AQA Psychology Memory 汇总

## AQA心理学记忆MSM工作记忆LTM干扰遗忘考点 A-Level AQA Psychology can feel like navigating a maze of competing theories, but the Memory topic is one of the most rewarding to master. It sits squarely in Paper 1 and draws on classic cognitive psychology experiments that are as fascinating today as they were decades ago. If you are aiming for an A or A*, you need more than rote memorisation of Atkinson and Shiffrin. You need to understand how the multi-store model works, why Baddeley and Hitch challenged it, what Tulving taught us about long-term memory, and how interference and retrieval failure explain everyday forgetting. This article covers all of that in a structured, bilingual format that helps you internalise the material in both English and Chinese. A-Level AQA心理学看起来像是充满竞争理论的迷宫，但记忆这一主题是整门课最值得深挖的内容之一。它稳稳落在Paper 1，涉及大量经典认知实验，今天读来依然精彩。如果你冲A或A*，光死记Atkinson和Shiffrin是不够的。你需要理解多存储模型如何运转，Baddeley和Hitch为何挑战它，Tulving教给我们关于长时记忆的什么，以及干扰理论和检索失败如何解释日常遗忘。本文以中英双语系统梳理这些内容，帮助你在两种语言中内化知识。 --- ### 1. The Multi-Store Model of Memory (Atkinson & Shiffrin, 1968) The multi-store model, or MSM, is the starting point for all AQA Memory essays. Atkinson and Shiffrin proposed that memory consists of three distinct stores: the sensory register, short-term memory (STM), and long-term memory (LTM). Information flows linearly through these stores. Environmental stimuli enter the sensory register, where they are held for a fraction of a second. If attention is paid, the information passes to STM. Through maintenance rehearsal (repeating the information), it can then transfer to LTM, where it is stored permanently. 多存储模型（MSM）是所有AQA记忆论文的起点。Atkinson和Shiffrin提出记忆由三个独立存储器构成：感觉登记器、短时记忆（STM）和长时记忆（LTM）。信息在这些存储器中线性流动。环境刺激先进入感觉登记器，停留不到一秒。如果给予注意，信息进入STM。通过维持性复述（不断重复），信息进而转入LTM，永久存储。 **Coding, Capacity, and Duration** Research has established the key features of each store. For the sensory register, coding is modality-specific (iconic for visual, echoic for auditory), capacity is very large, and duration is around 250 milliseconds. For STM, Baddeley (1966) showed that coding is primarily acoustic — participants made more errors with acoustically similar words like "cat, cap, cab" than with semantically similar ones. Jacobs (1887) used the digit span technique and found STM capacity to be around 7±2 items. Peterson and Peterson (1959) demonstrated that STM duration is about 18 to 30 seconds without rehearsal, using a trigram retention task with an interference task (counting backwards in threes). 对于感觉登记器，编码是模态特异性的（视觉为图像性，听觉为回声性），容量极大，持续时间约250毫秒。对于STM，Baddeley（1966）证明编码主要是听觉性的——被试对"cat, cap, cab"这类发音相似的词比语义相似的词犯错更多。Jacobs（1887）使用数字广度技术发现STM容量约为7±2个项目。Peterson和Peterson（1959）通过三辅音字母回忆任务配合干扰任务（倒数三个数），证明STM持续时间约为18到30秒（无复述时）。 For LTM, Baddeley showed coding is primarily semantic — errors were higher for words with similar meanings. Capacity is potentially unlimited. Bahrick et al. (1975) found that recognition of high school classmates' faces was about 90% accurate after 15 years and still 70% after 48 years, suggesting LTM can last a lifetime. 对于LTM，Baddeley证明编码主要是语义性的——意义相似的词错误率更高。容量可能是无限的。Bahrick等人（1975）发现高中同学面孔的再认正确率在15年后约90%，48年后仍达70%，说明LTM可以持续一生。 **Critical Evaluation** The MSM has strengths. It was the first influential model of memory and stimulated decades of research. Evidence from brain scans supports the idea of separate stores: the hippocampus is active during LTM tasks while the prefrontal cortex is active during STM tasks (Beardsley, 1997). The case study of HM (Scoville & Milner, 1957), whose hippocampus was removed to treat epilepsy, showed that he could no longer form new LTM but his STM remained intact — strong evidence for separate stores. MSM有其优势。它是第一个有影响力的记忆模型，激发了几十年的研究。脑扫描证据支持独立存储器的观点：海马体在LTM任务中活跃，而前额叶皮层在STM任务中活跃（Beardsley, 1997）。HM的案例研究（Scoville & Milner, 1957）——为治疗癫痫摘除海马体后，他无法形成新的LTM但STM完好——为独立存储器提供了有力证据。 However, the MSM oversimplifies memory. It portrays STM and LTM as unitary stores, which is contradicted by evidence. The case study of KF (Shallice & Warrington, 1970) showed that his STM for verbal material was severely impaired after a motorcycle accident, but his STM for visual material was intact — suggesting STM is not a single store. Similarly, Tulving's research on episodic and semantic memory demonstrated that LTM is not unitary either. The MSM also overemphasises the role of rehearsal. In everyday life, we rarely rehearse information to remember it — we remember events because we processed them meaningfully (Craik & Lockhart's levels of processing). 但MSM将记忆简化了。它将STM和LTM描绘为单一存储器，这与证据矛盾。KF的案例（Shallice & Warrington, 1970）显示他在摩托车事故后对言语材料的STM严重受损，但对视觉材料的STM完好——说明STM不是单一存储器。同样，Tulving关于情景记忆和语义记忆的研究也证明LTM不是单一的。MSM还过度强调复述的作用。日常生活中，我们很少通过复述来记忆——我们记住事件是因为我们对其进行了意义加工（Craik & Lockhart的加工层次理论）。 --- ### 2. The Working Memory Model (Baddeley & Hitch, 1974) Baddeley and Hitch developed the working memory model, or WMM, to address the limitations of the MSM's concept of a unitary STM. The WMM focuses on short-term memory as an active processor rather than a passive store. It proposes four components: the central executive, the phonological loop, the visuo-spatial sketchpad, and the episodic buffer (added by Baddeley in 2000). Baddeley和Hitch发展了工作记忆模型（WMM），以解决MSM将STM视为单一存储器的局限。WMM将短时记忆视为一个主动处理器而非被动储存器。它提出四个成分：中央执行器、语音环、视空间画板和情景缓冲器（Baddeley于2000年添加）。 **The Central Executive** The central executive is the supervisory component. It allocates attention to the slave systems, decides which information is attended to, and coordinates cognitive tasks. It has limited processing capacity and does not store information itself. The central executive is the least understood component — Baddeley (2003) acknowledged that it is more of an "attentional system" and that its exact functions remain somewhat vague. This vagueness is a limitation of the model, as it is difficult to test empirically. 中央执行器是监督性成分。它将注意力分配给辅助系统，决定处理哪些信息，并协调认知任务。它的处理容量有限，本身不存储信息。中央执行器是最不被理解的部分——Baddeley（2003）承认它更像一个"注意力系统"，其精确功能仍有些模糊。这种模糊性是模型的一个局限，因为它难以进行实证检验。 **The Phonological Loop** The phonological loop deals with auditory and verbal information. It has two subcomponents: the phonological store, which holds words we hear (the "inner ear"), and the articulatory control process, which allows for subvocal repetition (the "inner voice"). Evidence for the phonological loop comes from the word length effect: Baddeley et al. (1975) found that participants recalled shorter words better than longer words, because shorter words can be rehearsed more quickly in the articulatory loop before they decay. 语音环处理听觉和言语信息。它有两个子成分：语音存储（"内心之耳"），保存我们听到的词语；以及发音控制过程（"内心之声"），允许无声复述。语音环的证据来自词长效应：Baddeley等人（1975）发现被试对短词的回忆优于长词，因为短词在发音环中可以在消退前更快复述。 **The Visuo-Spatial Sketchpad** The visuo-spatial sketchpad processes visual and spatial information. It is subdivided into the visual cache (stores visual data about form and colour) and the inner scribe (records the arrangement of objects in the visual field, handles spatial relationships). Logie (1995) proposed this subdivision. Dual-task studies provide strong evidence: participants struggle to perform two visual tasks simultaneously (e.g., tracking a moving light while describing the layout of a building) but can perform a visual task alongside a verbal task — suggesting separate processing systems. 视空间画板处理视觉和空间信息。它细分为视觉缓存（存储关于形状和颜色的视觉数据）和内书写器（记录物体在视野中的排列，处理空间关系）。Logie（1995）提出了这一细分。双任务研究提供了有力证据：被试难以同时完成两个视觉任务（如追踪移动光点同时描述建筑布局），但可以同时完成视觉任务和言语任务——说明存在独立的处理系统。 **The Episodic Buffer** Baddeley added the episodic buffer in 2000 as the component that integrates information from the phonological loop, the visuo-spatial sketchpad, and long-term memory into a single, coherent episode. It acts as a temporary store with limited capacity, linking information across domains to form integrated units of visual, spatial, and verbal information with time sequencing — essentially creating our sense of an "episode" or event. Baddeley于2000年添加了情景缓冲器，作为整合语音环、视空间画板和长时记忆信息为一个连贯情景的成分。它扮演一个容量有限的临时存储器，将跨领域信息连接起来形成视觉、空间和言语信息与时间顺序的整合单元——本质上创造了我们的"情景"或事件感。 **Critical Evaluation** The WMM has significant strengths. The dual-task paradigm provides solid experimental support. Baddeley and Hitch (1976) demonstrated that participants could perform a verbal reasoning task while simultaneously repeating "the, the, the" (articulatory suppression), with only a small decrease in speed, but errors did not increase — suggesting the phonological loop and central executive are separate components. The case of KF, mentioned earlier, further supports the WMM: his brain damage selectively impaired his phonological loop while leaving his visuo-spatial sketchpad intact. Brain scan studies by Braver et al. (1997) found increased activity in the left prefrontal cortex during verbal working memory tasks and right prefrontal cortex during spatial tasks, supporting the separation of the phonological loop and visuo-spatial sketchpad. WMM有显著优势。双任务范式提供了坚实的实验支持。Baddeley和Hitch（1976）证明被试可以在进行言语推理任务的同时不断重复"the, the, the"（发音抑制），速度只略微下降但错误未增加——说明语音环和中央执行器是独立成分。之前提到的KF案例进一步支持WMM：他的脑损伤选择性损害了语音环而视空间画板完好。Braver等人（1997）的脑扫描研究发现言语工作记忆任务中左前额叶皮层活动增加，空间任务中右前额叶皮层活动增加，支持语音环和视空间画板的分离。 However, the model also has limitations. The central executive remains poorly specified. As Baddeley himself noted, it is a "homunculus" — a little person inside the head making decisions, which is not a satisfactory scientific explanation. The model also does not explain how working memory links to long-term memory in detail (though the episodic buffer partially addresses this). Furthermore, most evidence comes from laboratory experiments that lack ecological validity — in real life, we rarely perform dual tasks like those in the lab. 但模型也有局限。中央执行器仍不明确。Baddeley自己也承认它是一个"小人"——一个在脑中做决定的小人，这不是令人满意的科学解释。模型也未能详细解释工作记忆如何连接长时记忆（尽管情景缓冲器部分解决了这一点）。此外，大多数证据来自实验室实验，缺乏生态效度——现实生活中我们很少像实验中那样做双任务。 --- ### 3. Types of Long-Term Memory (Tulving, 1972) Endel Tulving revolutionised our understanding of LTM by proposing that it is not a single store but rather comprises distinct types. The key distinction is between episodic memory, semantic memory, and procedur

广告 / Sponsored

al memory. Episodic memory refers to the memory of personal experiences and specific events in time — your recollection of your first day at secondary school, or what you had for breakfast yesterday. These memories are time-stamped and require conscious effort to recall. Semantic memory is the store of general knowledge about the world — facts, concepts, and meanings that are not tied to a specific time or place. Knowing that Paris is the capital of France or that water freezes at 0°C relies on semantic memory. Procedural memory is the memory of how to do things — motor skills and actions like riding a bicycle, typing on a keyboard, or playing a musical instrument. These memories are typically implicit and do not require conscious recall. Endel Tulving革命了我们对LTM的理解，他提出这不是单一存储器而是由不同类型组成。关键区分是情景记忆、语义记忆和程序性记忆。情景记忆指对个人经历和特定时间事件的记忆——你高中入学第一天的回忆，或昨天早餐吃了什么。这些记忆有时间标记，需要意识努力才能提取。语义记忆是关于世界的一般知识——事实、概念和含义，不与特定时间地点绑定。知道巴黎是法国首都或水在0°C结冰依赖语义记忆。程序性记忆是如何做事的记忆——运动技能和动作，如骑自行车、打字或演奏乐器。这些记忆通常是内隐的，不需要有意识地提取。 **Evidence for Distinct Types** The strongest evidence for Tulving's distinction comes from clinical case studies of amnesia. HM, mentioned earlier, could not form new episodic memories after his surgery — he could not recall what he did an hour ago. However, his semantic memory remained relatively intact for memories acquired before the surgery, and he could still learn new procedural skills (like mirror drawing), even though he had no conscious recollection of having done the task before. This dissociation strongly suggests that episodic, semantic, and procedural memories involve different brain systems. 支持Tulving区分的最强证据来自临床失忆症案例研究。前面提到的HM在手术后无法形成新的情景记忆——他不能回忆一小时前做过什么。然而他的语义记忆（对术前获得的知识）相对完好，而且仍能学习新的程序性技能（如镜像绘图），即使他对做过这项任务毫无有意识的记忆。这种分离强烈表明情景、语义和程序性记忆涉及不同脑系统。 Brain scanning evidence further supports this. Tulving et al. (1994) used PET scans and found that episodic memory retrieval is associated with increased activity in the prefrontal cortex, while semantic memory retrieval is associated with increased activity in the posterior regions of the cortex. Procedural memory has been linked to the cerebellum and basal ganglia — brain structures distinct from those involved in episodic and semantic memory. 脑扫描进一步支持这一点。Tulving等人（1994）使用PET扫描发现情景记忆提取与前额叶皮层活动增加相关，而语义记忆提取与大脑皮层后部区域活动增加相关。程序性记忆与 cerebellum 和基底节相关——这些脑结构与情景和语义记忆所涉及的结构不同。 --- ### 4. Explanations for Forgetting Forgetting is not a single process. AQA requires you to understand two main explanations: interference theory and retrieval failure due to absence of cues. 遗忘不是单一过程。AQA要求你理解两种主要解释：干扰理论和由于线索缺失导致的检索失败。 **Interference Theory** Interference theory proposes that forgetting occurs because one memory disrupts or blocks another. This happens in two directions. Proactive interference (PI) is when old learning interferes with new learning — for example, when you keep calling your new teacher by your old teacher's name because the old teacher's name is firmly established. Retroactive interference (RI) is when new learning disrupts the recall of old memories — for example, when learning a new locker combination makes you forget the old one. 干扰理论提出遗忘发生是因为一段记忆干扰或阻塞了另一段记忆。这发生在两个方向。前摄干扰（PI）是旧学习干扰新学习——例如，你总是用旧老师的名字称呼新老师，因为旧老师的名字根深蒂固。后摄干扰（RI）是新学习干扰旧记忆的提取——例如，学会了新的储物柜密码让你忘记了旧的。 McGeoch and McDonald (1931) provided early experimental evidence. They had participants learn a list of 10 adjectives until 100% recall, then learn a second list. When the second list contained synonyms of the first list, recall of the first list dropped to just 12%. When the second list was unrelated (nonsense syllables), recall was 26%. The more similar the interfering material, the greater the forgetting — supporting the idea that interference causes forgetting. McGeoch和McDonald（1931）提供了早期实验证据。他们让被试学习10个形容词的列表直到100%回忆，然后学习第二个列表。当第二个列表包含第一个列表的同义词时，第一个列表的回忆率降至仅12%。当第二个列表无关（无意义音节）时，回忆率为26%。干扰材料越相似，遗忘越多——支持了干扰导致遗忘的观点。 Underwood (1957) reviewed multiple studies and found that participants who had learned 10 or more previous lists showed only 20% recall of the latest list after 24 hours, while those who had learned no previous lists showed about 70% recall. This is powerful evidence for proactive interference. Underwood（1957）回顾多项研究发现，之前学过10个以上列表的被试在24小时后对最新列表的回忆仅20%，而之前未学任何列表的被试回忆率约70%。这是前摄干扰的有力证据。 A criticism of interference research is that most studies use artificial laboratory tasks — learning lists of words or nonsense syllables. This lacks mundane realism. In everyday life, we rarely need to remember lists of similar words. However, Baddeley and Hitch (1977) addressed this by studying rugby players. They asked players to recall the names of teams they had played against during the season. The number of intervening games (which would cause interference) was a better predictor of forgetting than the passage of time — a finding that supports interference in a real-world context. 对干扰研究的一个批评是大多数研究使用人工实验室任务——学习单词或无意义音节列表。这缺乏日常真实性。现实生活中我们很少需要记忆相似单词的列表。但Baddeley和Hitch（1977）通过研究橄榄球运动员解决了这个问题。他们让运动员回忆本赛季对抗过的球队名称。中间比赛的场次数（会产生干扰）比时间流逝更好地预测了遗忘——这一发现支持了现实世界中的干扰理论。 **Retrieval Failure** The retrieval failure explanation takes a different approach. It suggests that memories are stored in LTM but are temporarily inaccessible because the retrieval cues present at the time of recall do not match those present at the time of encoding. This is known as the encoding specificity principle, proposed by Tulving and Thomson (1973). 检索失败解释采取了不同角度。它认为记忆存储在LTM中但由于回忆时存在的检索线索与编码时的线索不匹配而暂时无法获取。这被称为编码特异性原则，由Tulving和Thomson（1973）提出。 Godden and Baddeley (1975) provided a striking demonstration. They had divers learn a list of 40 words either on land or underwater, and then recall them either in the same or a different environment. Recall was 40% higher when the learning and recall environments matched. This shows the power of context-dependent forgetting. Similarly, Abernethy (1940) found that students tested by their regular teacher in their regular classroom performed better than students tested in an unfamiliar room by an unfamiliar instructor — supporting context-dependent retrieval. Godden和Baddeley（1975）提供了一个惊人的演示。他们让潜水员在陆地或水下学习40个单词，然后在相同或不同环境中回忆。当学习和回忆环境匹配时，回忆率高出40%。这显示了情境依赖遗忘的力量。同样，Abernethy（1940）发现由日常老师在日常教室中测试的学生比在不熟悉教室由不熟悉老师测试的学生表现更好——支持了情境依赖检索。 State-dependent forgetting is another form: Goodwin et al. (1969) found that participants who learned information while drunk recalled more when drunk than when sober — and vice versa. This suggests that internal physiological states also serve as retrieval cues. 状态依赖遗忘是另一种形式：Goodwin等人（1969）发现醉酒学习信息的被试在醉酒时比清醒时回忆更多——反之亦然。这表明内部生理状态也可作为检索线索。 A limitation of retrieval failure research is that it struggles to explain all forms of forgetting. It is difficult to apply to procedural memories or semantic memories that are not tied to specific contexts. Furthermore, the encoding specificity principle is somewhat circular: it essentially says that forgetting occurs because cues are absent, and we know cues were absent because forgetting occurred. 检索失败研究的一个局限是难以解释所有形式的遗忘。它很难应用于程序性记忆或不受特定情境约束的语义记忆。此外，编码特异性原则在某种程度上是循环的：它本质上是说遗忘发生是因为线索缺失，而我们知道线索缺失是因为遗忘发生了。 --- ### 5. Exam Tips for AQA Psychology Memory When you sit Paper 1, the Memory section is typically worth 24 marks — made up of multiple-choice, short-answer, and a 16-mark essay. Here are five actionable strategies to maximise your score. 当你坐在Paper 1考场，记忆部分通常值24分——由选择题、简答题和16分论文题组成。以下是五个可操作的策略帮助你最大化分数。 **Learn your AO3. Evaluations are the differentiator between a B and an A*. For every model (MSM, WMM) and every explanation (interference, retrieval failure), know at least two strengths and two limitations that use named researchers.** Generic statements like "the model is too simplistic" will not earn marks without supporting evidence. Always link your evaluation point to specific research: "This is supported by Shallice and Warrington's case study of KF, who...". 学好AO3。评价是B和A*的分水岭。对于每个模型（MSM, WMM）和每种解释（干扰、检索失败），需要知道至少两个用具体研究者命名的优点和两个局限。"模型太简化"这类笼统表述没有证据支撑不会得分。始终将评价点与具体研究联系起来，比如"这一点得到Shallice和Warrington关于KF案例研究的支持，KF......"。 **Use the PEEL structure for 16-mark essays: Point, Evidence, Explanation, Link.** Start with a clear topic sentence. Follow with specific research evidence (named researcher, year, brief method, key finding). Then explain what the evidence shows. Finally, link back to the question. For the 16-marker, aim for five to six well-developed paragraphs. 16分论文题使用PEEL结构：观点、证据、解释、回链。以清晰的主题句开篇。跟进具体研究证据（具名研究者、年份、简要方法、关键发现）。然后解释证据说明了什么。最后回链到问题。16分题目标五到六个充分展开的段落。 **Make comparisons between models explicit.** The MSM and WMM are often asked together. Know that the WMM was developed as an improvement on the MSM. The MSM sees STM as a passive, unitary store; the WMM sees it as an active, multi-component processor. These comparisons earn you high-level marks. 将模型之间的比较明确化。MSM和WMM经常一起考。知道WMM是作为MSM的改进而发展的。MSM将STM视为被动的单一存储器；WMM将其视为主动的多成分处理器。这些比较能帮你拿到高层级分数。 **Master the key researchers and dates.** Your marker will scan your essay for names. Atkinson and Shiffrin (1968), Baddeley and Hitch (1974), Baddeley (1966), Peterson and Peterson (1959), Tulving (1972), Godden and Baddeley (1975), McGeoch and McDonald (1931), Underwood (1957). If these names are absent, your essay cannot reach the top band. 掌握关键研究者和日期。阅卷人会扫描你论文中的名字。Atkinson和Shiffrin (1968)、Baddeley和Hitch (1974)、Baddeley (1966)、Peterson和Peterson (1959)、Tulving (1972)、Godden和Baddeley (1975)、McGeoch和McDonald (1931)、Underwood (1957)。如果这些名字缺失，你的论文就无法进入最高评分段。 **Time management: 16-mark essay in 20 minutes maximum.** Spend 2 minutes planning, 16 minutes writing, and 2 minutes checking. A planned essay has a clearer argument and scores higher than an unplanned, rambling one. Jot down your PEEL framework before you start writing. 时间管理：16分论文题最多20分钟。花2分钟规划，16分钟写作，2分钟检查。有规划的论文论证更清晰，得分高于无规划的散乱写作。动笔前先草拟PEEL框架。

🛍️ 微信小店好物推荐

fx-991CN X 中文版计算器-老王推荐

¥168.00 🛒 微信小店购买

新东方 乱序版SAT词汇词根+联想记忆法 新版 俞敏洪着 SAT绿宝书

¥51.00 🛒 微信小店购买

🎓 需要A-Level Psychology一对一辅导？从Paper 1到Paper 3，针对性攻克薄弱环节。添加微信获取课程方案和试听课安排。

📱 16621398022 同微信

更多咨询请联系16621398022（同微信）