The different types of memory stores and their characteristics

Memory is a crucial cognitive process that involves retaining information about past experiences. Psychologists categorise memory into three main stores: the sensory register, short-term memory (STM), and long-term memory (LTM). Each store varies in terms of duration (how long information lasts), capacity (how much information can be held), and coding (the way information is stored to form a memory trace).

Characteristics of the sensory register

• Temporary storage - Holds information from the senses (sight, sound, touch, taste, smell) for a very brief period as it is constantly receiving data from the environment.
• Rapid decay - Information fades quickly through spontaneous decay unless attention is paid to it.
• Limited capacity and duration - Can only retain a small amount of data for a fleeting moment.
• Sensory-specific coding - Information is stored based on the sense that detected it, such as visual for images or auditory for sounds.

Characteristics of short-term memory (STM)

• Brief retention - STM has a limited duration, holding information for a short time unless actively rehearsed.
• Restricted capacity - Can only store a small amount of information at once.
• Acoustic coding - Information is primarily stored based on sound, often through internal repetition.

Characteristics of long-term memory (LTM)

• Extensive storage - LTM has a potentially unlimited capacity and can retain information indefinitely.
• Semantic coding - Information is mainly stored based on its meaning, though visual or acoustic coding can also occur.
• Types of LTM - LTM is divided into distinct categories based on the nature of the information stored:
  • Episodic memory - Stores personal experiences and events, including details of time, place, emotions, and specifics of what occurred. These are declarative memories that can be consciously recalled.
  • Semantic memory - Holds factual knowledge and learned information, such as historical dates or vocabulary, without reference to when or where it was learned. These are also declarative and consciously recalled.
  • Procedural memory - Contains knowledge of skills and how to perform tasks, like riding a bike or typing, which cannot be consciously recalled as facts but are demonstrated through action.

Studies on the duration of memory in sensory register, short-term memory, and long-term memory

Various research studies have explored how long information remains in each memory store, providing insights into their temporal limits under different conditions.

Sperling (1960) study on sensory register duration

• Method - Participants in a controlled lab experiment viewed a grid of three rows with four letters each for a mere 50 milliseconds. They were then asked to recall either the entire grid or a specific row, cued by a tone (high, medium, or low) immediately after the display.
• Results - When recalling the whole grid, participants averaged four to five letters. When cued for a specific row, they recalled about three letters regardless of the row, suggesting they initially held most of the grid in memory.
• Conclusions - The sensory register briefly holds a large amount of information, but the memory trace fades rapidly before full recall is possible.
• Evaluation - As a lab experiment, it offers high control over variables and is replicable, ensuring scientific rigour. However, the artificial task of recalling letters after a tone lacks ecological validity, meaning it may not reflect real-world memory use.

Peterson and Peterson (1959) study on STM duration

• Method - Participants were presented with nonsense trigrams (three random consonants) and asked to recall them after delays of 3, 6, 9, 12, 15, or 18 seconds. During the delay, they performed an interference task of counting backwards in threes from a starting number to prevent rehearsal.
• Results - Recall accuracy was around 80% after 3 seconds but dropped to about 10% after 18 seconds.
• Conclusions - Without rehearsal, information in STM fades within approximately 18 seconds, indicating a very limited duration.
• Evaluation:
  • The controlled lab setting ensures reliable results due to strict variable management.
  • However, the use of nonsense trigrams reduces ecological validity as they lack real-world relevance.
  • Meaningful information might persist longer in STM.
  • Repeated trials with multiple trigrams could cause confusion, potentially skewing results after the initial attempt.

Bahrick et al (1975) study on LTM duration

• Method - A field study involving 392 participants who were asked to list names of former classmates in a free-recall test. They also completed a photo-recognition test (recalling names from pictures) and a name-recognition test (matching names to photos).
• Results - Within 15 years of leaving school, participants recognised about 90% of names and faces, with 60% accuracy in free recall. After 30 years, free recall dropped to 30%, while name recognition remained at 80% and photo recognition at 40% after 48 years.
• Conclusions - LTM can store information for very long periods in real-life contexts, with recognition outperforming free recall, suggesting a vast information store that requires cues for full access.
• Evaluation:
  • The real-world setting enhances ecological validity, reflecting genuine memory use.
  • However, uncontrolled variables in a field study reduce reliability, as it's unclear why some information was recalled.
  • Better recall of meaningful data (classmates) compared to lab studies may be due to rehearsal through ongoing contact or discussions, limiting generalisation to other LTM content.

Research on the capacity of short-term memory

The capacity of STM, or how much information it can hold at one time, has been a key focus of memory research, revealing its constraints and potential for enhancement through specific strategies.

Jacobs (1887) study on STM capacity

• Method - Participants were shown sequences of letters or numbers, repeating them back in order. The sequence length increased until recall failed.
• Results - On average, participants recalled about 9 numbers and 7 letters, with capacity improving with age during childhood.
• Conclusions - STM has a limited capacity of roughly 5 to 9 items, with individual variations such as age impacting performance. Numbers may be easier to recall due to a smaller set of possible items (10 digits vs. 26 letters).
• Evaluation - The artificial nature of recalling random sequences lacks ecological validity, as it doesn't mimic everyday tasks. Meaningful information might reveal a greater STM capacity. Repeated trials could also cause confusion, affecting later attempts.

Miller (1956) findings on STM capacity and chunking

• Key observation - Research reviewed by Miller indicated that STM typically holds around seven items, often referred to as 'Miller's magic number' of seven plus or minus two.
• Chunking strategy - Capacity can be increased by grouping individual items into meaningful units or 'chunks'. For example, a sequence of digits can be remembered as meaningful years, reducing the number of separate items to recall.
• Enhanced capacity - STM can hold approximately seven chunks, significantly boosting the total amount of information retained.

Investigations into coding in short-term and long-term memory

Coding refers to the format in which information is stored in memory, differing between STM and LTM, and influencing how effectively data is retained and retrieved.

Nature of coding in memory stores

• STM coding - Primarily acoustic, based on sound, often maintained by repeating information internally to keep it active.
• LTM coding - Mainly semantic, focusing on the meaning of information, which aids long-term retention, though visual or acoustic coding can also occur.

Baddeley (1966) study on coding differences

• Method - Participants were divided into groups and given four word lists: acoustically similar (e.g., man, mad, map), acoustically dissimilar (e.g., pit, cow, bar), semantically similar (e.g., big, large, huge), or semantically dissimilar (e.g., good, hot, pig). They recalled the words either immediately (from STM) or after a delay (from LTM).
• Results - Immediate recall struggled with acoustically similar words, indicating confusion in STM. Delayed recall showed difficulty with semantically similar words, suggesting interference in LTM.
• Conclusions - STM relies heavily on acoustic coding, while LTM depends more on semantic coding, as shown by the patterns of recall errors.
• Evaluation - The artificial task of recalling word lists lacks ecological validity, not reflecting typical memory use. The study overlooks other LTM types (like episodic or procedural) and coding methods (like visual). Using an independent groups design means participant variables weren't controlled, potentially affecting results.