Please enable JavaScript.
Coggle requires JavaScript to display documents.
Compare and contrast the effectiveness of deep versus shallow learning in…
Compare and contrast the effectiveness of deep versus shallow learning in relation to memory retention
levels of processing hypothesis (Craik and Lockhart, 1972) [4]
-
-
information is processed at varying depths. Deeper processing leads to a more durable memory trace - we remember it longer.
-
neuroplasticity
experience dependent neuroplasticity/ synaptogenesis - having more experience in something causes more synaptogenesis [W]
-
- "The Levels of Processing Theory lacks clarity regarding what constitutes “depth.” Some researchers argue that depth might be more complex and nuanced than the basic levels described. For instance, personal relevance and emotional salience can also affect memory but aren’t accounted for in Craik and Tulving’s model, which might oversimplify how depth relates to memory retention." [GPT]
-
episodic buffer [3]
stores multidimensional information and distributes it to other parts of the working memory, or to the long term memory
-
-
-
types of learning
shallow learning
-
Ebbinghaus (1885) [3]
Bartlett (1932) criticised his study: he learned meaningless material, not naturalistic. He performed a study that looked at how well participants could recall folk tales from other cultures, which found that people warped the folk tales to fit their pre-existing schemas. Therefore the LTM encodes semantically - meaning was derived from the information by making meaningful links to things we already know. [4]
their recollection of the folk tales were also shorter - maybe this is something worth finding a paper about?
taight himself nonsense syllables, learned them through maintenance rehearsal, recalled them
-
the distributed practice effect/ the spacing effect - evenly distributed learning results in better learning compared to cramming
expanding retrieval (Landauer and Bjork, 1978) [3]
when studying, you should alternate learning and testing. There should initially be a short delay between studying and testing that gets larger over time.
Pashler et al., 2007 - to optimise this further, you should have a study session, interstudy interval, second study session, retention interval, and then a test.
The interstudy interval should be 10-20% the size of the retention interval.
Bekerian and Baddeley (1980) proved that this is not a repetition effect/ being better at something you do repeatedly. Repetition of information didn't cause better learning, rather deliberate, meaningful practise (explicit learning) was better.
"expanding retrieval is therefore not just a repetition effect, rather a combination of the distributed practice effect, the testing effect and the generation effect" [3]
the testing effect (Roediger and Karpicke, 2006)
In Gurung's (2005) study about the correlations between different study techniques and final exam scores, self-testing was the strongest predictor of a good score
Participants had to learn a passage of text in 1 of 3 conditions:
read the passage 4 times (RRRR)
read the passage 3 times, then test (RRRT)
read the passage, then 3 tests (RTTT)
- 2 more items...
the generation effect (Metcalfe and Kornell, 2007)
when you "generate" material (actively recall it) you'll remember it better than if you passively engaged with/ reviewed it.
- 4 more items...
repetition priming - improvements in behavioural responses to stimuli (e.g. accuracy or reaction time) when you're shown similar stimuli beforehand. [W]
-
surface level processing/ memorisation of patterns (not a deep conceptual or meaningful understanding) facilitates processing
-
-
-
-
-
feedback (Kay, 1995)
when repeatedly recalling something, participants often kept making the same mistakes/ inocrrectly recalled the same pieces of information. This effect was eliminated when feedback was given
-
-
-
-
-
-
-
-
-
Slamecka and McElree "Normal Forgetting of Verbal Lists as a Function of Their Degree of Learning" (1983) - the degree of initial learning significantly affects the rate of forgetting. The forgetting rate decreases with deeper learning. How well you initially learn something impacts how well you can retrieve it in future.
Participants were split into 2 groups (low amount of studying/ high study) and tested immediately/ 1 day/ 5 days after studying. They had to learn auditory word lists, visual pairs of associated words, or meaningful sentences (different types of learning). High study group always recalled more.
Rohrer and Taylor "The effects of overlearning and distributed practise on the retention of mathematics knowledge" (2006) -
overlearning condition vs non-overlearning/ massed learning (1 study session) vs distributed learning. 4 conditions total. Tested 1 week and 4 weeks after learning. After 4 weeks, distributed practice performed better. No differences after 1 week, or between overlearning/ non-overlearning groups.
"Long term retention is boosted by distributed practice and unaffected by overlearning"
Overlearning doesn't impact how well you can recall information. Distributed practice is better when you're trying to learn something you'll have to recall after a longer time. Overlearning had slight short-term benefits compared to non-overlearning.
is there an artificial grammar learning paradigm that assesses how long participants can remember the grammar rules? If participants can remember grammar rules for a long time, thsi contradicts the idea that shallow learning does not enable long term retention of learned information. However it could be argued that learning grammar rules still requires deep processing despite the fact that doing so occurs unconsciously/ implicitly because the grammar rules are complex. It's not the kind of thing you could simply learn through repetition. They must require at least some understanding of meaning.
vvv
shallow procedural memories are highly contex dependent. are other types of memories that are learned shallowly also context dependent?
vvv
^^^
could find a paper which suggests that deep processing/ learning is linked with synaptic growth, enabling greater memory retention? whereas shallow learning doesn't impact synaptic change much (or memory retention by extension)