THE LIBRARY METAPHOR

-ANALOGUE TO CATALOGING A BOOK INTO A LARGE LIBRARY

-WHAT DO WE DO AT ENCODING
    -give it an address

-STORAGE

-WHAT DO WE DO AT RETRIEVAL
    -search and find


AT ENCODING

-ORGANIZING MATERIAL

-LEVELS (DEPTH) OF PROCESSING THEORY
DEPTH OF PROCESSING
ASSUMPTIONS:

1. PEOPLE ARE QUITE FLEXIBLE ON WHAT ASPECT OF THE INPUT THEY PROCESS; EVEN FOR THE EXACT SAME INPUT CAN EXTRACT DIFFERENT INFORMATION

2. ACCESSIBILITY FROM LONG TERM MEMORY INCREASES AS ONE PROCESSES "MORE DEEPLY' (WITH DEPTH = THE EXTRACTION OF MEANING, SUCH THAT PAYING ATTENTION TO PHYSICAL CHARACTERISTICS IS LESS DEEP THAN PAYING ATTENTION TO WHAT THE ITEMS SOUND LIKE WHICH, IN TURN, IS LESS DEEP THAN PAYING ATTENTION TO THE DEFINITION OF THE WORD)

IF ONE JUST PAID ATTENTION TO THE VISUAL "SHAPE" OF THE WORD ONE WOULD KNOW THAT IT IS A SHORT WORD, HAS A MIDDLE VOWEL ETC

IF ONE PAID ATTENTION TO THE SOUND OF THE WORD, ONE WOULD KNOW THAT IT RHYMES (SOUNDS LIKE) SOME OTHER WORDS, SUCH AS FOG, OR BOG

IF ONE PAID ATTENTION TO THE MEANING OF THE WORD, ONE WOULD ACCESS KNOWLEDGE, SUCH AS THAT IT IS A 4-LEGGED MAMMAL, "MAN'S BEST FRIEND" ETC


TESTING THE THEORY

ONE CAN MANIPULATE THE "ORIENTING TASKS" TO INDUCE PEOPLE TO USE DIFFERENT CODES TO ENCODE THE ITEMS

example

TIME 1:
present items under one of two conditions
condition #1:
WORD IS ROCKET
-DOES IT RHYME WITH POCKET

versus
condition #2
-WORD IS ROCKET
-DOES IT FLY

HAVE A LONG LIST OF SUCH ITEMS

TIME 2
test for memory using a recognition memory task

PICK THE WORD PRESENTED EARLIER
ROCKET MISSILE POCKET

QUESTION: IF ORIENTING INSTRUCTIONS INDUCED Ss TO USE DIFFERENT CODES, WHAT TYPE OF ERRORS SHOULD ONE MAKE?

LET US DO THE FOLLOWING TYPE OF EXPERIMENT

3 GROUPS; EACH GET THE SAME SET OF WORDS (FOR 200 MSEC) BUT GET DIFFERENT ORIENTING INSTRUCTIONS

PHYSICAL: IS THE WORD PRINTED IN CAPITAL LETTERS

SOW

RHYME: DOES THE WORD RHYME WITH "HOW"

SOW

MEANING: DOES THE WORD BELONG TO THE "ANIMAL" CATEGORY;

SOW

AT A LATER TIME UNEXPECTEDLY TEST FOR MEMORY
RECOGNIZE THE WORDS THAT YOU MADE DECISION ON EARLIER

% CORRECT RECOGNITION
    PHYSICAL=60%
    RHYME=75%
    MEANING= 85%

get similar pattern if memory tested via recall, cues recall, intentionally

note predicted by how long one took to make the decision initially


INCIDENTAL VERSUS INTENTIONAL MEMORY

NOTE: RFFERS TO CONDITIONS AT THE TIME MATERIAL IS BEING ORIGINALLY PRESENTED (TIME 1)

INTENTIONAL MEMORY: Ss KNOW THAT THEY WILL LATER BE TESTED FOR THEIR MEMORY OF THE ITEMS

INCIDENTAL MEMORY MATERIAL IS PRESENTED TO Ss BUT THEY ARE NOT AWARE OF LATER MEMORY TEST (THAT IS THE TEST ITSELF IS UNEXPECTED)

USUALLY THIS INVOLVES AN "ORIENTING TASK"
-EG A COVER STORY THAT ENABLES THE RESEARCHER TO BE SURE THAT THE MATERIAL IS STUDIED WHILE, AT THE SAME TIME, ENSURING THAT Ss DON'T HAVE AN EXPECTATION THAT
THEY WILL BE TESTED ON THE MEMORY FOR THE MATERIAL
-AN EXAMPLE: MIGHT BE PRESENTING A LIST OF WORDS TO-BE RATED ON SOME DIMENSION (SUCH AS EASE OF FORMING A MENTAL IMAGE)... Ss BELIEVE THAT THE TASK IS JUST A RATING TASK AND ARE THEREFORE SURPRISED WHEN AFTERWARDS THEY ARE ASKED TO RECALL (RECOGNIZE ETC) THE ITEMS


IS DEEPER ALWAYS BETTER?

ELABORATION: CAN GET Ss TO PROCESS AN ITEM TO THE LEVEL OF MEANING, BUT IN SOME ORIENTING TASKS, GET SUPERIOR MEMORY OVER OTHER "DEEP" TASKS.. APPEARS TO DEPEND ON
DEGREE OF ELABORATION, THAT IS, DEGREE THE MATERIAL IS RELATED TO OTHER MEANINGFUL INFORMATION

TRANSFER APPROPRIATE PROCESSING: IS THE INITIAL PROCESSING APPROPRIATE TO HOW TIlE MEMORY WILL BE USED (TESTED). THUS, PROCESSING FOR MEANING IS GOOD IF ONE IS TESTING FOR MEANING

CONSIDER THE FOLLOWING STUDY
TIME 1: MANIPULATE CODE
PROCESSING FOR MEANING
DOES THE ITEM USE
ELECTRICITY; RadiO		 PROCESSING FOR APPEARANCE
DOES THE WORD CONTAIN
A CAPITAL "0" RadiO

 
TIME 2: TEST FOR MEMORY
EITHER A SEMANTIC TEST		 OR   a PHYSICAL TEST

WHICH WORD PRESENTED
 
RadiO
ChaIR
BooK		 RadiO
RADio
raDIO


CONSIDER TWO POSITIONS ON "FORGETTING"

AS A CONSEQUENCE OF AVAILABILITY: THE INFORMATION IS NO LONGER IN THE MEMORY STORE

AS A CONSEQUENCE OF ACCESSIBILITY: THE INFORMATION IS IN THE MEMORY STORE BUT YOU CAN'T ACCESS (OR GET TO IT)
-SUGGESTS A SEARCH PROCESS

THERE IS EVIDENCE FOR THE ACCESSIBILITY POSITION
CONSIDER THE FOLLOWING:
-THE TIP OF THE TONGUE PHENOMENON

-MANIPULATION OF RETRIEVAL CUES
EXAMPLE:
TIME 1: PRESENT A CATEGORIZED LIST (OF THE SORT USED IN CLUSTERING STUDIES): EG., A LIST OF 80 WORDS, MADE UP OF 4 WORDS FOR EACH OF 20 CATEGORIES (NOTE THE CATEGORY NAMES ARE NEVER PRESENTED):

TIME 2
ASK FOR FREE RECALL:

TIME 3
GIVE CUED RECALL; THAT IS THE 20 CATEGORY NAMES

FIND MUCH HIGHER RECALL UNDER THE CUED RECALL CONDITION

WHY DO RETRIEVAL CUES WORK?

TULVINGS ENCODING SPECIFICITY THEORY

A RETRIEVAL CUE IS EFFECTIVE IF AND ONLY IF IT WAS GENERATED AT ENCODING

THAT IS
AT TIME 1
INFORMATION IS ENCODED IN A CERTAIN CONTEXT

AT RETRIEVAL: CUE REINSTATES THE CONTEXT WHICH PERMITS ACCESS TO THE INFORMATION ONE IS TRYING TO REMEMBER

OTHER EXAMPLES OF CONTEXT-SPECIFIC MEMORY
-STATE DEPENDENT MEMORY
-MOOD AND MEMORY
-ENVIRONMENTAL CONTEXTS

EXAMPLE WITH GLOBAL AMNESIA Ss

PHASE 1; PRESENT HOMOPHONES IN A CONTEXT THAT BIASES TOWARDS ONE OF THE TWO MEANINGS

EXAMPLE:
I WILL ASK A QUESTION ABOUT A SET OF OBJECTS. THE TASK IS TO ANSWER WHETHER A GIVEN ANSWER IS TRUE OR FALSE

IS PART OF SOME MUSICAL INSTRUMENTS == REED
IS A TYPE OF METAL==STEEL
etc

PHASE 2 (AT A LATER TIME)
EITHER AN EXPLICIT TEST: RECALL THE "ANSWERS"

OR AN INDIRECT TEST (SPELLING)
READ OUT LOUD A SET OF WORDS; Ss ARE TO SPELL EACH WORD

HOW IS THE WORD SPELLED (EXAMPLE: REED/READ; STEEL/STEAL)

FINDINGS:
ON THE DIRECT EXPLICIT TEST:
NORMAL CONTROL Ss PERFORM WELL; AMNESIACS HAVE NO MEMORY AT ALL FOR THE ITEMS

ON THE INDIRECT IMPLICIT SPELLING TASK

BOTH NORMAL AND AMNESIACS SPELL THE WORDS WITH THE BIASED MEANING


WHAT PRODUCES THE DIFFERENCE BETWEEN EXPLICIT AND IMPLICIT MEMORY TESTS

ONE EXPLANATION:

EXPLICIT MEMORY IS CONCEPT-DRIVEN; DEPENDS ON THE KNOWLEDGE BASED INTERNAL PROCESSES INVOLVED IN GIVING AN ITEM AN "ADDRESS" IN PERMANENT MEMORY AND
THE CONTEXT-DRIVEN PROCESSES RESPONSIBLE FOR FINDING IT AGAIN

IMPLICIT MEMORY IS DATA-DRIVEN: DEPENDS ON PERCEPTUAL PROCESSES UNIQUE TO THE APPEARANCE OF THE ITEMS

CONSIDER A STUDY SUCH AS:
 
PHASE 1
NO CONTEXT
(READ WORDS ALOUD)		 VERSUS CONTEXT
(GENERATE OPPOSITES)
example of one such item
xxx-COLD		 hot-???
PHASE 2: (TEST FOR MEMORY)

RECOGNITION; WAS COLD ON THE LIST?
-IN THIS CASE PEOPLE RECOGNIZE FAR MORE ITEMS IN THE GENERATE THAN IN THE READ CONDITION (GENERATION EFFECT)
OR
PERCEPTUAL IDENTIFICATION;
FAST PRESENTATION: WHAT WORD WAS PRESENTED?
-BETTER PERFORMANCE N THE READ THAN GENERATE CONDITION

WHY DO RETRIEVAL CUES WORK. ..AN ALTERNATIVE POSITION
GENERATE-RECOGNIZE MODELS

RECALL IS A TWO STAGE PROCESS; FIRST YOU HAVE TO GENERATE POSSIBLE "ANSWERS" AND FROM THOSE ANSWERS YOU HAVE TO RECOGNIZE THE CORRECT ONE

CUES WORK BY LIMITING THE NUMBER OF GENERATED ANSWERS TO HIGHLY PROBABLE CORRECT

THUS, IF ONE HAS AN ITEM SUCH AS "ROBIN" ON A LIST TO REMEMBER

AND ONE GETS THE CUE BIRD.. A PERSON GENERATES PLAUSIBLE (HIGH ASSOCIATE) RESPONSES, OF WHICH ROBIN IS LIKELY TO BE ONE..

RECALL= P(CORRECT GENERATION) x (CORRECT RECOGNITION, GIVEN CORRECT GENERATION)

THUS FORGETTING CAN OCCUR IF A PERSON FAILS TO GENERATE THE SEARCHED FOR ITEM, OR IF THE ITEM WAS GENERATED BUT A PERSON FAILS TO THEN RECOGNIZE IT AS WHAT THEY ARE LOOKING FOR

NOTE: THIS IS A CONTEXT-INDEPENDENT POSITION;
A RETRIEVAL CUE WORKS IF IT HAS A HIGH PROBABILITY OF GENERATING WHAT ONE IS SEARCHING FOR, REGARDLESS OF WHETHER IT WAS OR WAS NOT AROUSED AT ENCODING


EXPLICIT (DIRECT) VERSUS IMPLICIT (INDIRECT) TESTS OF MEMORY

CONSIDER DOING A RECALL OR RECOGNITION TASK: THIS IS A DIRECT TEST OF MEMORY IN AS MUCH AS IT INVOLVES THE CONSCIOUS KNOWLEDGE THAT ONE IS INQUIRING ABOUT SOME EXPERIENCE FROM THE PAST
TO REMEMBER THE INFORMATION PRESENTED AT TIME 1 YOU HAVE TO (1) REMEMBER BEING PRESENTED INFORMATION AT TIME 1 (2) EMPLOYING PROCESSES THAT WILL ALLOW YOU
TO ACCESS THIS INFORMATION

BUT WHAT IF WE TAP MEMORY QUITE DIFFERENTLY. NAMELY TESTING TO SEE IF INFORMATION PRESENTED AT TIME 1 INFLUENCES PERFORMANCE AT TIME 2 EVEN IF THE PERSON IS NOT AWARE OR CONSCIOUS THAT THE INFORMATION AT TIME 1 IS ACTUALLY BEING STUDIED

EXAMPLE
TIME 1: STUDY A LIST OF WORDS
AT TIME 2: INSTRUCT Ss THAT WE ARE MEASURING DIFFERENT PSYCHOLOGICAL SKILLS. WHAT WE ARE INTERESTED HERE IS VERBAL ABILITIES AND THIS WILL BE TESTED
SUCH AS
    -WORD COMPLETION TASK
            T_L_PH__E

QUESTION: IS THE S LIKELY TO COMPLETE THE PATTERN WITH A WORD PRESENTED AT TIME 1

SUCH AS: PERCEPTUAL IDENTIFICATION; PRESENT WORDS AT A VERY FAST RATE
QUESTION: DO Ss IDENTIFY PREVIOUSLY PRESENTED WORDS MORE RAPIDLY


AN ALTERNATIVE VIEW OF ENCODING, STORAGE. AND RETRIEVAL

BARTLETT (1932)
AND THE CONCEPT OF MEMORY SCHEMATA

-READ THE "WAR OF THE GHOSTS"
-REPRODUCE IT
        -note: PEOPLE CONDENSE IT, MAKE IT MORE MEANINGFUL BY ADDING OR DELETING INFORMATION

-HOW COULD THIS OCCUR?

BARTLETT ARGUED THAT INFORMATION IN PERMANENT MEMORY IS STORED AS STRUCTURED KNOWLEDGE UNITS (SCHEMATA); EACH SCHEMA (PLURALSCHEMATA) IS LIKE A SKELETON CONTAINING THE BASIC CENTRAL KNOWLEDGE

WHEN ACTIVATED, SUCH UNITS (SCHEMA) GUIDE WHAT WE PAY ATTENTION TO, WHAT INFERENCES WE DRAW DURING THE ACT OF ENCODING (CONSTRUCTIVE MEMORY), HOW WE STORE INFORMATION (WE INSTANTIATE-OR FILL IN THE BLANKS)
NOTE: THE INSTATIATION MIGHT BE AN INFERENCE

---GIVE AN EXAMPLE

AT RETRIEVAL: WE ACTIVATE THE SCHEMA, ATTEMPT TO RECONSTRUCT THE INSTANTIATIONS: THIS COULD LEAD TO DRAWING INFERENCES AT RETRIEVAL (WHAT IS LABELLED RECONSTRUCTIVE MEMORY)


Psych 235 Home Page