coordinated operation of active mental processes within a multi component memory

(Sensory Memory, Short-Term/Working, Long Term Memory)

Vary in duration & capacity

Provides nice ordering, but there are problems

Computer Model: information must be translated into a useable form

Without Computer Model: limited span of conscious apprehension,

things can break through our attention, adaptive way of processing large amounts of info without using up limited capacity

sensory register

(characteristics)

information has no meaning (otherwise wouldn't need a SR)

short duration (taking in a lot of info and need to move it)

no awareness (pre conscious)

large (as compared to STM)

each modality has its own SR (five senses)

pre-categorical (no meaning)

patter recognition

(necessary qualities)

identification of complex arrangement of sensory stimuli

flexibility (same patter as 2 different things, or different pattern as same thing)

speed

accuracy

decision mechanism (how do we know we have the right pattern?)

specific pattern stored in memory

decision mechanism : LTM search for perfect match

Problems : speed (many templates), flexibility (ex. handwriting we have never seen before, don't have a template for it)

more flexible than template (can see different patterns as same thing)

decision mechanism : close enough match

prototype : abstract patters stored in LTM

problems : inadequately developed, what is "close enough"?

analysis by synthesis

(feature theory)

identification of patters by breaking down parts

decision mechanism : greatest number of overlapping features

problems : search asymetries, defining feature before hand, what is necessary to distingush pattern?

visual search tasks : easier to detect objects that are defined by single features. Confusing when objects have similar features

feature + template theory

geon : geometric icon (basic form) 36 geons, 3 geons = 1.4 billion 3 geon objects

decision mechanism : certain geons have particular relationship, way shapes/geons correspond to each other

degraded objects study

problems : same as different flexibility, face recognition, selective deficit (some patterns broken down, some taken as whole)

categorical : directional relationship (objects)

co-coordinate : (face recognition)

Cooper & Woojan (2000)

- categorical predicted better when both eyes moved

- co-coordinate predicted better when one eye moved

attention

(what happens to unattended information)

filter theories :

- early selection selects sensory info that will then be processed for meaning (attention operates before meaning is determined)

- late selection selects processed events (attention operates after meaning has been determined

early selection theory

attention like switch (on or off)

physical properties control attention (NOT controlled by meaning)

2 sources (sometimes called channels) of input

Example, one in each ear

Attend to one message only

shadowing

("Moray" Study)

(to test Broadbent Switch Theory, must know which message attention is switched to)

subject must say words as they are heard

Moray : shadowing in a dichotic listening task

Group 1 and Group 2 (one group heard name, then told to stop, other group didn't hear name)

Group 2 (heard name) stoped

inconsistant with early selection (neither group should of stopped, name shouldn't make a difference)

Sometimes people detect the meaning of infor on what should be the unattended channel

modified early selection theory

not all or none switch, gradient like volume control

allow for more than one input to be processed at a time

what happens to unattended info? gone.

Marcel's Experiment

(late selection support)

Independent Variable

Word Triplets

Polysemous word (more than one meaning) always in middle

Congruent : hand/palm/wrist

Incongruent : tree/palm/wrist

Unassociated (baseline) : clock/race/wrist

Marcel's Experiment

(late selection support)

2nd Indepdent Variable

Pattern Mask (present word briefly then show something on top)

Marcel masks second item of word triplet

Awareness of Meaning of masked word?

No

Early Selection : mask blocks the processing of sensory information

Late Selection : mask interfers with attentional selection of the meaningful processed information

Marcel's Experiment

(late selection support)

Results

NO MASK

Congruent : faster than baseline, priming

Incongruent : slower than baseline, negative priming

Unassociated (baseline)

all meanings of 2nd word are activated, attention selects one of those meanings. Selection of one meaning inhibits activation of inconsistant meanings

MASK

Congruent : faster than baseline, priming

Incongruent : faster than baseline, priming

Unassociated (baseline)

all meanings of 2nd word are activated, attetional selection is blocked by mask and so all meanings remain active to affect repsonse to 3rd word.

Capacity Theory

limited capacity

different tasks require different levels of capacity (CAN do more than one thing at a time)

attention : meaning allocating capacity

Evidence (Tyler, Hertel, McCallum, Ellis)

Anagrams (difficult & easy) + tone task

more effort = more recall, but difficult anagram = more capacity and less capacity for tone task

makes no demands upon capacity

reading automatic? Stroop shows yes (can't avoid reading when trying to say color

Kahneman/Henik prove otherwise :

words in circle and square to side of fixation point, but both are in visual field.

Object defining what you should be paying attention to

Results : less interference when ink color in square is incongruent with word (looks like early selection because object selection before meaning is applied)

objects are selected for filtering (early selection)

selection one at a time

once selected, all properties are processed automatically (late selection)

difficulty of selective attention ("invisible gorilla")

Treisman, Kahneman, Burkell

- name where gap is

- reading faster when they're separated (Precieved as one object)

- when apart (have to select object, gap selection worse)

Study Phase

- study list of city names & non-city names

- get mild shock when studying city names

Test Phase

- dichotic listening with shadowing of one message (passages in both ears)

Results

significant GSR (Galvanic Skin Response) on non-shadowed message

- shadowing not interrupted (therefore according to early selection, attention isn't switching)

- but the greater GSR to city names in non-shadowed ear indicate that the meaning of the non-shadowed material was processed

supports late selection

recognition of mono-oriented objects

orientation important for faces, but not for objects (ex. house)

prosopagnosia : inability to recognize faces specifically

an analogy

model is NOT the same as a theory

computer as model of the mind (general symbol manipulation, specific function, learning, retains knowledge)

Circulatory Problem (studying something based off of the mind to study the mind)

knowledge acquired from sensory experience

(what you are currently experiencing or know from experience)

Limitations

- don't always experience what is reality

- cannot know something unless you have previously experienced it

non-sensory experience, thinking and reasoning

true knowledge of a concept cannot be acquired empirically

abstract

(we combine rationalism & epiricism)