The Multi-Store Model of Memory (AQA A-Level Psychology): Revision Notes
Coding, Capacity & Duration of Memory
Understanding how memory works requires examining three fundamental characteristics of different memory stores: how information is stored (coding), how much can be held (capacity), and how long it lasts (duration). These features help distinguish between short-term memory (STM) and long-term memory (LTM) within the multi-store model.
These three characteristics form the foundation for understanding how different memory stores operate and interact within the multi-store model of memory. Each characteristic reveals distinct differences between STM and LTM that have important implications for learning and recall.
Key definitions
Coding refers to the format in which information is stored within different memory stores. This determines how our brain processes and represents information internally.
Capacity represents the amount of information that can be held within a memory store at any given time. Different memory stores have vastly different storage limitations.
Duration measures the length of time that information can be retained within memory before it becomes unavailable for recall.
Coding in memory stores
Research has demonstrated that STM and LTM process information using different coding systems, which has important implications for how we learn and remember. STM primarily relies on acoustic coding (sound-based storage), while LTM uses semantic coding (meaning-based storage).
Baddeley's research on coding
Alan Baddeley conducted influential experiments in the 1960s to investigate how information is coded in different memory stores. His research involved presenting participants with four different types of word lists to memorise and recall.
Worked Example: Baddeley's Coding Experiment
Participants: Four groups received different word lists Procedure:
- Group 1: Acoustically similar words (e.g. cat, cab, can)
- Group 2: Acoustically dissimilar words (e.g. pit, few, cow)
- Group 3: Semantically similar words (e.g. great, large, big)
- Group 4: Semantically dissimilar words (e.g. good, huge, hot)
Participants attempted recall either immediately after presentation (testing STM) or after a 20-minute delay (testing LTM).
Key Findings:
- Immediate recall: Performance was significantly worse with acoustically similar words → STM uses acoustic coding
- Delayed recall: Performance was significantly worse with semantically similar words → LTM uses semantic coding
These results demonstrate that STM processes information based on how words sound, while LTM focuses on meaning and understanding.
Evaluation of coding research
The research provided clear evidence for different coding systems and used controlled experimental conditions with results that have been replicated in subsequent studies.
Critical Limitation: Artificial Stimuli
The word lists used lacked personal meaning for participants, limiting the generalisability of findings to real-world memory tasks. When processing meaningful information, people might use semantic coding even in STM tasks, suggesting the findings may not apply to everyday memory activities.
Capacity of memory stores
Research into memory capacity has revealed significant differences between STM and LTM, with important practical implications for learning and information processing. While STM has a strictly limited capacity, LTM appears to have unlimited storage potential.
Research evidence
Early research established fundamental principles about STM capacity that continue to influence our understanding of memory limitations.
Worked Example: Jacobs' Digit Span Research (1887)
Joseph Jacobs developed a technique to measure STM capacity by testing participants' ability to recall sequences of digits or letters in the correct order.
Procedure: Gradually increasing the length of sequences until participants could no longer recall them accurately.
Key Findings:
- Average span: 9.3 items for digits
- Average span: 7.3 items for letters
- Conclusion: STM has limited capacity
Miller's observations (1956): George Miller examined patterns in everyday life and noticed that many things are organised in groups of seven (musical notes, days of the week, deadly sins). This led to his famous conclusion that STM capacity is approximately items.
However, Miller also identified that people can increase their effective memory capacity through chunking - grouping individual items into meaningful units. For example, recalling the letters F-B-I-C-I-A-N-H-S as three chunks (FBI-CIA-NHS) rather than nine separate letters.
Evaluation of capacity research
The early research established important principles, but some limitations must be considered when interpreting the findings.
Methodological Concerns with Early Research
Jacobs' research was conducted in 1887 when psychological research methods lacked adequate controls. Participants may have been distracted during testing, potentially affecting the validity of results through uncontrolled confounding variables. However, subsequent research has confirmed these findings, supporting their validity.
Contemporary Challenge to Miller's Estimate
More recent research by Cowan (2001) reviewed multiple studies and concluded that STM capacity may be closer to four chunks rather than seven items, suggesting Miller may have overestimated capacity.
Duration of memory stores
Understanding how long information persists in different memory stores provides crucial insights into memory processes and the importance of rehearsal and consolidation. Research shows a stark contrast between the brief duration of STM and the potentially lifelong storage of LTM.
STM duration research
Worked Example: Peterson & Peterson Study (1959)
This landmark study measured precisely how long information remains in STM without rehearsal.
Participants: 24 undergraduate students Procedure:
- Each participant completed eight trials
- Received a consonant syllable (trigram like YCG) and a three-digit number
- Counted backwards from the number to prevent mental rehearsal
- Recall tested after retention intervals of 3, 6, 9, 12, 15, or 18 seconds
Key Finding: Memory performance declined rapidly as the retention interval increased, demonstrating STM duration of typically only seconds without active rehearsal.
LTM duration research
Worked Example: Bahrick et al. Study (1975)
This study investigated how long memories can persist in LTM using real-world memories over extended periods.
Participants: 392 American high school graduates aged 17-74 Procedure:
- Photo-recognition test: Identifying classmates from yearbook photos (50 photos total)
- Free recall test: Naming graduating classmates without prompts
Key Findings:
- Within 15 years: 90% accuracy in photo recognition
- After 48 years: 70% accuracy in recognition
- Free recall: 60% accuracy after 15 years, 30% after 48 years
- Recognition consistently outperformed free recall at all time intervals
This research demonstrates that LTM can indeed store information for decades, though some forgetting occurs over very long periods.
Evaluation of duration research
Both STM and LTM duration studies have contributed valuable insights, but each faces distinct methodological considerations.
STM Research Limitation: Ecological Validity
Peterson and Peterson's use of meaningless consonant syllables doesn't reflect typical memory activities where we try to remember meaningful information. This artificial approach may not generalise to real-world memory tasks, though it remains relevant for situations where we need to remember relatively meaningless information like phone numbers.
LTM Research Strength and Limitation
Strength: Bahrick's study examined genuine, personally meaningful memories, providing higher ecological validity and more applicable findings to everyday memory experiences.
Limitation: The real-world nature meant confounding variables couldn't be fully controlled. Participants might have looked at yearbook photos or encountered former classmates over the years, potentially rehearsing these memories and affecting results.
Key Points to Remember:
-
STM uses acoustic coding (sound-based), while LTM uses semantic coding (meaning-based), as demonstrated by Baddeley's research
-
STM capacity is limited to approximately items (though possibly closer to 4 chunks), but can be increased through chunking strategies
-
STM duration is very brief (seconds) without rehearsal, while LTM can store information for decades with gradual decline over very long periods
-
Recognition memory typically outperforms free recall, especially for long-term memories
-
Research validity varies - studies using artificial stimuli may have limited real-world application, while naturalistic studies provide better ecological validity but less experimental control