' or 'indicates'. This demonstrates your grasp of the tentative, self-correcting nature of psychological knowledge.
在日常语言中，人们常会说某个研究“证明”了某事，但在科学中，“证明”一词是应当避免的。心理学发现可以支持或反驳某一理论，但永远不能提供绝对的证明，因为新的证据可能会在日后出现，而且所有研究都包含局限性。卡尔·波普尔的证伪原则提醒我们，科学主张必须具有可检验性和潜在可证伪性。在CAIE AS考试作答中，应避免写“这证明了……”，而要使用“表明”、“支持”、“为……提供了证据”或“指出”等措辞。这能体现你理解心理学知识是暂时性的、自我修正的这一特性。
6. Sampling Bias is Unimportant | 抽样偏差不重要
Students often overlook the impact of the sampling method when evaluating a study, assuming that any group of participants will produce broadly applicable results. Yet a sample drawn exclusively from university psychology undergraduates, for example, is unlikely to represent the wider population in terms of age, education level, cultural background or socio-economic status. This lack of representativeness limits the generalisability of the findings. To correct this thinking, always link the sampling method (random, stratified, volunteer, opportunity) to the concept of population validity. When a study uses a biased sample, you must explain exactly which characteristics are over- or under-represented and how this might distort the results.
学生在评价研究时常常忽视抽样方法的影响，以为任何一组参与者都能产生广泛适用的结果。然而，一个完全由大学心理学本科生组成的样本，在年龄、教育水平、文化背景或社会经济地位方面都不太可能代表更广泛的总体。这种代表性的缺乏会限制研究结果的可推广性。要纠正这一想法，必须始终将抽样方法（随机、分层、自愿、方便）与总体效度这一概念联系起来。当一项研究采用了有偏差的样本时，你应具体说明哪些特征被高估或低估了，以及这可能会如何扭曲结果。
7. Observations are Always Objective | 观察总是客观的
Many candidates mistakenly believe that because observations involve watching real behaviour, they are free from bias. In practice, observer bias — where the researcher's expectations influence what is noticed and recorded — can seriously affect reliability and validity. For example, if an observer expects boys to be more aggressive than girls, they may interpret the same rough-and-tumble play differently based on the child's gender. To reduce subjectivity, researchers use structured coding systems, clearly operationalised behavioural categories, and multiple observers to calculate inter-observer reliability. Correcting this misconception also helps you to discuss the trade-off between ecological validity and control in naturalistic versus controlled observations.
许多考生错误地以为，因为观察涉及观看真实行为，所以它们不存在偏差。事实上，观察者偏差——即研究者的期望会影响他们注意到和记录的内容——会严重影响信度和效度。例如，如果观察者预期男孩比女孩更具攻击性，他们可能会根据孩子的性别对同样的打闹行为做出不同的解读。为了降低主观性，研究者会使用结构化的编码系统、操作性定义清晰的行为类别，并安排多位观察者以计算观察者间信度。纠正这一误区还有助于你讨论自然观察与受控观察之间在生态效度和控制程度上的权衡。
8. Ethical Considerations are Just Bureaucracy | 伦理考虑只是官僚程序
Some learners treat ethics as a box-ticking exercise that gets in the way of important research. However, historically, the absence of ethical safeguards caused real harm. Milgram's (1963) study of obedience induced extreme stress in participants who believed they were delivering dangerous electric shocks, while Watson and Rayner's (1920) Little Albert experiment intentionally conditioned a fear response in an infant without any long-term desensitisation or consent. Today, the BPS Code of Ethics and Conduct requires psychologists to observe principles of informed consent, avoidance of deception unless scientifically justified, debriefing, the right to withdraw, confidentiality and protection from physical and psychological harm. When evaluating a study in the AS exam, linking its procedures to these ethical guidelines demonstrates a high-level critical perspective.
一些学习者将伦理视为一项打勾练习，阻碍了重要的研究。然而，历史上，伦理保障的缺失曾造成切实的伤害。Milgram（1963）的服从研究让相信自己在施加危险电击的参与者产生了极度压力，而Watson和Rayner（1920）的小阿尔伯特实验则故意在一名婴儿身上诱发恐惧反应，却没有进行任何长期的脱敏或征得同意。如今，英国心理学会（BPS）的伦理行为准则要求心理学家遵守知情同意、避免欺骗（除非有科学理由）、事后解说、退出权、保密以及保护免受身心伤害等原则。在AS考试中对研究进行评价时，将其实验程序与这些伦理指南相联系，能够展示高层次的批判性视角。
9. Replication is Unnecessary | 重复研究无用
Students occasionally argue that once a well-known study has been published, repeating it wastes time and resources. In fact, replication is a cornerstone of the scientific method. A single study might produce significant results by chance (Type I error) or fail to detect a true effect due to a small sample (Type II error). If an experiment cannot be replicated — as highlighted by the recent 'replication crisis' in psychology — its findings lack reliability and should be treated with caution. In CAIE AS Psychology, you can strengthen your evaluations by pointing out that test-retest reliability and independent replications improve confidence in the consistency of psychological evidence.
学生偶尔会说，一旦某项著名研究已经发表，重复它就是在浪费时间和资源。事实上，可重复性是科学方法的基石。单一研究可能偶然地产生显著结果（I类错误），或因样本量小而未能检测到真实效应（II类错误）。如果一项实验无法被重复——正如近期心理学“重复危机”所凸显的那样——其发现便缺乏信度，应谨慎看待。在CAIE AS心理学中，你可以通过指出重测信度和独立重复能够提升对心理学证据一致性的信心，来增强你的评价。
10. Self-Report Data is Always Accurate | 自我报告数据总是准确的
Questionnaires and interviews seem straightforward, leading many students to assume that participants will naturally give honest, accurate answers. The reality is quite different. Social desirability bias can cause respondents to give answers they think are more socially acceptable rather than truthful ones. Demand characteristics may lead participants to guess the aim of the study and alter their responses accordingly. In addition, memory flaws and misinterpretation of questions can reduce validity. Correcting this misconception means recognising that self-report measures should often be triangulated with other methods, such as observations or physiological recordings, and that researchers can use techniques like anonymity assurance, filler items and reverse-scored questions to minimise bias.
问卷和访谈看似简单直接，致使许多学生以为参与者理所当然会给出诚实准确的回答。现实却大不相同。社会期望偏差可能导致受访者给出他们认为更能被社会接受的答案，而非真实的答案。需求特征可能让参与者去猜测研究目的并相应改变自己的回答。此外，记忆缺陷和对问题的误解也会降低效度。要纠正这一误区，就要意识到自我报告测量通常需要结合其他方法（如观察或生理记录）进行三角验证，并且研究者可以采用匿名保证、填充题以及反向计分题等技术来减少偏差。
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