5. L5 principles and dissociations of memory

Question 1

abstraction

Answer 1

• People memorise conglomerations of ideas
• Not verbatim information
• People appear to lose specific surface information (shortly after study)
• People retain “gist”: Generalised, abstracted information
• People infer things they never heard or saw
• Explanations and models must accommodate abstraction
• Most of our log term memory is devoted to things which are devoid of detail but maintain the gist of the info

Question 2

Deese-Roediger-McDermott false memory paradigm

Answer 2

• Study list of words
• cloud, drop, water, sun, storm, shower,…
• In subsequent memory test, up to 70% of participants remember the item “rain”
• “rain” is not on the list…
• …but all list items are associates of rain
• false memory can be seen as an instance of generalisation or abstraction

Question 3

hyper specificity

Answer 3

• The idea that abstraction is encoded but specific details about a stimulis are also encoded and we show facilllitation for that item in a test
• Memory for surface features persists
• Manifest primarily on indirect tests of memory (priming; but also some forms of direct tests such as image recognition)
• Explanations and models must accommodate hyper specificity
• Favours models that retain all aspects of each stimulus experience
• Indirect tests of memory have little to do with memory tests as they are conventionally understood (i.e., direct tests with intentional retrieval)
• Now search for specificity in direct tests of memory more generally

Question 4

hyper specificity is long lasting

Answer 4

Kolers 1976
- Participants quicker at rereading same text they had read a yer ago
Mitchell (2006)
- Found priming after 17 years
With the camel shapes

Question 5

hyper specificity and encoding context

Answer 5

• Any stimulus we encode is encoded in a specific context, e.g. * time of encoding * place * other stimuli (people, things) present * task specifics * etc.
• At a basic level, context can influence encoding efficacy,
  ○ eTask set e.g., depth-of-processing
  ○ Unitization (feature binding demands) easier to remember vs. (Karlson et al., 2010)
  ○ Synchronicity better memory for words chanted together with others (von Zimmerman & Richardson, 2016)
  More generally, context matters because context elements are always encoded with a stimulus, and can thus be used as retrieval cues

Question 6

hyper specificity encoding context examples

Answer 6

Encoding Specificity Principle (Thomson & Tulving, 1970)
- match between study and test matters
- target word = flower
- strong cue = bloom
- weak cue = fruit
- remembered best at test when cue was the same as what they studied, whether it was weak or strong cue.

Encoding specificity: Godden & Baddeley (1975)
- learning and recalling test material under water
- Impairment if you went from one context to the other context
- It also works with internal states, e.g. relating to substance intake (alcohol, nicotine, caffeine) and mood (i.e., state-dependent memory)

Question 7

abstraction vs hyper specificity

Answer 7

Specificity apparent with
- Memory tasks that require perceptual processing of specific cues
- Performance benefits when a cue present at study is repeated at test
- e.g., repetition priming (implicit) or recognition of images (explicit).

Abstraction apparent with
- non-specific, direct (explicit) tests
- mainly free recall
- Abstraction is then adaptive and useful
- Free recall of prose
- But can also be maladaptive when you are led into abstraction
Abstraction can contribute to false memories