AQA Psychology A-level: Memory
Part 1 — Coding, Capacity and Duration of Memory
Coding: Refers to the format or 'type' of information stored in each memory store.
Short-term Memory (STM): Acoustic coding (Baddeley, 1966).
Long-term Memory (LTM): Semantic coding (Baddeley, 1966).
Experimental Findings:
More mistakes are made with acoustically-similar words right after learning.
More mistakes with semantically-similar words after 20 minutes.
Capacity: Volume of information a memory store can hold.
STM Capacity: Approximately 7 +/- 2 items (Miller).
LTM Capacity: Considered unlimited.
Miller’s Theory: People tend to remember groups of 7 (e.g., 7 days in a week); chunking helps recall information.
Jacobs' Findings: Mean letter span = 7.3; mean digit span = 9.3.
Duration: Time information can be stored in memory.
STM Duration: 18-30 seconds (Petersen et al, 1959).
Study method: Participants recall consonant syllables after counting down from a 3-digit number (prevents mental rehearsal).
LTM Duration: Unlimited (Bahrick et al, 1975).
Findings: 90% photo recognition of classmates drops to 70% between 15-46 years post-graduation.
Critique of Historical Research:
Jacobs lacked standardization; confounding variables (noise, difficulty of words) may have influenced results.
Strength of Bahrick's Study: Use of meaningful stimuli increases ecological validity and generalize to real-life learning.
Critiques of Petersen and Miller:
Low mundane realism due to artificial stimuli impacting generalizability of findings.
Recent findings suggest STM capacity may be closer to 4 chunks, rather than Miller's estimate of 5-9.
Part 2 — The Multi-Store Model of Memory (MSM)
Overview: MSM illustrates how memory is stored, transferred, retrieved, and forgotten.
Memory Stores:
Sensory Register:
Contains sub-stores for each of the 5 senses (e.g., echoic store for auditory info).
Capacity: Huge. Duration: Less than half a second.
Transfers to STM requires attention.
Short-term Memory (STM):
Encoding: Acoustic (Baddeley).
Capacity: 7 +/- 2 items (Miller).
Duration: 18-30 seconds (Petersen).
Maintenance Rehearsal: Repeating information retains it in STM; lack of rehearsal leads to forgetting.
Long-term Memory (LTM):
Encoding: Semantic.
Capacity: Unlimited.
Duration: Long-term (over 46 years, Bahrick et al).
Retrieval Process: Moving info back to STM requires retrieval; after retrieval, it undergoes maintenance rehearsal.
Types of Long-term Memory:
As categorized by Tulving: procedural, semantic, and episodic.
MSM views LTM as a single store, failing to represent different LTM types and their recall processes (conscious vs. unconscious).
Craik and Watkins (1973) Critique: Emphasizes that the type of rehearsal (elaborative vs. maintenance) is vital for transferring memories to LTM, contrary to MSM's suggestion of quantity.
Strengths of MSM:
Acknowledges qualitative differences between STM and LTM (e.g., coding and duration differences).
Limitations:
Shallice and Warrington's study of amnesiac patient KF showed that STM isn't a unitary store; KF's recall varied by stimulus type, suggesting multiple STM types.
Part 3 — Types of Long-Term Memory
Types of Long-term Memory:
Episodic Memory: Personal memories of events with associated details (e.g., weddings, first meetings).
Semantic Memory: Knowledge about the world, concepts, words (e.g., meanings of words).
Procedural Memory: Memories of learned skills (e.g., riding a bike, playing piano).
Recall Differences:
Episodic and semantic require conscious recall; procedural is typically unconscious.
Neurological Basis:
Petersen et al. found episodic memories tied to the right prefrontal cortex, while semantic memories are tied to the left.
Practical Applications:
Understanding of different LTM types can improve treatment for cognitive impairments affecting episodic memory (Belleville et al).
Distinct Memory Types:
Cohen and Squire's distinction: Declarative (episodic and semantic, conscious recall) vs. Non-declarative (procedural, unconscious recall).
Case Studies:
HM and Clive Wearing: Both struggled with episodic memory but retained procedural and semantic memory, highlighting brain area specialization.
Part 4 — The Working Memory Model (WMM)
WMM Overview: Proposes that STM consists of:
Central Executive: Limited capacity; allocates tasks to slave systems.
Phonological Loop: Processes auditory info; includes articulatory process and phonological store.
Visuo-spatial Sketchpad: Manages visual and spatial info; capacity approximately 4-5 chunks (Baddeley).
Episodic Buffer: Integrates data from other stores; connects STM to LTM.
Critique of Central Executive: Indefinite definition; lacks thorough explanation of its processes.
KF Study: Support for WMM; KF’s visual STM was better than auditory, indicating separate processes.
Dual-task Performance Studies: Participants performing visual and verbal tasks simultaneously show decreased performance, supporting capacity limitations of the central executive.
Neuroscanning Evidence:
Braver et al. linked cognitive load with prefrontal cortex activation, supporting WMM’s premise of central executive allocation.
Part 5 — Explanations for Forgetting: Interference
Interference: Recall of one memory inhibits another, leading to forgetting.
Types:
Proactive Interference: Old memories inhibit new ones.
Retroactive Interference: New memories inhibit old ones.
Retroactive Interference Example: McGeoch and McDonald (1931) showed learning synonyms led to reduced recall of original list compared to control.
Findings:
Interference is more prevalent with similar materials, supported by research showing commonalities in memory types.
Critique:
Artificial stimuli in studies (e.g., word lists) lead to low ecological validity; real-life materials are more meaningful.
Validity:
Lab experiments yield high validity through control of confounding variables.
Baddeley and Hitch Study: Rugby players’ recall revealed newer games interfered with older, supporting retroactive interference concept.
Methodological Critique: Short time spans in studies fail to reflect real-life memory recall conditions, challenging the validity of interference as a reason for forgetting.
Part 6 — Explanations for Forgetting: Retrieval Failure
Retrieval Failure: Forgetting occurs when cues at encoding are absent during recall (Tulving’s Encoding Specificity Principle).
Types of Cues: Context-dependent and state-dependent.
Context-Dependent Forgetting: Godden and Baddeley (1975) underwater study found matching conditions led to significantly higher recall.
State-Dependent Forgetting: Carter and Cassaday (1998) demonstrated through anti-histamine-induced state changes that matching internal cues led to better recall.
Critique of Retrieval Failure:
Low ecological validity due to extreme conditions (e.g., underwater) suggesting retrieval failure may not apply to everyday forgetting.
Godden and Baddeley (1975) Follow-up: They found no significant difference in recognition tasks between matching and non-matching conditions, implying limitation of retrieval failure explanation.
Cyclical Reasoning in ESP: Overreliance on assumptions regarding cue significance may not always lead to retrieval failures, raising methodological concerns.
Part 7 — Factors Affecting the Accuracy of Eyewitness Testimony: Misleading Information
Eyewitness Testimony (EWT): Recollection of crime details; vulnerable to misleading information (e.g., leading questions).
Leading Questions Study: Loftus and Palmer (1974); different verbs prompted different speed estimates of cars, demonstrating how phrasing can manipulate perception.
Response Bias Explanation: Leading questions can skew answers without altering memory; however, they also support the substitution explanation when memory is genuinely altered.
Post-Event Discussions: Conversations following an event can introduce inaccuracies (Gabbert et al, 2003); 71% inaccuracies in discussion groups vs. 0% in control.
Memory Conformity: Co-witnesses may adopt incorrect details from others’ accounts.
Methodological Concerns:
Young targets lead to own-age bias; older participants may be misrepresented in accuracy studies (Anastasi and Rhodes).
Demand Characteristics: Participants may skew responses due to social desirability (Zaragosa and McCloskey).
Ecological Validity Concerns: Artificial stimuli diminish realism; anxiety from real-life situations influences EWT accuracy differently.
Part 8 — Factors Affecting the Accuracy of Eyewitness Testimony: Anxiety
Anxiety: Physiological response to stress; can negatively or positively impact EWT.
Negative Impact Study: Johnson and Scott (1976); high-anxiety scenario led to reduced accurate identification.
Positive Impact Study: Yuille and Cutshall (1986); real-life shooting witnesses retained high accuracy even with increased anxiety.
Yerkes-Dodson Law: Inverted U-shape relationship between arousal and performance, suggesting moderate anxiety yields best performance.
Surprise vs. Anxiety: Pickel’s study showed accuracy increases due to surprise, not just anxiety, complicating interpretations.
Ethical Issues: Exposure to distressing scenarios breaches participant rights; requires cost-benefit analysis of ethical costs vs. knowledge gained.
Field Study Biases: Real-life conditions introduce uncontrolled variables, undermining reliability.
Part 9 — Improving the Accuracy of Eyewitness Testimony: Cognitive Interviews
Cognitive Interview (CI): An approach designed to enhance EWT accuracy through specific techniques.
Stages of CI:
Report Everything: Encourages recall of all details, regardless of perceived importance.
Reinstate the Context: Helps reduce context-dependent forgetting by recalling conditions at the time of the event.
Change the Perspective: Minimizes schema biases by encouraging recounting from alternate viewpoints.
Reverse the Order: Prevents lying and influences perception by altering event chronology.
Enhanced CI: Fisher et al. (1987) focused on social interactions with interviewers to improve comfort and truthfulness in witness reporting.
Practical Value Concerns: CI may be time-consuming and require specialized training, limiting its implementation in police settings.
Counterproductive Increases: CI raised recall of incorrect details (Kohnken et al, 1999), suggesting its introduction could complicate accuracy.
Optimized CI: Milne and Bull (2002) suggested using effective components of CI, such as context reinstatement and reporting everything, could yield improved accuracy even with limited training.