Week 7 - Attention in Space and Time

psychological function of spatial attention

allocate our limited processing capacity to important stimuli in environment - 2 processes

1. locate relevant stimuli

2. process and identify stimuli

laboratory analogue of cheetahs-in-the-savannah

find M in small set size of letters vs large set size and measure mean response time

• response time decreases as set time increases

• used letters bc easy to program

• easy to quantify similarity by number of common features

pop-out effect

a unique visual stimulus immediately captures attention in a field of otherwise familiar items = target stands out esaily without deliberate searching

pop-out effects in lab

• pop-out targets = little or no change in search time with set size

• non pop-out targets = large changes in search times with set size

feature targets

item in visual search task defined by single, distinct characteristic = stands out from surrounding distractors

parallel search for feature targets

target differs from distractors by single visual feature = brain detects quickly

• reaction time doesn’t increase with display size

• target present and target absent trials take the same time

• compare contents of each location with mental rep of target at same time = reject distractor and locate target

conjuction targets

object in a visual search task defined by a specific combination of two or more features = don’t stand out

• e.g. red vertical bar among red horizontal, blue vertical and blue horizontal lines

effect of conjunction targets on pop-out effect

reaction time increases with display size

• slop twice as steep for target absent than target present trials

• target absent trials take longer

= serial search

serial search

need to focus attention on individual items = items examined one at a time

• stop when target is found

• searched half of display in target-present trial

• all of display searched in target-absent trial

• constant scanning rate predicts 2:1 slope ratio

feature integration theory - treismand and gelade 1980

role of attention is to bind features into perceptual compounds

• each feature (colour, shape) registered in its own feature map

• attention brings these separate features together into unified perception of object

• without attention = illusory conjunctions

illusory conjuctions

brain incorrectly combines features from multiple separate stimuli into a single, unified object

feature integration theory - feature vs conjunction targets

• conjunction = targets require feature binding = need to focus attention = serial search

• feature = don’t need feature binding = don’t need to focus attention = parallel search

problems with FIT - enns and rensink 1990

• pop out sometimes depends on complex properties not just simple features

• high level not low level properties predict pop out

• inconsistent with idea that pop-out only occurs at level of simple features

Wolfe 1998 - 1 million trials for visual search

search slopes from 2500 sessions = slopes very variable

• no evidence of 2 separate groups of slopes representing purely parallel and purely serial earch

• parallel and serial are opposite ends of continuum

• Wolfe suggests better described as inefficient or efificient search

guided search theory - wolfe, cave and franzel 1989

2 stage theory

• initial stage = candidate list of possible targets

• 2nd serial stage = checks candidate list for targets

• efficiency depends on similarity of target and distractor

• predicts range of search slopes

effects of similarity in guided search theory

• similar targets and distractors = large candidate list = inefficient search

• dissimilar targets and distractors = small candidate list = efficient search

guided search 2.0 wolfe 1994

• guided search 1.0 = controlled by similarity btwn targets and distractors

• 2.0 = controlled by priority map - depends on salience of stimuli

guided search 6.0 Wolfe 2021

scene guidance - real world images rather than isolated elemnts = meaning maps not just salience

The stroop effect 1935

1st demo of a focused attention failure

• name colour of ink the word is written = measure reaction time

• fast with compatible e.g. word red in red font

• intermediate with neutral e.g. rain in blue font

• slow with incompatible e.g. word blue in yellow font

why does focused attention fail in stroop effect

automatic processing of irrelevant attributes

• parallel processing of colour naming and word reading

• word reading = fast and involuntary

• word name available before colour name = creates output interference

• asymmetrical = ink colour doesn’t interfere with word reading

output infterference

information that has already been processed competes with or interferes with the response you are trying to produce

automaticity

processes that become fast, efficient and largely involuntary through practice and learning

• basis for skill acquisition

what makes a process automatic

learned stimulus-response associations

• repeated practice = stronger association = respond quickly without

controlled and automatic processing - shiffrin and schneider 1977

searched for target items among distractors in rapidly presented sequences

• find digit among letters or letter among digits. 2 variables

• manipulated target memory set, display size of distractors and mapping types

mapping variables shiffrin and schneider 1977

• consistent mapping = roles of targets and distractors remained constant thruout trials - letters always target and numbers always distractors or vice versa

• varied mapping - target and distractor categories changed across trials = 1 trial letter is target next is distractor

results from shiffrin and schneider 1977

• consistent mapping = performance became automatic with practice independent of memory set and display size = high accuracy - participants reported that targets pop-out

• varied mapping = never became automatic

= automatic requires consistency of target set membership

The Eriksen Flanker Task

• participants shown central character surrounded by other characters (flankers)

• if > respond right and if < respond left

• compatible = > > > > = shorter reaction time and higher acuracy

• neutral = ++>++ = baseline

• incompatible = < < > < < = longer reaction time and lower acuracy

zoom lens model - eriksen and st james 1986

• use large spotlight to locate central target = zoom out

• use narrow spotlight to identify target = zoom in

• gradual narrowing of spotlight while target still being identified

influence of flankers on zoom lens model

• strong influence early in processing bc attention initially covers wider area

• as attention narrows influence of flankers decrease

attentional blink

occurs when we temporarily struggle to detect a second target because our attention is still processing the first target

rapid serial visual presentation (RSVP) task

for attentional blink - items are presented very quickly, one after another

• each item displaued for 100ms

• each item masked by following item

• participants must identify 2 targets

• T1 = identify white letter, T2 = detect whether X appeared

• measured T2 accuracy depending on time lag btwn T1 and T2

results of RVSP task

• T2 performance declines and then recovers = attentional blink = lag 1 sparing

• worst performance not immediately after T1 but some time later

• blink occurs only if T1 is processed - if ignored no blink

why does attentional blink occur

• nontarget rejection is fast

• decision process that identifies targets is slow

• processing T1 temp closes attentional gate = unaware of other targets

• once T1 processed = agate reopens = T2 can be detectd more easily

why does lag 1 sparing occur

decision process for T1 takes time to begin

• if T2 appears immediately after T1 = may be processed before attentional gate fully closes = escape attentional blink

where does attentional bottleneck occur

Wolfe guided search 6 - when info is selected and transferred into visual working memory

• VWM has 4-5 item capacity but items enter serially at rate 1 item per 50-60ms

• decision process - asynchronous diffuser evaluates mulitple items simultaneously once availble

• full VWM = must shift attention so further items can be slected and processed

Zen Kōan

question by which you attain enlightment or go mad trying