Explicit Memory, Forgetting and Amnesia Flashcards

1
Q

Describe a study which demonstrates the importance of level of representation

A

Text shown in 1) surface form (italics)
Text shown in text base form
Tex shown in mental model form (the prince holds the princess vs princess is_held by_prince- ) asked to picture in minds eye

And just asked whether it is in italics or not (low level of processing-implicit memory)

This effects how well it is remembered- the mental model is remembered best. This is a strong effect. Surface form remembered least well

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What “but” is there to this representation conclusion? (describe a study)

A

It depends what you want the person to get from the learning experience. (Transfer-appropriate learning)

students responded to words using either a meaning-based (deep-level) task, such as whether the word “plane” made sense
in the sentence “The __________ had a silver engine,” or a rhyme-based (shallow-level) task, such as whether the word “eagle” made sense in the sentence “__________ rhymes with legal.” Later, students took either a standard recognition test (a direct memory test) or a rhyming recognition test (an indirect test) in which they indicated whether a new word rhymed with one that they had heard earlier (e.g., “regal”). The results reveal that memory is better when the encoding and retrieval processes match than when they do not.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What are context effects? How effective are they?

A

Episodic memory improved when tested in the same context as you learned (Divers learning and testing underwater or on land). Or is it? Effects, if any, are usually very weak.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

How did the lecturer improve Baddeley’s study in a replication and could he replicate it?

A

could not replicate the context effect myself (60% of the experiments in psychology cannot be replicated)

But important difference with Godden and Baddeley: I removed a large confound–Divers underwater had to always get out, even in the water-water condition (and into the water in the land-land condition)

–In the original they could just rest
–But not in the land-water or water-land conditions

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Describe state-dependent memory with an example

A

While sober studying and sober testing results in best performance, sitting an exam sober after studying drunk results in a worse exam than sitting it drunk: i.e in the same state

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is meant by the retrieval practice effect?

A

If you try to remember something and with some effort are just able to do so, this gives a large learning effect. The learning benefit is much greater than merely relearning the material

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What depends one how effective this practice effect is?

A
  • The more you repeat, the better your memory performance
  • Long intervals are more effective than short
  • There is often an optimum repetition interval but how long depends on the person, the material, and the degree of prior learning
  • The computer can help optimise this, given a good learning model (Brainscape :))
  • Conclusion: testing yourself (at increasingly long intervals) is a very effective wayto learn
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

How effective is mere exposure?

A

little effect (what is on a 10e bill?, periodic table)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Why is mere exposure not that effective

A

you must pay attention to and consciously elaboratethe to-be-learned materials

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Describe spacing effects

A
  • Performance appears to depend on distribution
  • At a short retention interval: performance decreases with increasing P1-P2 (temporal distance)
  • At a long retention interval: performance increases with increase in P1-P2

where P1 and P2 are learning events

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What reasons are suggested for these spacing effects?

A

Longer inter-session intervals ensures
–More consolidation in between sessions
–Less habituation (less efficient processing)
–More diverse memory trace, responsive to more different cues (more contextual variability)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What did Jaap the chap discover from his study on memory and news stories?

A

Could formulate a decent forgetting curve

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What did Ebbinghaus use to formulate his forgetting curve?

A

Non-sense syllables

First tried but decided against

  • Poem stanzas
  • Tones
  • Numbers
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Describe how to calculate savings in two ways

A

The relative relearning time:
10 minutes vs 5 minutes = 50% savings

Initial learning:
8 times relearning until perfect vs 4 time = 50% savings

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What did Jaap change in his replication of the Ebbinghaus forgetting curve study? Describe the participants

A

Ebbinghaus 7 months, Jaap 2.5
E: 12-45 replications per time interval, J:10
J: Used dutch and removed syllables with too much meaning like SEX E: did not
E: does not state in detail when he learned what list, so we had to guess how many interfering lists there were

The subject was also a co-author and acted like an experimenter like ebbinghaus and did 70 hours of re-learning

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What materials did Jaap use in his study?

A

70 lists (10 per retention interval). Each list consisted of 104 syllables
–8 ‘rows’ of 13 syllables
–Rows were the main unit of learning
–They were rehearsed until once correct in ordered recall

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Following Ebbinghaus, Jaap did a practice phase. Why was this and what did it consist of?

A

To familiarize the subject with the savings measure, so there would be no ‘meta-learning’ during the actual experiment

  • 14 lists were learned and relearned after 20 min (1 list = 8 ‘rows’ of 13 syllables)
  • 19 lists were learned (and not relearned)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Describe the experimental phase of Jaap’s study

A
  • 69 lists learned between 1 Dec ’11 and 13 Feb ’12
  • Lists distributed as much as possible over the 10 weeks
  • Not possible for the 31-day interval
  • Lists for 31-day interval were learned at the beginning (early Dec) to fit within internship period
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

How was a 13-syllable row recounted and why?

A

One row was softly spoken at 150 beats per minute (following Ebbinghaus qua speed)

  • About 5.2 seconds per list
  • More like a complicated foreign sentence or long word than standard word list

Words specified in a specific rhythm (dih-dah, dih-dah dih-dah), E rathered a 3/4 rhythm (dih-dih-dih-dah, dih-dih-dih-dah)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Describe the relearning phase of Jaap’s replication

A

Relearning was done after one of 7 retention intervals (following Ebbinghaus):
–20 min, 1 hour, 9 hours, 1 day, 2 days, 6 days, or 31 days

  • First relearning attempt was voice-recorded and later checked for correctness per syllable
  • Then relearning proceeded until once correct
  • Savings scores were calculated
21
Q

Where did Jaap find a spike in the data, replicated in other studies and what does he argue from this

A

There is a jump in the retention after 24 hours compared to those before and in contrast to the general forgetting curve, likely the effect of consolidation during sleep

Considered as ‘noise’ by Ebbinghaus (he later ran more trials for the 24-hour data point)

22
Q

Were there any time of day effects found?

A

–E. did notice large time-of-day effects and corrected for these (up to 13% better savings in the morning)
–Heller et al. did not find time-of-day effects

Jaap did find them

23
Q

How did the study replicate overall? (3)

A
  • On the whole: Remarkable similarity in the forgetting curves
  • Curves are not identical though: Ebbinghaus’ curve is not “The Forgetting Curve” (depends on person, materials, method used- savings, free recall, recognition etc)

•Strange: Drost’s 31-day point is much lower than the rest
–Likely: interference because all 31-day lists were learned at beginning of 10-week test period

24
Q

Were there primacy and recency effects found?

A

Yuh, very large serial position effect (negates effects for equivalent syllables)

25
Q

What type of functions do forgetting functions typically follow?

A

Power functions: A power function gives a straight line when plotted in log-log coordinates

Q(t) = (1 + u1(t))^-a1
Q(t) is savings as a function of time t

26
Q

How well do these savings follow a power function

A

Pretty good dawg, they roughly follow one. Even better if controlling for the 24 hour jump

27
Q

What did Harry Bahrick demonstrate about forgetting?

A

For very well learned materials, forgetting may stop (permastore).

Forgetting continues until about 3-12 years (depending of the materials) and then halts

28
Q

What is proactive interference and what paradigm demonstrates it?

A

Use a Brown-Peterson paradigm with three words; Have subjects recall after a filled interval of 30 s (counting backwards). Recall on subsequent trials gets worse and worse. This is proactive interference; previous knowledge interfering with new knowledge.

29
Q

What can cause a release from the proactive interference?

A

Unexpected stimuli cause a release from the proactive interference (e.g change in category gives an increase in performance)

30
Q

What is meant by associative interference? What effect is associated with it?

A

Associative interference reflects the complexity of newly learned information. The disruption of memory is not based on temporal order (as it is with proactive and retroactive interference) but on the number of associations with a concept.

For example, if you have just learned five things about Jenny, you will be slower to verify any one of these than if you had learned only one thing. Often, associative interference is described in terms of the fan effect. The term fan effect assumes that information is stored in a propositional memory network, with nodes representing individual concepts and links representing the associations among them. During retrieval, the more links “fanning” off of a concept, the greater the interference from the competing associations and retrieval time increases accordingly.

31
Q

Describe a study measuring the fan effect

A

Stanford University lists of sentences to memorize, such as “the doctor is in the park” or “the lawyer is in the museum.” The number of associations with the person and location concepts (e.g., doctor or park) was varied from one to three. Thus, there were one to three places that a person could be in, and one to three people in a location.

After memorization, a recognition test was given in which students indicated whether probe sentences were studied or not. Unstudied sentences were recombinations of people and locations, such as “the doctor is in the museum.” The results were that, as the number of associations with a concept increased, response time also increased.

32
Q

What is meant by the paradox of the expert and how do you get away from it?

A

A worrisome implication of associative interference is that the more you know, the harder it should be to remember. However, experts in an area actually have more information than novices with no deficit in remembering. This is known as “the paradox of the expert”.

A way out of this paradox is to use chunking. Information that is integrated into a common memory trace reduces the amount of interference because there are fewer traces to compete with one another

33
Q

Name and describe two forgetting theories

A

Interference theory
•Forgetting is the result of the competition between two memories

Inhibition theory
•Forgetting is the result of an active monitoring process whereby information that is “in the way” is suppressed

34
Q

Practice of cue A leads to a reduction in reproduction ofB

What can inhibition theory assume Is occurring? (2)

A

In order learn A, the (stronger) item B is inhibited and hence (disabled).

Item B is inhibited: the connection from the cue to B is weakened

35
Q

What is observed in the part set cueing task?

A

In a free recall task, if three items from the list are given, as a cue, then less items are recalled than if no cue is given

36
Q

What reasons could be given for the results of the part-set cueing task?

A

You are not just recalling 1 by 1 independently, you are recalling a number of associations. If you give people three, you ‘walk through’ this network in a manner which does not maximise performance

37
Q

How did Jaap investigate the learning curve?

A

Having a British person learn dutch through testing and feedback.

38
Q

What is the use in investigating the learning curve?

A
  • It is very useful to know the shape of learning
  • We can then predict and perhaps optimise the learning process
  • It is also of great theoretical relevance as it hints of underlying mechanisms
39
Q

What is the power law of learning?

A

p(t) = 1 - (t+1)^-u

40
Q

What does the u represent?

A

The learning rate: if u is high, the errors go down v quickly

41
Q

What does the p represent?

A

The probability that you will learn the translation (0-1)

42
Q

How do people argue against this power law of learning?

A

They claim that it is an exponential curve

p(t)= 1-e^ut

43
Q

What is the big difference between these curves?

A

In exponential the error rate goes down at an exponential rate per day e.g 32 errors, 16, 8, 4 etc

With Power function the rate at which it goes down slows down with time e.g 50% then 40% etc

Power function takes longer

44
Q

What did Jaap find in many cases?

A

Learning curves are in fact s shaped, especially for difficult tasks. Not much visibly progress at first then sharp ride before levelling off at 100.

This makes the mathematical form more complex

45
Q

How does task complexity correlate with the curve shape?

A

For easy, simple tasks, measured in a single individual the learning curve tends to be exponential

For complex tasks or averaged over many subjects, it tends to be a power function

46
Q

There is strong evidence for the S shaped curve if:

A

–learning rate is low

–perceived complexity or difficulty is high (many chunks: words, syllables, letters)

This is material dependent

47
Q

What does the learning rate largely depend on?

A

Learning rate seems to be an individual difference

48
Q

What are the overall conclusions on these curves?

A
  • It is not fruitful or even feasible to find the learning curve equation
  • We must zoom in on the underlying mechanisms and develop formal models for these
  • Only then can we discover the shape of learning