認知心理学8. Encoding for retrieval: 思い出すためのエンコード

認知心理学もくじ

• 認知心理学もくじ

Today

• If distinctiveness helps us to encode separate memory traces, what factors help us to recall those memories?

As we say yesterday, when levels of processing theory came out the focus was primarily on the levels.
As experiments were done, there were criticisms to the theory.
Simple explanation of forgetting, differences recall for semantic processing.
So the emphasis changed from focusing on the levels to focusing on elaboration.
Studies on elaboration gave us another piece of the puzzle.
It helped explain as to why we see differences in recall for deep level processing strategies, but it also gives more insight into forgetting.
The more we elaborate on information, the slower it should decay from memory or the slower we should forget it.
However,even though elaboration is a good way to increase the probability that the info will be retained in ltm, Recall will be even better if we can make it distinct form other types of information in LTM.

But first a little more about distinctiveness…

• primary distinctiveness:
  • An item that is distinct from other items in the immediate context
  • involves making the item distinct relative to the immediate context.
• Now the examples I gave you yesterday had to do with words. But we can go beyond the learning of words and apply it to different settings.

• secondary distinctiveness:
  • An item that is distinct from other items in LTM.
  • involves making an item distinct relative to other items in LTM.
• Again the example I showed you use words that were orthographically distinct from each other. But again we can apply this to different settings. For example: Bikes, Cat.
• But today were going to examine move away from encoding and take a closer look at how memory codes influence are ability to retrieve info.

• processing distinctiveness:
  • is a result of a memory code we create rather than the items ''characteristics''.
  • Caricatures
  • House Bunny

Encoding and Retrieval

• Appropriate retrieval cues
  • Eyewitness reconstruction of a crimes context
  • Differences in yes and no answers for cued recall of elaborated sentences.
    • The small lady angrily picked up the red ( ).

Although Craik and Lockharts original LOP (levels of processing) theory had very little to say about retrieval, it focused mainly how words are coded.
However, the experiments we have been looking at certainly to provide us information concerning encoding and retrieval.
We saw previous lectures that having eyewitnesses recall the context of a crime can help enhance recall.
Talked about simple, medium, and complex, elaborate sentences.
Ss said yes when the word fit the sentence and no for words that did not.
And they found that recall was better for words that they had responded yes to.
Well Craik and Tulving argued that the reason for this was that having an elaborate framework in which to encode the word only works if the word fits the context.
Suggests that under certain conditions some retrieval clues will be better than others.

Encoding Specificity Principle

• The specific encoding operations performed on what is perceived determine what is stored.
• And what is stored determines what retrieval cues are effective in providing access to what is stored. (Tulving & Thomson, 1973, p. 369)
• Studied: STRAWBERRY - JAM
• Recognition cue: GRAPE - JAM /or/ TRAFFIC - JAM
  • Grapeの方がより簡単に思い出せた。

For example, as we saw yesterday when Ss studied Straw-jam recall was better when Ss were cued with grape rather than traffic.
Because grape-jam and traffic are from different semantic categories while strawberry-jam an grape- come from the same semantic category.
Grape-jam has a closer association to strawberry jam.
Using appropriate cues.
Tatiana used a word that began with the suit of the card, and used the remaining letters to as a separate recall device for the value or number of the card.

But this definition can be dissected into two parts

• The first part
  • very much like the LOP (levels of processing).
• The second part states…
  • Use same encoding strategy and see different recall based on recall strategies.
  • So a cue that is effective in one condition may not be as effective in another condition.

• States that memory traces differ in durability and content.
• The content of the memory trace determines which retrieval info. facilitates recall.
  • ex: mood dependent memory
    • People seem to have better recollection when their mood during retrieval matched their mood during encoding.
    • 悲しいときにencodeした場合は、悲しいときの方が思い出しやすい。

Fisher & Craik (1977)

記憶成績は符合化情報と検索情報の一致の程度だけでなく、符合化情報の形によっても決定されるということをFisher & Craik (1977)は指摘する。
彼らは意図的記憶手続きを用い、符合化情報もしくは検索情報として意味的情報と音韻的情報を与える実験を行って両者の組み合わせが手がかり再生に及ぼす効果を検討している。
その結果,手がかり再生率全体としては、符合化情報の主効果が認められ、意味的情報化音韻的情報よりも手がかり再生率が高かったが、符合化情報と検索情報の一致度が高いほど再生成績がよいということも明らかになった。
彼らは、これらの結果から、符合化時の処理の深さと符合化時の文脈が検索時にどの程度回復されるかということの両方によって、記憶成績が決定されると述べている、つまり、記憶に関する自称を理解するためには、符合化時に深い処理をすればするほど記憶成績がよくなるという処理水準説と符合化特定性仮説の両方が必要だというのである。
言い換えれば、符合化情報の型及び符合化時と検索時の文脈の一致度の両要因を考慮することの必要性が示されたわけである。
(偶発記憶に及ぼす符合化情報と検索情報の組み合わせの効果)

• Manipulated both L.O.P. and encoding specificity.

Associated with sleet?
or	HAIL		Encoding
Rhymes with pail?

• Test:

Identical	associated with sleet?
Similar	associated with snow?		Retrieval
Different	rhymes with sail?

Now lets look at how the encoding specificity principle applies to the different semantic and phonemic processing levels.

• Deep semantic processing produces better recall than acoustic shallow processing.
• Interaction between encoding and retrieval.
• Effectiveness of retrieval cue depends on how the word was coded.
• The specific encoding determines the best retrieval cues.

The many factors of effective retrieval

• Context dependent memories
• Godden & Baddeley (1975)
  • IV: location of encoding / retrieval (underwater vs. on land)
  • DV: + of words recalled
  • 地上で覚えさせられた人は地上で、地下で覚えさせられた人は地下での方がよく思い出せた。

• state-dependent memory:
  • better memory when encoding & retrieval occur in the same state.
• Eich (1975) Marijuana study
  • Study list of words high or not high.
  • Recall lists of words high or not high.

• Results

			Encoding (study)
		Normal cigarette	Marijuana cigarette
test(recall)	Sober	(match)25%	(mismatch)20%
	joint	(mismatch)12%	(match)23%

• encodingの時に飲んでた人は、テスト中にのんでた方がいい点数。

• Mood-dependent memories
• Eich, Macaulay & Ryan (1994)
  • IV: mood at encoding/ retrieval
  • DV: + of events recalled

• Other contextual factors:
  • Background music
  • Posture
  • Wall colors

Transfer-Appropriate Processing

• The value of a particular learning strategy is relative to the particular goal
• Encode information based how you want to retrieve it
  • Multiple-choice tests: details
  • Essay Test: organization

• d’Ydewalle & Rosselle (1978)
• Subjects told they would get a multiple choice test or an essay test.
• Better performance when Ss received the test they expected.
• Each group used a different learning strategy.
• どんなテストのタイプかを教えられている方がいい点数。

• problem-oriented acquisition:
  • Encoding to help with problem-solving
  • Encoding material in a manner that is helpful for its later use in solving problems (CP 150)
  • framing affect
• fact-oriented acquisition:
  • Encoding to help with the recall of facts.
  • Encoding material in a manner that emphasizes factual knowledge without emphasizing its application (CP 150)

Another distinction that can be made between different test questions are those in which the instructor asks you to recall factsand those in which you need to apply a concept to solve a problem

Adams et al. (1988)

• Present Problem:
  • Uriah Fuller, the famous Israeli super-psychic, can tell you the score of any baseball game before it starts. What is his secret?
  • A man living in a small town married 20 different woman in the same town. All are still living, and he has never divorced one of them. Yet he has broken no law. How?

• Results
• Reading the answers to the problems before hand did not help.
• Recall =36%
• Subjects encoded for fact retrieval not problem-solving
• People had encoded the statements as facts and therefore it did not help with problem solving.

• Alter the statements to promote problem solving.
  • Can a baseball game have a score different from 0-0…..before it starts?
  • Is it o.k. to give drugs to a lot of people…..if you’re a doctor
  • Is it possible to marry several people each week…. if one is a minister.
• re-framing a problem in such a way, promote problem solving

• Results
• Framing the sentences to encourage problem-solving increased recall.
• Recall = 56%
• Subjects now encoded for problem-solving retrieval not facts.

People had encoded the statements as facts and therefore it did not help with problem solving.

Using encoding specificity

• Transfer appropriate processing
  • Encoding material in a manner related to how the material will be used later.

Feats of Memory

• Is it just the memory strategy or are there other factors involved?
  • deep processing vs. shallow processing
• Transfer-appropriate processing
  • fact-oriented acquisition
• Context dependent memory
• State dependent memory

2525

認知心理学もくじ

• 認知心理学もくじ