Chapter 3: Models of memory Flashcards
The multi-store memory model (Atkinson & Shiffrin)
Human memory consists of:
1. sensory memory
2. short-term memory storage
3. long-term memory storage
Information going from 1 –> 2 requires attention.
Rehearsal keeps stuff in 2., continuous rehearsal transfers it to 3.
Retrieval recalls information from 3 to 2
Sensory memory
Stuff you notice, information is not processed, consists of all the senses. Short duration of 2-5 sec. Potentially unlimited capacity
Short-term memory storage
can undergo primitive transformation, processing of words, sounds. Capacity of 5-9 chunks. Duration usually 30 seconds
Long-term memory storage
Place for storing large amounts of information for indefinite periods of time, capacity is not defined nor even approximated by researchers. Retrieval of memories might be what proves to be the challenge, not storage.
Sperling testing the multi-store memory model (1960)
A: testing the existence of iconic memory
M: repeated-measures design. “Partial-report technique”.
P: Participants were shown a tachistoscopic image of a grid with alphanumerics flashed up for 50ms.
Condition one: whole-report. Participants were given an empty grid and write down as much as they could remember in the exact positions. The rest they guessed.
Condition two: partial report. 50 ms after having seen the entire grid, a sound cue told the participants which row (out of 3) to recall.
R: Condition one remembered a mean 4 out of 12, condition two. Condition two usually remembered the entire row.
C: As condition two did not know which row to look at until after seeing it, we know that participants had access to the almost the entire grid for at least a short period.
E: This is line with the multi-store memory model, as the information goes stays in the sensory memory very briefly and can be processed before it decays.
Glanzer & Cunitz (1966)
A: Investigate the serial position effect.
M:
P: Participant were shown a 20 word list and were asked to recall in a free order. In another condition, it was still free recall but there was a filler task inbetween so they could not rehearse.
R: People remembered the first words (primacy effect) and last words (recency effect) better. Second condition eliminated recency effect (because no rehearsal).
C:
Criticisms regarding the Multi-store memory model
- It focuses on the structure, not the process of memory. This is a problem because memory is defined as a cognitive process.
- It’s an oversimplification of the transfer of STM to LTM. It’s not just rehearsal, it can be something that triggers an emotional response for example.
- There have been identified different types of memory that are stored differently: episodic, procedural, semantic, emotional, perceptual.
- Flow of the memory is bidirectional, not linear. E.g. knowledge is used from LTM to process information.
Working memory model (Baddeley & Hitch, 1974)
Focuses on the structure of STM through the dual-task technique (two memory operations simultaneously).
- A central executive (simply a construct) coordinates three subsystems: visuospatial sketchpad, phonological loop & episodic buffer.
- Visuospatial sketchpad: inner eye, all information visual and spatial
- Phonological loop: the inner ear; holds someone’s voice, and the inner voice allows for rehearsal and lets us read a book in our head: An articulatory rehearsal component (inner voice [does: 1) visual stimuli-sounds (phonological store) 2) rehearsal of sounds-LTM (auditory process)]
- Episodic buffer (added in 2000): integrates information from other stores and links to LTM structures.
Typical conceptualization of memory
Stages – sensory, short-term memory, long-term memory.
Processes – encoding, storage, retrieval.
Types – implicit memory (procedural, emotional, perceptual memory), explicit memory (semantic memory, episodic memory).
Types of encoding
Acoustic encoding, which usually applies to incoming auditory inputs like music or a teacher’s voice.
Visual encoding, which transforms incoming visual inputs including images, faces and words printed on a page.
Elaborative encoding, which connects or relates new inputs to existing memories, which in turn makes the new inputs easier to remember.
Semantic encoding, which focuses on the meaning of an input or how it might be applied.
Support and limitations for the working memory model
Conrad & Hill (1984) - investiaged phonological similarity effect (rhyming letters)
Baddeley, Lewis, Vallar (1984) - articulary suppression
Baddeley (1996) - central executive, attention switches
Limited by it’s complexity, though that is also it’s strength. It is hard to investigate the whole model, only specific parts of it at a time.
Conrad & Hull (1964)
A: investigate phonological similarity effect.
M: repeated measures-design
P: The recalling of rhyming letters (b, d, c, g, p) vs non-rhyming (r, t, h, p, r, x).
R: Non-rhyming letters are easier to remember as the rhyming letters can be mixed up.
C: The non-rhyming words were easier to remember as the rhyming ones and similar sounds and hence could be confused
E: Supports the working memory model
Baddeley, Lewis, Vallar (1984)
A: Investigated the effects of articulatory suppression on the phonological similarity effect (blocking inner voice; preventing rehearsal)
M: repeated measures design
P: Repeating the same sound over and over while doing an experimental task, as to fill up the inner voice. Either they were asked to remember rhyming or non-rhyming words orally or in written form.
R: Orally, the rhyming words were harder to remember similar to Conrad & Hill (1984) and for the written mode, the recall was the same for rhyming and non-rhyming.
C: The fact that written mode had no difference was concluded to be because the articulatory suppression didn’t interfere with memory because of the visuospatial sketchpad.
E: Supports the working memory model
Baddeley (1996)
A: Investigate the central executive and how it works with attention (attention switches)
M: repeated measures
P: Participants required to play a random sequence (so not to create a pattern) on a keyboard while doing one of three other things: 1) reciting the alphabet, 2) count 3) alternate between 1) and 2) (A1, B2, C3)
R: The random sequences were more random in condition 1) and 2) but less so in 3)
C: 3) requires attention, therefore we mess up more and are less automated. Hence the central executive works on attention switches.
E: supports working memory model
