Schemadriven Memory and Structural Alignment

1
Schema-driven Memory and Structural Alignment
C. Hunt Stilwell Arthur B. Markman
Experiment 3 Mean Number of Properties Recalled
Experiment 2 OLD Sentences Mean Recognition

• Experiment 1 Mean Recall
• Results For both sub-targets, participants
  recalled more when they received the
• analogous Base Story (m 6.33 Target 1, 4.67
  Target 2) than when they received the
  non-analogous

• Schematic Memory
• Schematic memory research flourished in the
  1960s and 1970s
• Researchers produced several findings relating
  to memory for thematic information.
• Schemas select information for encoding
  participants recall schema-relevant information
• better than schema-irrelevant information, and
  schemas facilitate recall for relevant info
• (e.g., Bransford Johnson, 1972).
• Schemas are abstract participants remember the
  gist, rather than the surface details
• (Bransford Franks, 1971).
• Schemas allow for inferences about instances,
  including false ones (e.g., Sulin Dooling,
• 1974).
• Memory takes place through an on-line
  reconstructive process from abstract gist
• representations (Bartlett, 1935).
• Lacking a concrete theoretical framework,
  however, this line of study, and the questions it
  posed,
• gave way to the list-learning research of the
  1980s.
• Structural
  Alignment
• Analogy has been shown to influence memory by
  highlighting relevant information, and ignoring
• irrelevant information (Markman Gentner,
  1997 Schustack Anderson, 1979). This is
  similar
• to the role of schemas.
• It is known that analogy, and comparisons in
  general, use the process of structural alignment

• Recall Task 1 Mean Recall
  Target 1 Mean Recall Target 2
  Mean Recall
• Results
• In first recall, participants recall more when
  they have read the analogous Base Story (m 5.89
  Target 1,
• 3.83 Target 2) than when they saw the
  non-analogous Base Story (m 4.64 Target 1, 3.11
  Target 2).
• In second recall, participants recalled more
  from Targets for which the non-analogous story
  had been
• read in encoding (m 5.26 Target 1 , 3.75
  Target 2) than they had in the first recall (m
  4.64 Target 1, 3.11 Target 2).
• Conclusions Replicate Experiments 1 2
  replicate Anderson Pichert, with a change in
  analogical correspondence (change in
• perspective) leading to recall of previously
  unrecalled information.

2a 2b
Experiment 2 New Sentences Mean Recognition

• Summary of Results
• Experiment 1 Replicate Bransford Johnson -
  memory is better for analogous, or
  schema-relevant information
• than for non-analogous, or schema-irrelevant
  information.
• Experiment 2a Replicate Sulin Dooling with
  short delay participants do not make many
  intrusions. Consistent with
• Alba Hasher with short delay,
  schema-relevant and schema-irrelevant information
  recognized equally well in a recognition
• task.
• Experiment 2b Replicate Experiment 1
  Replicate Sulin Dooling with long delay-
  schema-relevant information
• intrudes on memory for analogous (thematic)
  information frequently.
• Experiment 3 Replicate Experiments 1 2b
  replicate Anderson Pichert - a change in
  analogical correspondence
• (shift in perspective) yields recall of
  previously unrecallable information.
• Conclusions
• The goal of these studies was to replicate
  major findings from the schema-driven memory
  literature using an
• analogical paradigm, allowing us to then carry
  over what we know about analogy, and the
  structural alignment
• process, to the study of these memory
  phenomena. Experiments 1 2 replicated two major
  findings, and the
• preliminary results from Experiment 3 are
  consistent with a third major finding. Thus, we
  can conclude that
• schematic memory is driven by analogy.
• We can now begin to carry over what we know
  about analogy to the domain of schematic memory.

• Sulin Dooling, (1974)
• During encoding presented participants with
  biographical characters in which the main
• character was either given the name of a
  famous person (e.g., Hellen Keller) or given a
• novel name (e.g., Carol Harris).
• Delay of 5 minutes or 1 week
• Recognition Task When the main character was
  given the name of a famous person,
• participants mistakenly remembered having seen
  information about that character that was not
• presented in the original biographical story
  with 1 week delay, but not with 5 minute delay.
• Conclusion Schema-relevant, or thematic
  information intrudes on memory for thematic
• information. This effect increased with the
  length of delay (5 minutes vs. 1 week).
• Experiment 2a and b
• Goal Replicate experiment 1 using a
  recognition task replicate Sulin Dooling
  (1974).
• Prediction With a short delay (20-30
  minutes), participants will be able to
  distinguish
• between old and new sentences (Sulin
  Dooling, 1974. In addition, they will not
• show a preference for schema-relevant
  information (Alba Hasher, 1983). With a
• long delay, participants will mistakenly
  recognize intrusion sentences that are
• schema-relevant, and will remember
  schema-relevant old sentences better than

2a
2b

• Conclusions
• In Experiment 2a, there was no preference for
  schema-relevant over schema-
• irrelevant information at a short delay with
  a recognition task, consistent with Alba
• Hasher (1983).
• In Experiment 2b, as in Experiment 1, memory
  for analogous, or schema-
• relevant, information was better than memory
  for non-analogous, or schema-
• irrelevant, information, this time in a
  recognition task.
• Replicate Sulin Dooling using the analogical
  reasoning paradigm -
• information relevant to the analogical match
  (schema-relevant information)
• present in one domain intrudes on memory for
  the other domain after long delay (2b),
• but not after a short one (2a).

• Encoding
• Comparison Task Participants were given pairs
  of stories, one story at a time, and asked to
• rate their similarity.
• Target Stories The second story in each pair
  served a target, that contained two
• sub-stories (Sub-story 1 Sub-Story 2) that
  had different sets of relations and objects,
• but the same characters. Each Target Story
  was given a title.
• Base Stories The first story in each pair
  served a base. For each Target Story, there
• were two separate Base Stories, each
  corresponding to one of the two sub-stories in
• the Target Story.
• For each Base-Target pair a participant
  received, the Base formed the Schema, and
• the Target sub-story to which it corresponded
  was analogous, or relevant to the
• Schema, while the Target sub-story to which
  it did not correspond, was irrelevant to
• the Schema.
• After the Encoding Task, participants received
  a Filler Task or Delay.
• Retrieval
• Participants were given the titles of the
  Target Stories from each pair, and either a
  recall or
• recognition task for information from the
  Target Stories.

• Anderson Pichert, (1978)
• Presented participants with the story of a boy
  who skipped school, during encoding.
• Participants were asked to read the story from
  the perspective of either a Home Buy or a
• Burglar
• First Recall Task Asked to recall as much as
  they could from the original story. Participants
• recalled more information relevant to the
  perspective they took than information relevant
  to
• the other perspective.
• Perspective Shift Told to now think about the
  story that they had ready from the other
• perspective.
• Second Recall Task Asked participants to
  recall the original story again. Participants
  now
• recalled information relevant to their new
  perspective that they had not remembered in the
• first recall task, as well as the information
  relevant to their first perspective that they had
• remembered in the first recall task.
• Conclusion With a shift in perspective,
  participants are able to recall information the
  had
• previously not remembered.
• Experiment 3
• Goal Replicate experiment 1 using a
  recognition task replicate Anderson Pichert
  (1978).
• Prediction Memory for information from the
  analogous Target sub-story (schema-
• relevant information) will be better than
  memory for information from the non-analogous

• Bransford Johnson (1972)
• Participants were given paragraphs containing a
  target sentence that does not fit with the
• schema invoked by the story..
• E.g., Peace March story containing a
  sentence about a craft landing on an alien
• planet.
• During retrieval task, participants fail to
  recall the schema-irrelevant sentence, unless
  they
• had been given a title that facilitated the
  creation of a schema within which that sentence
  is relevant
• (e.g., Landing on an alien planet.)
• Conclusion Memory is better for
  schema-relevant than schema-irrelevant
  information.
• Experiment 1
• Goal Replicate Johnson, et a. (1972) using the
  analogical reasoning paradigm described
• above.
• Prediction Memory for information from the
  analogous Target sub-story (schema-
• relevant information) will be better than
  memory for information from the non-analogous
• Target sub-story (schema-irrelevant
  information).
• Participants 60 undergraduates at the
  University of Texas, Austin. Participants
  received
• course credit for their participation.
• Method
• Encoding Comparison Task

• Alba Hasher (1983). Is memory schematic?
  Psychological Bulletin, 93(2), 203-231.
• Anderson, R. C. Pichert, J.W. (1978). Recall
  of previously unrecallable information following
  a shift in perspective. Journal
• of Verbal Learning and Verbal Behavior,
  17(1), 1-12.
• Bartlett, F. C. (1932). Remembering.
  Cambridge, England University Press.
• Bransford, J. D. Franks, J. J. (1971). The
  abstraction of linguistic ideas A review.
  Cognitive Psychology, 2(4), 331 - 350.
• Bransford, J. D. Johnson, M. K. (1972).
  Journal of Verbal Learning and Verbal Behavior,
  11(6), 717-726.
• Gentner, D. (1983). Structure mapping A
  theoretical framework for analogy. Cognitive
  Science, 7, 155-170.
• Markman, A. B. Gentner, D. (1993). Structural
  alignment during similarity comparisons.
  Cognitive Psychology,
• 25(4), 431-467
• Markman, A.B. Gentner, D. (1997). The effects
  of alignability on memory. Psychological
  Science, 8(5), 363-367.
• Schustack, M. W. Anderson, J. R. (1979).
  Effects of analogy to prior knowledge on memory
  for new information.
• Journal of Verbal Learning and Verbal
  Behavior, 18(5), 565-583.
• Stilwell, C. H. Markman, A. B. (2001). The
  fate of irrelevant information in analogy.
  Proceedings of the 23rd Annual Conference
• of the Cognitive Science Society.
• Sulin, R. A. Dooling, D. J. (1974). Intrusion
  of a thematic idea in retention of prose.
  Journal of Experimental
• Psychology, 103(2), 255-262.

• Method
• Encoding Comparison Task. Each Base Story
  contains information relevant to the
• theme which is not contained in the
  corresponding Target Story.
• Filler Task (15-30 minutes)
• Retrieval 1 Recognition Task
• Change of Perspective Read the Base Stories
  they had not seen in the Comparison
• Task
• Retrieval 2 Recognition Task