Memory and Forgetting: Lecture Notes

Forgetting and False Memory

Objectives

• Understand major theories of why forgetting occurs.

• Evaluate empirical data supporting these theories.

• Understand how people study "false memory" and what this tells us about memory formation and retrieval.

• Appreciate the limitations of these methods.

Forgetting

• When we talk about “forgetting”, we’re interested in how information is lost from long-term memory.

• "". . .the inability to recall something now that could be recalled on an earlier occasion” (Tulving, 1974)

Key questions:

• Why do we fail to recall?

• Why do we recall incorrect information?

• Why are some memories resistant to forgetting?

Ebbinghaus (1879)

• Self-experimentation using novel nonsense words.

• Encoding method: "Train to criterion".

• Retrieval method: "Savings in learning".

Ebbinghaus' Forgetting Curve

• The forgetting curve illustrates how much we forget each day.

• There's a very quick initial loss of information.

  • 20 minutes: 58% retained

  • 1 hour: 44% retained (halfway gone!)

  • 1 day: 33% retained

  • 6 days: 25% retained

Ebbinghaus’ Key Findings

• Forgetting curve: A decrease in memory retrieval performance (forgetting) is non-linear.

• Fast forgetting initially, with more gradual forgetting later.

• Associative processes play a role.

• Variations in difficulty of material affect forgetting.

• Stress/sleep impact forgetting.

Modern Methods

• Ebbinghaus’ methods are very labor-intensive!

• Commonly used memory performance tests today:

  • Recognition: Identifying previously presented items from new ones (e.g., "Did you see ‘bread’ or ‘apples’?").

  • Free recall: Effortful retrieval from memory without external cues (e.g., "Which foods did you see?").

  • Cued recall: Retrieval from memory using an associated cue or hint (e.g., "Can you complete this food item ‘a[underscore]l[underscore]s’?").

Interference Theory

• There’s lots of content in memory. We’d fill it up quickly…

• Why do we forget?

• Memories interfere with each other and compete for storage.

Proactive Interference

• Old memories seem to stop new memories from sticking.

• It’s called “Pro-active” because the effects of the interference are felt in the future.

  • Example: moving house and navigating home; taking the wrong turns due to old memory of previous house location.

  • Example: general likes and dislikes (e.g., food, actors, people) are not based on the most recent experience.

Bäuml & Kliegl (2013)

• Study task: learn lists of words

• Conditions:

  • “remember” (PI): Study List 1, Instruction: Remember List 1, Study List 2

  • “no PI”: --- Filler task, Study List 2

• Recall task: Recall List 2 items

Proactive interference

• Typically thought to be due to problems with retrieval.

• People search through irrelevant memories, and this is where interference occurs.

• Can such retrieval issues be mitigated (controlled)?

• If so, can we just tell people to forget information?

Bäuml & Kliegl (2013) Experiment

• Study task: learn lists of words

• Three Conditions:

  1. “remember” (PI): Study List 1, Remember List 1, Study List 2

  2. “no PI”: --- Filler task, Study List 2, Recall task: Recall List 2 items

  3. “forget” (PI): Study List 1, Forget List 1, Study List 2

• These data suggest PI is controllable

• This “directed forgetting” technique is very effective.

Retroactive Interference

• New memories seem to “overwrite” or push out older memories.

• It’s called “retro-active” because the effects of the interference are felt on the memories from the past.

  • Example: learning a second language and forgetting words from the first language.

  • Example: useful to keep memory current; face of an old close friend/relative: What did their face look like 5 or 10 years ago?

Effects of RI

• RI seems to be strongest when the new information resembles the old information.

• Misra et al. (2012) studied picture naming in bilinguals – Chinese speakers proficient in English

  • If they named pictures in English first, their time to name pictures in Chinese was impaired.

  • Suggests that the formation of new memories (English names) leads to inhibition of competing old information (Chinese names)

Understanding Forgetting

• PI and RI suggest that the hard disk analogy is a poor one…

• Memories are associated.

• Memories have semantic content.

• Memory is relational.

• These properties lead to interference, resulting in storage and retrieval issues.

Cue-Dependent Forgetting

• Recall the albums of…

  • One Direction

  • Coldplay

  • Foo Fighters

Cued Recall

• Recall is better when cues are provided: cued recall is generally better than free recall.

• Relevant cues are anything that is associated with the item at the time of encoding.

• Associated items can be external, such as the context/room in which it is learnt, or internal, such as the mood state.

Godden and Baddeley (1975)

• Forgetting is greatest when there’s a poor match between the information in the memory trace and that available at retrieval (Tulving, 1979)

• Encoding specificity principle

Why Do We Forget?

• Because old (PI) and new (RI) information interferes with retrieval.

• Because the retrieval context doesn’t match encoding.

• Why is Ebbinghaus’ forgetting curve non-linear?

• Why is this initial dramatic forgetting followed by almost no forgetting?

• Repeated encoding

Consolidation

• New memories are in a fragile state and need to undergo a transfer to long-lasting memories.

• Transfer from hippocampus to the neo-cortical regions.

• When recent memories are retrieved, the hippocampus is more active.

• At this point, memories are assumed to be vulnerable to interference.

Failures of Consolidation

• Moulton et al. (2005): drinking alcohol before encoding impairs memory formation

• What about drinking alcohol after encoding?

  • Prediction: Alcohol should disrupt new memories, so less interference.

  • Moulton et al. confirmed this pattern; better memory for information encoded before alcohol.

• Failure to rehearse.

• Failure to sleep.

• Intoxication!

Reconsolidation

• This theory is now a very hot topic

• Suggests a paradox of memory: Successful retrieval potentially weakens that memory

• Retrieval places memory in a “labile state” and can be modified

• Huge implications for the clinic (anxiety, trauma, phobia) and society…

False Memory

• We’ve seen that memory is subject to errors and interference.

• It is not a veridical copy of experience. It is fragile and can be malleable.

• This has huge implications for society (crime, trauma, phobia, addiction)

Eye-Witness Testimony

• Let’s think back to the car video you watched earlier.

  • A: Approximately how fast was the car travelling when it smashed into the car that was stopped at the intersection (mph)?

  • B: Approximately how fast was the car travelling when it contacted the car that was stopped at the intersection (mph)?

Speed Estimates

• Participants estimated the speed to be much greater with "forceful" verbs.

  • Smashed: $40.8$

  • Collided: $39.3$

  • Bumped: $38.1$

  • Hit: $34.0$

  • Contacted: $31.8$

Loftus Results

• It is claimed that the “memory” of the event was significantly altered by the prompt.

• A false memory was created.

• Are there alternative accounts of these effects? Or problems with this interpretation?

Problem 1

• Maybe the prompts are recalling true past memories.

• “smashing” might evoke a memory of a recent car crash with broken glass

• This might conflict with recall of the event, and so a false memory hasn’t been created.

• Can we develop a better test of false memory?

Truly "False Memories"

• Can participants create false memories for events that could not have possibly taken place?

• Wade, Garry, Read, & Lindsay (2002)

Method (Wade et al., 2002)

• Interview 1:

  • Free recall

  • Guided imagery: close their eyes and picture being back in the hot air balloon… think… what the weather was like, and what they saw when they looked over the side of the basket.

  • Subjects given photo booklet and asked to focus on recalling the forgotten event(s) each night.

  • < 3-7 days >

• Interviews 2 & 3:

  • Guided imagery techniques used. < 3-7 days >

Result (Wade et al., 2002)

• On this basis, 50% of the participants in this experiment generated a false memory

• How did they classify false memory?

  • To be classified as having a clear false memory, a subject had to report memories of the critical balloon ride, including consistent elaboration of information not depicted in the photograph.

  • A partial false memory occurred when a subject consistently elaborated on the false photograph (such as reporting feelings, who was present, etc.) but did not indicate memories of taking the balloon ride per se.

  • When a subject tried to recall the false event and described images of it but did not claim these images as memories, it was categorized as trying to recall.

  • When a subject failed to attempt any recall, this was categorized as no memory.

Problem 2

• Are these just “demand characteristics” (DC)?

• DC: when a participant’s response reflects an expectation brought about by the experimenter, rather than the manipulation per se.

• In car crash experiment, participant reads “smashed” and concludes a higher speed is desired by the experimenter

• This is difficult to control for, and it should be considered an issue with this type of research.

Can This Be Solved? (Laney et al., 2008)

• Aim: develop a false memory for childhood likes and dislikes for foods.

• Red herring: planted clues to lead participants towards another purpose to the experiment (diet, health)

• At the end they asked participants to openly state what they thought the purpose was.

• Results

  • They found that participants who believed the “red-herring” as the purpose, still showed false memories.

  • Those who believed the “true purpose” (examining false memories) showed equivalent levels of false memories.

False Memories

• Memory is not veridical (not based entirely on reality)

• Can be distorted in various ways:

  • Imagination

  • Source monitoring errors

  • Information from before and after memory event

  • Reconsolidation might be key here

Why Are These Findings Important?

• Recall of accurate memories is critical for society, e.g.

  • Eye-witness testimony

  • Abuse / trauma