Edexcel心理学 Social Influence 服从 偏见

Edexcel心理学 Social Influence 服从 偏见

社会心理学这一章在 Edexcel A-Level Psychology（9PS0）里是 Paper 1 的 Section A，内容是服从与偏见两条线。服从考 Milgram 的实验与三个变式，解释考代理理论与社会影响理论；偏见考社会认同理论、现实冲突理论，以及人格、情境、文化三类影响因素。答题时最容易被忽略的是方法部分，问卷、抽样、统计与伦理都写在同一份报告里。这篇按知识点把两半内容串起来，配上数字与答题位置，方便对照自己的笔记补漏。

Social psychology forms Section A of Paper 1 in the Edexcel A-Level Psychology specification (9PS0), and it splits cleanly into obedience and prejudice. The obedience side expects Milgram's experiment, three of his variation studies, agency theory and social impact theory; the prejudice side expects social identity theory, realistic conflict theory, and the personality, situational and cultural factors that change how prejudice is expressed. What students usually neglect is the methods requirement, because the questionnaire, sampling decisions, statistics and ethics all have to be justified in one report rather than recited as a list. Use the sections below as a checklist against your own notes.

服从研究的基线 从 15 伏到 450 伏

Milgram 在 1963 年招募 40 名 20 至 50 岁男性，广告写记忆与学习实验，地点在耶鲁大学。被试当教师，学习者每错一次就升 15 伏，最高 450 伏。学习者由同谋扮演，先抱怨心脏，到 300 伏左右彻底沉默。实验者用四句固定台词施压，从请继续到你没有选择必须继续。26 人推到顶，占 65%，40 人全部越过 300 伏。Edexcel 官网的 specification 9PS0 在 1.1.2 把这项研究列为必学内容。很多学生以为 450 伏那一步最难，真正让人停手的是学习者沉默之后的那几十伏。被试事后普遍报告手心出汗、发抖、想退出却停不下来。

Milgram's 1963 experiment recruited 40 men aged 20 to 50 through a newspaper advert promising a memory and learning study at Yale. All 40 passed 300 volts and 26 reached 450, a result that still unsettles students who expect ordinary people to walk out early, and most participants reported trembling, sweating and wanting to stop while continuing anyway. The specification published by Pearson Edexcel (9PS0, 1.1.2) treats this study as the baseline against which every variation is judged.

三个变式 办公楼 电话 普通人下令

变式研究的价值在于把情境拆开看。Experiment 10 把实验室搬到 Bridgeport 一间破旧办公楼，服从率降到 47.5%，权威的合法性被地点削弱。Experiment 7 让实验者只通过电话下令，服从率掉到 20.5%，即时性一断，压力就松。Experiment 13 换成一名普通被试给指令，服从率约 20%，制服与头衔带来的强度消失。三项变式都指向同一件事，服从不是人格缺陷，而是情境变量在起作用。考场上写变式必须带实验编号与具体数字，只写情境变了会丢分。常见的错误是把 47.5% 记成 65%，那是基线不是办公楼。

The variation studies matter because they take the situation apart, and the specification names three of them for good reason. The rundown office block in Bridgeport produced 47.5% obedience, the telephone condition 20.5%, and orders from an ordinary participant around 20%, so removing the university building, the physical presence of the experimenter or the uniform all cut the figure. Many students lose the mark here by quoting 65% for every condition instead of the number attached to that experiment.

代理理论 agentic state 与道德压力

代理理论由 Milgram 在 1974 年提出，用两个状态解释服从。自主状态下人按自己的意愿行动，并为结果负责。代理状态下人把自己看成权威的工具，责任移交给上级。从自主到代理的转换伴随道德压力，也就是明知不对却继续执行的不适。缓冲因素让这种不适减弱，比如把学习者隔在墙后，看不见痛苦。约束因素把人留在代理状态，比如已经收了钱、已经按下第一下。这套解释的弱点是不解释 35% 的不服从者。阅卷时看到只写服从是因为权威，没有 state 与责任转移的机制，通常只给一到两分。

Agency theory, set out by Milgram in 1974, separates an autonomous state from an agentic state in which people act as the instrument of an authority and hand responsibility upwards. Moral strain describes the discomfort of carrying on while knowing the action is wrong, and buffers such as a wall between teacher and learner make that discomfort easier to ignore, while binding factors keep people in the agentic state once they have taken the money and pressed the first switch. The theory struggles to explain the minority who refused, which is the evaluation point examiners look for first.

社会影响理论 强度 即时性 数量

社会影响理论由 Latane 在 1981 年提出，把影响拆成三个变量。强度指权威的地位与权力，即时性指影响源在时间与空间上的接近，数量指影响源的人数。三者不是相加，而是相乘，任何一项接近零，整体影响就大幅削弱。理论还区分影响源与影响对象，多个对象会分摊影响，人数越多每多一人带来的增量越小。用它解释变式很顺，电话下令降低即时性，普通人下令降低强度。局限是概念偏抽象，很难量化验证，也说不清同一情境下个体的反应差异为什么那么大。

Social impact theory, proposed by Latane in 1981, treats influence as the product of strength, immediacy and number rather than their sum, so a value near zero on any one of them collapses the effect. It maps neatly onto Milgram's variations, because a telephone instruction cuts immediacy and an ordinary person giving orders cuts strength, and the law of diminishing returns explains why extra sources add less each time. The weakness is abstraction, since the three variables are hard to measure independently and the theory says little about why individuals differ so much in the same situation.

人格 性别 文化 如何改变服从率

个体差异这一条常被写得太笼统。Adorno 等人在 1950 年提出权威人格，用 F 量表测量，高分者对权威顺从、对弱者敌意。Milgram 自己也发现，不服从的被试更倾向质疑实验的伦理与目的。性别方面，Milgram 1963 年的样本全是男性，后续重复研究里男女差异并不稳定，写性别时不要下绝对结论。文化方面，Meeus 与 Raaijmakers 在 1986 年的荷兰实验里用心理压力替代电击，服从率依然很高，说明情境结构比国别更能预测行为。Edexcel 的 specification 1.1.3 要求把人格、性别、文化、情境四条都覆盖。很多学生以为文化差异一定显著，实证结果比想象中温和。

Individual differences are usually written too vaguely, yet the specification (9PS0, 1.1.3) asks for personality, gender, culture and situation as four separate factors. Adorno's 1950 work on the authoritarian personality used the F scale to link deference to authority with hostility towards the weak, while Milgram noted that disobedient participants questioned the purpose and ethics of the procedure; gender differences across replications are unstable, and Meeus and Raaijmakers found high obedience in the Netherlands in 1986 using psychological stress instead of shocks. The safe exam line is that situation structure predicts behaviour better than nationality, which is why culture alone rarely explains a result.

社会认同理论 分类 认同 比较

社会认同理论由 Tajfel 与 Turner 在 1979 年与 1986 年系统提出，用三步解释偏见。社会分类把人分成我们与他们，社会认同把自我价值绑定在群体成员身份上，社会比较在有利维度上抬高内群体。Tajfel 在 1971 年的最简群体研究里，让 14 至 15 岁男生按画作偏好分组，分组与资源分配毫无关系，被试仍然给内群体更多。这说明偏见不需要真实冲突，分类本身就能生出偏向。理论的长处是把偏见从人格问题拉回群体过程。局限是它解释不了为什么某些群体被贬得更狠，也解释不了同群体内部的态度差异。

Social identity theory, developed by Tajfel and Turner in 1979 and 1986, runs through categorisation, identification and comparison: people sort themselves into us and them, tie self-worth to group membership, and then favour the in-group on dimensions that matter. Tajfel's minimal group studies in 1971 gave 14 and 15 year old boys points to allocate after nothing more than a preference for paintings, and in-group favouritism appeared anyway, which is the strongest evidence that prejudice does not require real competition. What the theory handles less well is the size of hostility between particular groups and the variation in attitudes inside one group.

现实冲突理论 罗伯斯山洞的三个阶段

现实冲突理论来自 Sherif 在 1966 年的总结，实证基础是 1954 年与 1961 年发表的罗伯斯山洞实验。22 名 11 岁男孩被带到俄克拉何马州的营地，分成响尾蛇与老鹰两队。第一阶段各自形成群体规范与领袖，第二阶段用拔河、寻宝等零和竞赛制造敌意，第三阶段设置共同目标，比如修水箱、拉卡车，敌意下降。理论主张偏见源于对有限资源的竞争，资源可以是地位、土地或工作。评估时可以说它把偏见与社会结构连起来，也可以说营地情境人造、男孩背景相似、敌意被研究者诱导出来。

Realistic conflict theory, summarised by Sherif in 1966, rests on the Robbers Cave experiment reported in 1954 and 1961: 22 eleven year old boys were split into two groups, given separate norms and leaders, then set against each other in zero-sum tournaments before being pushed into superordinate goals such as repairing the water supply. Prejudice rose with competition for scarce resources and eased once both groups needed each other, which is the theory's central claim. Evaluation cuts both ways, since the camp was artificial, the boys came from similar backgrounds and the hostility was partly engineered by the researchers.

偏见的个体差异 情境与文化

同样的群体冲突，有人积极参与，有人保持距离，差别在个体与社会位置。Cohrs 等人在 2012 年用同伴报告数据检验意识形态态度与偏见的关系，右翼权威主义与社会支配倾向都能预测偏见，且不是自评偏差。情境方面，竞争、威胁感、群体边界清晰度都会推高偏见。文化方面，集体主义与个体主义社会对内群体偏好的表达方式不同，直接比较分数容易失真。写这一节容易堆结论，扣分点多在没有把因素与机制连起来。常见的错误是把权威人格当成偏见的唯一解释。Edexcel 官网的 specification 1.1.5 把人格、情境、文化三条并列，答题时三条都要落地。

Cohrs and colleagues in 2012 used peer report data to show that right-wing authoritarianism and social dominance orientation predict prejudice without relying on self-report alone, which strengthens the individual difference argument. Situation matters too, because competition, perceived threat and clear group boundaries all raise the temperature, and culture shapes how in-group preference is expressed rather than whether it exists at all. Answers that list these factors without linking each one to a mechanism tend to sit in the middle band.

经典研究与三项当代研究

Topic 1 的 studies 部分要求一项经典研究与一项当代研究。经典研究固定是 Sherif 等人的罗伯斯山洞，考试可能问程序、结果或评估。当代研究三选一，Burger 在 2009 年复刻 Milgram，70 名被试在 150 伏前停下，70% 越过该点，他还加了示范拒绝以降低伤害。Reicher 与 Haslam 在 2006 年做 BBC 监狱研究，15 名男性随机分组，暴政没有自动出现，共享认同才是关键。Cohrs 等人在 2012 年用同伴报告做偏见研究。考场上只写自己学的那一项，写三项不会加分。很多学生以为当代研究要背全部三篇，其实只需要一篇写透。

Topic 1 requires one classic study, Sherif's Robbers Cave, and one contemporary study chosen from three. Burger replicated Milgram in 2009 with 70 participants, stopping at 150 volts after a modelled refusal, and 70% of them went past that point; Reicher and Haslam ran the BBC prison study in 2006 with 15 men and found that tyranny was not automatic but depended on shared identity; Cohrs used peer reports in 2012 to test ideological attitudes and prejudice. Write only the study your centre taught, because three summaries earn no extra credit.

社会学取向的问卷实操与统计

Methods 部分围绕一次自编问卷展开。题目可以开放、封闭或排序量表，抽样有随机、分层、自愿、机会四种，报告里要写清选择理由。量化数据要算集中量数与离散量数，平均数、中位数、众数配极差与标准差，用条形图与频数表呈现。质性数据用主题分析，把回答编码归类再提炼主题。伦理按英国心理学会 2009 年的准则写，尊重、胜任、责任、诚信四条都要落到知情同意、保密与退出权。替代假设要写清是差异还是相关。常见的错误是只报平均数不报离散程度，阅卷时这类答案在方法题上很容易停在低档。

The methods section is built around the practical questionnaire, so the report has to justify open, closed or ranked items and one of the four sampling techniques: random, stratified, volunteer or opportunity. Quantitative analysis means central tendency alongside dispersion, with range and standard deviation plus a bar chart and frequency table, while qualitative answers go through coding and thematic analysis. Ethics follow the British Psychological Society code of 2009, covering respect, competence, responsibility and integrity through consent, confidentiality and the right to withdraw, and the alternate hypothesis must state whether a difference or a relationship is predicted.

试卷结构与 2 小时 15 分的时间切法

Paper 1 的试卷代码是 9PS0/01，Section A 就是社会心理学。Pearson Edexcel 在 2026 年 3 月发布的 Issue 4 把 Paper 1 与 Paper 2 的考试时间从 2 小时改为 2 小时 15 分钟，这一点要按新版准备。全卷 90 分，Section A 到 Section D 共 70 分，Section E 的 issues and debates 占 20 分。计算器允许使用，公式与统计表在附录 4 里给出。时间可以按分值切，70 分的选择与短答大约 95 分钟，两道 10 分论述留 40 分钟。AO1 考知识，AO2 考应用，AO3 考评估，8 分与 12 分题里 AO3 至少要占一半。很多学生把时间全花在前面的短答，最后一道 10 分题只写了两行。

Paper 1 is 9PS0/01 and Section A is social psychology. Pearson Edexcel Issue 4, published in March 2026, extended Papers 1 and 2 from two hours to two hours and fifteen minutes, so revision plans built on the older timing are out of date. The paper carries 90 marks, with Sections A to D totalling 70 and Section E on issues and debates worth 20, calculators are permitted and the formulae and statistical tables sit in Appendix 4. Answers worth 8 or 12 marks need evaluation to carry at least half the credit, and running out of time on the final extended response is one of the most common ways to drop a grade.

常见错误与扣分点

最常见的几类失手都能提前防。把代理理论与社会影响理论混着写，两个理论的机制完全不同，一个是责任转移，一个是强度与即时性。变式的百分比记错，47.5、20.5、20 三个数字要对应到实验编号。把罗伯斯山洞与最简群体研究对调，一个讲竞争，一个讲分类。评估只写有局限，不写局限在哪一步导致什么后果。另一类问题出在术语，obedience、conformity、dissent 三个词在答案里要前后一致。错得最多的是把不服从写成反抗权威的人格表现，其实情境变量已经能解释大部分差异。写完最后一段再回头看题干，通常能捡回两三分。

Four mistakes account for most lost marks, and all of them are avoidable. Agency theory and social impact theory get merged even though one moves responsibility and the other multiplies strength, immediacy and number; the variation percentages drift away from their experiment numbers; Robbers Cave and the minimal group studies are swapped even though one is about competition and the other about categorisation. Evaluation that says a study has limitations without naming the step that produces the consequence rarely reaches the top band, and inconsistent use of obedience, conformity and dissent costs precision marks.

牛津高阶英汉双解词典 第10版 商务印书馆¥168📱 去京东购买京东精选

需要一对一辅导？Need One-on-One Tutoring?

A-Level、GCSE、IB 与 KS2-KS3 全科辅导，社会心理学这类需要逐题批改与评估训练的科目尤其适合小班。

更多咨询请联系16621398022（同微信）

Visit aleveler.com