Episodic Reminding: The effectiveness of augmenting memory by using reminders of previous events

1
Episodic RemindingThe effectiveness of
augmenting memory by using reminders of previous
events

• G. A. Radvansky
• University of Notre Dame
• David E. Copeland
• University of Southern Mississippi

2
Augmented Cognition

• For an augmented cognition device, one role it
  could play is in reminding people of things.
• One type of reminding is of previous, similar
  experiences (episodic cuing).
• The current study attempts to identify whether
  this type of information is useful to people.

3
Memory and Cuing

• People can sometimes forget, or at least not
  realize that they need information.
• Memory can be improved with some assistance. This
  assistance is called cuing.

4
Cuing Helps Memory

• Semantic cuing
• doctor cues nurse
• Context cuing
• Encoding specificity
• Content cuing
• A guitar can be used to paddle a canoe.
• Cue guitar

5
Cuing Hurts Memory

• Part-set cuing
• People are provided with part of a list.
• People find it harder to remember any given item
  than if no cue had been given.

6
Episodic Cuing

• Episodic cuing
• Reminding people of specific events in the past
• Related to the current situation.
• Does episodic cuing help or hurt memory retrieval?

7
Forsythe Industries

• Subjects told that they were put in charge of
  running a factory.
• Need to learn a set of tasks
• Each task composed of
• Task number
• Three components
• Client name
• Client need

8
Study Screen
9
Forsythe Industries

• After studying the information, people need to
  recall the task number, components, client name,
  and client need.

10
Training Recall Screen
11
Forsythe Industries

• Task number and components are more semantic in
  nature.
• They are consistent across episodes
• Client name and need are more episodic, and are
  seen only once during training.

12
Forsythe Industries

• After training was completed people were given a
  recall test.
• No Episodic Cue (Control)
• Just the task number
• Episodic Cue (Experimental)
• Task number plus client name and need.

13
Testing Recall Screen
14
Experiment 1

• N 37
• This experiment assessed the influence of 3
  memory phenomena on episodic cuing
• Recency
• Frequency
• Type of Practice (massed vs. spaced)

15
Experiment 1

• Recency
• Episodic cue was from either a recent event or a
  distance event.
• Frequency
• Episodic cue was from a task type that occurred
  either 2 or 8 times during training.
• Type of Practice
• Episodic cue was from a task type in which
  training was either massed or distributed.

16
Experiment 1

• Results
• Episodic cuing aided memory
• Cued .30
• Uncued .37
• This is a significant improvement

17
Experiment 1

• Effect of Delay
• Short .39
• Long .28
• Effect of Episodic Cuing on Delay
• Cued Uncued
• Short .38 .39
• Long .35 .20

18
Experiment 1

• Effect of Frequency
• 2 times .25
• 8 times .42
• Effect of Episodic Cuing on Frequency
• Cued Uncued
• 2 times .29 .20
• 8 times .44 .40

19
Experiment 1

• Effect of Practice
• Massed .29
• Spaced .38
• Effect of Episodic Cuing on Practice
• Cued Uncued
• Massed .35 .23
• Spaced .38 .37

20
Summary

• Episodic cuing generally boosted performance.
• Cuing better aided retrieval of
• Older episodes
• Massed practice episodes
• Cuing had no benefit for
• Newer episodes
• Space practice episodes
• Cuing had no differential effect as a function of
  frequency.

21
Experiment 2

• N 46
• This experiment assessed the influence of 2
  memory phenomena on episodic cuing
• Interference
• Frequency (again)

22
Experiment 2

• Interference
• When there are an increase in the number of
  associations with a concept, memory performance
  may worsen.
• Frequency
• Episodic cue was from a task type that occurred
  either 2 or 6 times during training.

23
Experiment 2

• Results
• Episodic cuing did not aid memory
• Cued .28
• Uncued .29
• There is no significant difference

24
Experiment 2

• Effect of Frequency
• 2 times .31
• 6 times .27
• Marginally significant difference
• Effect of Episodic Cuing on Frequency
• Cued Uncued
• 2 times .31 .31
• 6 times .25 .28

25
Experiment 2

• Effect of Overlap
• No overlap .30
• Partial 1 .30
• Partial 2 .30
• Complete .27
• Effect of Episodic Cuing on Overlap
• Cued Uncued
• No Overlap .32 .27
• Partial 1 .27 .33
• Partial 2 .28 .32
• Complete .27 .27

26
Summary

• General episodic cuing is not universally found.
• Episodic cuing of high frequency items can hurt
  memory retrieval.
• The status of interference on cuing is uncertain.

27
Experiment 3

• N 29
• This experiment assessed the influence of
  episodic cuing on memory for structure rather
  than content.

28
Experiment 3

• Task numbers corresponded to 7 structures.
• Linear
• Circular
• Divergent
• Convergent
• Mediated
• Partially Independent
• Independent

29
Experiment 3
30
Experiment 3
31
Experiment 3

• Results
• Episodic cuing did not aid memory
• Cued .16
• Uncued .18
• There is no significant difference
• Similar patterns observed for the different
  structure types.

32
Summary

• Once again, there was no clear benefit to
  episodic cuing.
• However, performance was very low.

33
Conclusions

• Episodic cuing helped performance when
  information was (a) older, (b) mass practiced,
  and (c) relatively simple.
• Episodic cuing did not help performance when
  information (a) newer, (b) spaced practiced, and
  (c) more complex information.
• Episodic cuing may hurt performance for higher
  frequency information.