Eysenck: Attention ch.5

Cross Modal attention

in real world we combine and integrate info from diff sense modalities. Info from auditory and visual modalities are combined to faciliitate understanding of what they are saying.

Cog neuroscience findigns re 4 features of automaticity

...

Cogntiive bottleneck theory

...

Divided Attention- Dual Task Performance

...

Focused auditory attention

...

Instance Theory

...

evaluate-

...

Endogenous spatial attention and exogenous interact cross modally

1.individual voluntarily directs attention to spatial location 2. involuntary attention given to spatial location determined by aspects of stimulus there (intensity, threat, value) Directing Auditory and or tactile info to a location can attract visual attention to same place an vice versa

limitations

1.only focuses on VIS and AUD inputs 2. there is some disruption to performance even when tasks engage different modalities 3. assumes that a given strategy is used when pp perform dual tasks but performane and effort depedn on whether pp engage in serial or parallel processing 4. assumption that several tasks engaging different pools of resources can be performed simultaneously without interference, minimises propblems ass. with higher level processes of coordinating & organising demands of tasks being carried out simultaneously.

Automatic PRocessing- practice makes perfect

Shiffrin & Schneider (1977) drew theoretical distinction between controlled and automatic processes controlled- limited capacity, require attention, and can be used flexibly in changing circumstances, Automatic- no capacity limitation, do not require attention, hard to modify once learned (inflexible).

What exp did Lavie conduct?

1995- pps detect target letter x or z in on of 6 positions in a row. in high perceptual load cond. other 5 positions were occpied by non0target letters- in low, none of the positions were occupied. A large distractor letter was presented which was either incompatibleor neutral. it was predicted that nature of distractorshould effect time to id target when perceptual load is low but not high. Mcevoy et al (2007) 1. completely task irrelevant distractors (cartoon figure) interfered with performance as much as task relevant distractors (letter). 2. interfering effects of two distractors eliminated with high perceptual load. Neuroimaging? Schwartz et al (2005)- draughtboard distractors prodiced less activation in v1 v2 and v3 (related to visual processing, when there was high perceptual load. Working memory? Fockert et al (2001) pps classified written names as pop stars or politicians uner h/l working memory load conditions (remember digit string) with distraction provided by famous famous. task was affected by distractors under high working memory load (more Face activity i visual cortex). Evaluation- distractor stim. effects depend upon perceptual load (low increasor distractor impact) and load on exec control (high increases impact of distractor).

What was BBs filter model?

1. 2 stimuli presented at teh same time fain access in parallel to a sensory buffer 2. one of the inputs is then allowed through a filter on the basis of its physical characteristics- the other remians in the buffer for later processing. 3. filter prevents overloading of limited capacity mechanism beyond the filter which processes input thorougly in terms of its meaning.

Evaluation

1. Corrects tradtional implicit assumption that att processes in each sense modality operate independently of those in all other modalities. 2. cross-modal effects reearch is limited in that it has focuse too heavily on effects in attention at expense of effects in stimuli and object identification so we do not know how different modalities combine to facilitate object recognition. 3. lots of empirical findings but weaker on theoretical understanding- so predicting strength of a particualr cross modal effect is difficult. 4. more research in naturalistic conditions away from artificial tasks needed.

How do we get beyong concluding that there are associations between individual expectations about imminent stimuli and and activation of a goal directed system to a causal statement?

1. Ruff et al., pps decided which of 2 stimuli had greater contrast. When TMS was applied to ventral system it produced predicted effects on patterns of brain activation for several visual areas, and on perceptual performance strengthening the case for claiming that the ventral system influences attention in a top-down fashion. 2. Stimulus driven system is typically damaged in Visual neglect patients supporting corbetta et al. 3. Folk et al, 1992- evidence that stimulus driven attention is captred more by distractors resemling task relevant stimuli than by salient distractor stimuli. They used targets and distractors defined by colour or abrupt onset. When pps looked for abrupt onset targets, abrupt onset distractors captured attention but colour distractors did not. When pps looked for colour targets, colour distractors captured attention but abrupt onset distractors did not. 4. Indovine & MAcaluso, 2007- fMRI assessed the effects of different distractor types on activation of stimulus driven sustem or ventral network. pps reported orientation of a coloured T in presenec of T in different colour (task relevant distractor) or flickering draughtboard (salient distractor). Ventral network was activated by task relevant but not salient distractors.

Why is attentuation theory more likely than Deutsch and Deutsch late selection theory?

1. Treisman and Riley (1969)- pps shadowed one of two auditory messages but were told to stop shadowing and tap when they detected target in either message. There should be attentuated processing on non-shadowed message and so fewer targets should be detected on that message. This is what was found- Acc. to D&D there should be no difference because there is perceptual analysis of all stimuli. 2. Neurophysiological studies provide evidence against D&D. Coch et al., 2005, ERPs 100ms after probe presentation were greater when probe was presented on the attended message than the unattended one in a dichotic listening task in which pps attended to one of two auditory messages an had to detect probe targets in attended/unattended message- more processing of attended than unattended. D&D woul predict no difference.

Main assumption s of Feature integration Th. (Treisman and Gelade, 1980)

1. distinction between features of objects (colour, size, lines) and objects 2. features of objects in env are processed without attention (rapidly, in parallel). 3. followed by Slower Serial process of feature combination 4.features can be combine by focused att. to location of object (att. acts as glue binding unified object percept). 5. feature combination can be influenced by stored knowledge 6. in absence of focused att. features are combined randomly producing illusory conjunctions. Predicts?- Fit predicts that illusory conjunctions occue because of problems combining features to form objects at a late stage of processing.

what are limitations of this resaerch In real world visual search?

1. distractors are heterogeneous 2. targets are very rare in imp situations such as airport security checks (to explore this Wolfe et al 2007 presented pps with xrays of bags and targets were weapons (knive or guns). when targets a[ppeared on 50% of trials, 80% of them were detected, however when targets appeared on 2% of trials (more ecologically valid), detection rate fell to 54%. 3. RT measures are limited because they can vary with differences in discriminability, processing speed, or unknown mixtures of the two effects therefore speed-accuracy trade-off by McElree and Carrasco is an improvement.

What limitations does attentional resaerch suffer?

1. focuses on external environment rather than attention to the internal environment 2. what we attend to in the real world is driven by our current goals and emotional states, however attention research often results in the experimenter instructing and thus determining what is attended to.

What are the limitations of BBs filter model

1. idea that unattended message is rejected early is dubious- Underwood (1974) found naive pps detected 8% of digits on non-shadowed ear compared to experienced researcher (67%) 2. Allport, Antonis and Reynolds (1972) degree of similarity between 2 messages has a major impact on memory for the non-shadowed message-when shadowing of auitorily presented message was combine with picture presentation, memory for pictures was good. so disimilar inputs can be processed more fully than in BBs model 3. Early studies assumed that there was no processing of meaning of unattended message because pps had no conscious awareness of hearing them, however, Von Wright, Anderson, and Stenman (1975) meaning can be processed without conscious awareness- GSK to words previously conditioned to associate with electric shock, same with words that souned similar to conditioned words- indicates that unattended message was processed for meaning even though pps were not consiously aware. 4. Moray (1959) found that pps responded to their own name 33% of the time. This probability of detecting one's own name depends on ind. differences in working memory- LWM 65% HWM 20% (COnway, Cowan and Bunting (2001) suggesting HWM reflects ability to control focus of attention and ignore unattended message.

Evaluation of BB filter model

1. inflexible 2. assumes filter selects info on basis of physical features but Gray and Wedderburn (1960) show that it can select on meaning- who 6 there, 4 goes 1- presented ear by ear, but preferred order of report was determined by meaning which is oppsite of what Cherry found with digits- this is inconsistant with filter theory. HOWEVER recent support of BBF model comes from LAchter et al 2004.

What does GS th. state?

1. initial processing of basic features produces activation map with each item in vis. display having its own level of activation. 2. Attention is directed to items on basis of their level of activation starting with most activated. 3. this explains longer search times when distractors share one or more features with targets. Crit of FIT- there should be no effect of set size is early processing is parallel, and substantial effect is it is serial- findings fall between these extremes.

Support for Multiple Resources-

1. more interference when 2 tasks share modality (Treisman and DAviesm 1973) or 2. demand similar response (McLeod 1977). 3. brain imaging resaerch indicates that tasks very different activate widely separated brain areas- suggesting ifferent resources.

What factors determine dual task performance (DTP)?

1. task similarity- Treisman & DAvies (1973): in monitoring 2 tasks, interference is greater when stim. are in same sense modaility. McLeod (1977) performance worse with high response similarity. tracking task with manual responding compared with tone id task. some pps respond vocally and some respond with hand not involved in tracking task (this cond. slowed performance & produced errors). 2. practice- Spelke et al (1976) 2 students wth 5 hours training a week for 4 months. T1 reading short stories for comp whilst writing words to dictation was initially hard but after 6 weeks they could read as rapidly and with as much comprehension when taking dictation. T2 Also they wrote down names of categories to which dictated words belonged whislt maintaining normal reading speed & comp. Crit: focused on accuracy rather than speed. 3. task difficulty - sullivan (1976)- shadowing (repeating aloud) an auditory message and detecting targets n a non-shadowed mesage. -when Shadowing task made harder by using less redundant messages fewer targets were detected on non-shadowed message. Segal & Fusella (1970)- performance determined more by similarity than difficulty

There are limitations to Corbetta and Shulman's findings what are these?

1. tasks used to id goal-directed attentional system differed from tasks used to id stimulus-driven system. 2.most studies considered one attentional system so failed to provide comparisons of patterns of brain activation in 2 systems.

What majoy attentional systems are involved in Focused visual attention?

2 major systems invovled in visual attention. 1. voluntary, endogenous, goal directed- involved when peripheral cues were presented 2. involuntary, exogenous, stimulus driven- involved when uninformative peripheral cues are presented. Stimuli salient in colour' motion etc, are most likely to be attended to via this system.

Decision Integration Hypothesis- Palmer et al (2000)

Argued in contrast to FIT that 1. parallel processing is involved in both single feature and conjunctive searches. 2. observers form internal representations of target and distractor stim, and visual search involves decision making based on discriminability between target and distractor items regardless of whether targets are single or madee of conjunctive features). 3. conjuncion searches aare harder because their is less discriminability between target and distractor stim. 4.visual search is slower for large set sizes because the decision making process is more complex.

What alternative theories are there Deutsch and Deutsch (1963)

All stimuli are fully analysed with most important determining the response- bottlenech is nearer the response end of processing system thatn in Treisman's theory.

What is the complex FIT that Treisman (1993) proposed?

4 kinds of attentional selection can operate at various levels depending on task demands- 1. sel. by location, 2. sel. by features (surface ef. features, and shape def. features), 3. selection on basis of object-defined locations, 4. late selection that determines object file controlling ind. response.

Evidence for multiple spotligts theory?

Awh and PAshler (2000) five by five displays containing 23 letters and 2 digits- identity of two digits reporte, just before display was presented, pps were given two cues indicating probable locations of two digits (invalid on 20%) crucial condition ws one in which cues were invalid- with one of the digits being presented in between cued locations- performance was much lower on this condition supporting multiple spotlight th. rather than zoomlens which would have given digit full attention as it would have been contained in its beam.

What evidence did BB cite?

BB 1958 cited the findings from Cherry's shadowing task, and data from a memory task in which 3 pairs of digits were presented dichotically (3 digits one after the other to one ear at the same times as three different digits were presented to the other ear. PPs mainly recalled digits ear by ear rather than pair by pair.

Focused visual attention- why has most rearch focused on visual attention?

BEcause auditory attention has less of brain devoted to it, it is easier to control visual stimuli, more issues can be explred in visual attention.

Central capacity and multiple resources (SYNTHESIS?)

Baddeley argues for synthesis- proposes a hierarchically structured processing system with central exec (att. control) at top involved in coordination and control of beh, an specific processing mechanisms at bottom- phon loop, visuospatial sketchpad.

ventriloquist illusion

Bonath et al (2007)- ERPS and fMRI showed that this illusion involves processing within auditory cortex matching the apparent visual source of sound (therfore vision dominates sound) perhaps because location of env events is normally indicated more preceisely by visual information so we rely more on vision.

Evidence that FIT occurs late in processing- Brain imaging?

Braet & Humphreys (2009)- TMS administered at intervals after onset of visual display- more illusory conjunctions when TMS applied relatively late but not early.

How is focused attention (selective attention) studied?

By presenting individuals with 2 or more stimulus inputs simultaneously and instructing them to respond to only one- enables us to study the fate of unattended stimuli DICHOTIC LISTENING TASK

What is underadditivity and how has cog neuroscience finding of underadditivity shed light on divided attention and support central capacity?

Common sense says that the emands on two tasks performed together equal the sum of the demands of the two tasks performed seperately. Underadditivity is when brain activity in dual-task conditions is actual less than total activity in the two tasks performed on their own. Just et al, 2001- pps performed 2 tasks (1. aud comp, 2. mental 3D shape rotation to decide if figures matched) on assumption that different brain areas would be involved. RESULTS? 1. performance on both tasks was impaired under dual compared to single task condition, 2. language task activated temp lobe and mental rotation act. parietal lobe, 3. brain activation in language associated regions decreased by 53% in dual task compared to single, mental rotation parietal areas decreased by 29% in dual task. Thus brain activity in dual tasks is less than actibity of tasks performed alone- Underadditivity.

When is performance faster?

Duncan & Humphreys- 1. when there is similarity among distractor stim. and 2. when the target and distractors are similar.

Location and object based?

Egly et al (1994)- behavioural evidence- RT studies- task- detect a target stim. ASAP. Cue presented before target was either invali or vali. Invalid cues were in same object as target or different object. detection was slower on invalid trials. Neuropsychological evidence- same displays presented to right parietal area brain damaged patients. When cue was presented to same side as damage but target was presented to opposite side patients showed slowed detection.

Spotlight or Zoom lens, or multiple spotlights?

Eriksen and St James (1986) developed spotlight into Zoom lens. Whereas spotlight states that visual attention illuminates a small part of visual space with little seen outside its beam, they argued that we can increase the area of focal attention at will just as a zoom lens can be adjusted.

Evaluation

FIT has influenced thinking on early stages of sensory encoding to higher order characteristics of att. control, but its limitations are: 1, conjunction searches do not typically involve parallel processing followed by serial search. 2. search for conjunctions are faster than predicted by theory (e.g. grouping of distractors; distractors sharing no features with target) are incorporated into GS TH. 3. set size effects o not only depend on nature of target (single feature or conjunctive feature) but the nature of distractors (similarity to each other). 4 FIT predicts that att. deficits of neglect and extintion patients should disrupt search for conjunctive but not single feature targets, however patients detect both types more slowly than controls but impairment is greater with conjunctive targets.

Evidence from Brain damage?

Friedman-Hill (1995)- brain damaged patient who had problems with accurate location of visual stim. produced many illusory conjunctions (combine many features).

How have HAhne et al (2006) rectified theses limitations?

Hahne et al (2006) tried to eliminate problems using tasks which assumed that top down processes would be used most extensively when the cue was informatiive in contrsat to bottom up processes which would occur after the target stimulus had been presented and would be used when cue was uninformative. 1. They found no overlap between brain areas associated with TD & BU processing, 2. regions overlapped iwth those of Corbetta and SHulman

CEntral capacity vs. Multiple Resources

How to explain finding that performance is impaired when tasks are paired than performed separately? 1. Central capacity can be used flexible across range of activities but has strictly limited resources- o dual task performance depends on demands each task makes on those resources (Kahneman, 1973)- so when two tasks are similar (use same central capacity, performance is impaired). 2. Wickens (1984) processing system consists of ind. processing mechanism in form of multiple resources- which explains why degree of similarity between two tasks is imp- they compete or same specific resources and produce interference. 3D structure of processing resources with 3 successive stages of processing (1. encoding= perceptual processing of stim (VIS or AUD), can involve spatial or verbal codes 2. central processing=, 3. responding= manual or vocal responses) 2 theoretical assumptions- 1. several pools of resources based on distinctions amon stages of processing, modalities, codes, and responses; 2. If two tasks use different pools there is no disruption when performing 2 tasks.

What did Posner find (1980)?

In his studies of covert att. where att, shifts to a spatial location in absence of eye movements- pps responded rapidly when they detected onset of light. Shortly before light onset they were presented with central cue (arrow pointing left or right) or peripheral cue (brief illumination of box outline). These cues were mostly valid (indicated where target would appear), but sometimes invalid (innaccurate information about location of target light). Valid cues produced faster responding to light onset than neutral cues whereas invalid cues produced slower responding than neutral cues. Findings were comparable for central and peripheral cues in absence of eye movements. cues valid on central cues they were ignored, but peripheral cues affected performance

Eimer et al (2003) also found what to conclude that cross modal effects do not depend heavily on visual modaility but that endogenous spatial attention is controlled by high level system that influences att. processes within each sense modaility.

In lighted or dim conditions visual tactile cross modal effects were very similar.

What alternative theories are there? Treismans attentuation theory-

In shadowing tasks sometimes breakthroughs occur from unattended channel when a word is highly probably in context but only sometimes (e.g. 6% of time) due to lowered thresholds of expected stimuli. Filter attenuates analysis of unattended information so that bottleneck is more flexible than in BBs model. Treisman proposed that stimulus analysis proceeds through a hierarchy-analysis based on physical cues,syllable pattern and specific words to analyses based on grammatical structure and meaning. If there is insufficient capacity for full stimulus analysis, tests at top of hierarchy are ommited.

What does Broadbents Theory account for?

It accounts for Cherry's basic finding that unattended messages are rejected by the filter and receive little pricessing. It also accounts for performance on BB's dichotic listening task. The filter selects one input on the bases of the most prominent physical characteristicsdistinguishing the two inputs (ear of arrival)

What other criticism should be levelled at trad approach?

It says nothing about how serial processing associated with controlled processing turn into parallel processing associated with automatic processing.

What evidence supports Zoom lens?

La Berge (1983) 5 letter words presented and probe requireing rapid response was occasionally presented instead or immediately after word., in the spatial position of any of the five letters. Pps had to categorise middle letter (focusing att), or whole word (broad beam). Att. spotlight can have a very narrow or broad beam. Muller et al, 2003- presented with four sqaures in a semi circle and cued to focus att. on one, two, or all four squares. objects were then presented and pps decided whether a target white circle was among them. when a target was present it was always in a cued square. targets were detected fastest when attended region was small (0nesquare)- activation in early visual areas was most widespread when attended region was large an liited when attended region was small. Supports notion of an attentional beam than can be wide or narrow.

What is selected- an area or region of space, or given object/s?

Location-based attention-O'Craven et al (1999)some evidence that location based inhibition of return is stronger. pps given two ovals of different colours ( one to left of fixation and one to right) and inicated the orientation of the one in a given colour, each oval was superimposed on a task irrelevant face or house- making use of fact that FFA is activate when faces aare processed whereas parahippocampal area is activated when houses are processed)- fMRI indicated more processing of stimulus superimposed on attended oval than stimulus on unattended oval).

Evidence?

McElree & Carrasco (1999)- RT decision measurements are limited becauase speed of performance is based on pps willingness to accept errors- 1.when speed of performance is controlled by requiring pps to respond rapidly following a signal patterns of performance were more similar for feature and conjunction search than would be predicted by FIT. 2. set size effects conjunction more than single feature search (as predicted by FIT)- but could also be explained by fac that increasing set sixe redices discriminability between target and istractor items more for conjunction searchs than feature searches. 3. set size effect on conjunction seach was smaller than expected on serial processing moels. fMRI evidence- Leonards (2000) cerebral netweks in efficient feature and inefficient conjunction search overlap completely suggesting that feature and conjunction search involve similar processes (as assumed by decision integration hyp). However Anderson (2007) found some overlap but more activation of inferiior middle frontal cortex with conjunction search than feature search because former tyep placed emands on attention?

We know that task relevant distractor stimuli have more effect thaan salience on disrupting task performance, but What Th. did Lavie develop to emphasis alternative factors?

Perceptual load theory- 1. susceptibility to disraction is greater when task involves low perceptual load than high- because high load engages full capacity in relevant processing leaving no capacity left over. 2. distraction is also greater when high load is on executive cognitive control functions (working memory). than when there is a low load ecause cog control is needed for actively maintaining a distinction between targets and distractors esp when it is har to discriminate between target and distractor stim.

How oes Guided Search Theory (Wolfe, 1998) integrate FIT?

Replaces Treismans assumption that initial feature processing is nec. in parallel and subsequent processing serial with notion that processes are more or less efficient.

How does Lavie shed light on the situation?

She argues that there is sometimes early selelction (BB, 1958) and sometimes late selection (D&D, 1963).

How does the stroop task highlight that attentional processes are actually relevant to automatic processes (contrary to the traditional approach of- e.g. Schiffrin & SChneider 1977)

Stroop often assumed to involve automatic porcessing of colour words (which slow down naming of colours in which words are priinted). However Kahneman and Chajczyk (1983) used version in which colour word was presented close to a strip of colour and colour had to be named- which reduced stroop effect compared to a condition in which colour word and colour strip were in same location.

Support for Central Capacity

Supported by 1. Bourke et al., 1996 (4 tasks- pps performed two together-least demanding task interfered least with other tasks, and task making greatest demands on central capcity interefered most). 2. Brain imaging

Moors and De Houwer are imp why?

THey argued that automaticity should be defined in terms of varous features that distinguish it from non-automaticity. So...that we can make relateive statements such as e.g. process x is more/less automatic than process y (NOT so that we can claim tat a given processis 100% automatic or non-automatic. Automaticity is ass. with 4 features not always found togeher- goal-unrelated, unconscious, efficient, fast.

What did Thornton and Gilden suggest as a way of determining if parallel or serial search was used (2007)

That first item processed in serial processing will always be a taret and so detection time should not vary as fnction of set size, but if target detection time decreases as number of targets increase, parallel processing is indicated as inds. are taking in all information from all targets at same time. using single and multiple taregt trials with 29 search tasks in which set size was 1, 2, or 4 three patterns emerged: 1. target det time increased modestly with increasing set size on single target trials but decreased with increasing set size when all stimuli were targets. found with tasks where targets and distractors only differed along a single feature- Parallel processing indicated 2. target detection times increased raplidly with increasing set size for single target trials and when all stim were taregts- found with complex visual tasks involving detection of specific direction of rotation- indicates serial processing 3. moderate increases of set sixe on target detection times with single targets and no effect when all stim were targets. found for conjunction saerch tasks where targets were defined by conjunction of features (more consistent with parallel models. Therefore some visual search tasks involve parallel (72%) whereas others involve serial (28%). parallel processing was found on tasks that were especially complex and had longest average target-detection times.

What else does Corbetta et al pose?

That task relevant stimuli are more likely to attract attention from ventral netweo than salient stimuli (dorsal) but that there are interactions between both

What did Cherry find in the COcktail Party effect (1953)?

That the ability to follow just one conversation amongst many in a hum, involves using physical differences (sex of speaker, voice intensity, location). When 2 messages in the same voice were presented to both ears at once, listeners found it hard to seperate out the two messages on the basis of meaning alone.

What was concluded from Cherry's studies?

That unattended auditory information receives almost no processing (very little memory for unattended words even when presented 35 times (Moray, 1959).

IS there evidence for exogenous system?

Using brain imagery Corbetta and Shulman, 2002- identified stimulus driven ventral network involved when unexpected, potentially important stimulus is presented. has circuit breaking function (redirecting vision from its current focus) right-hemisphere ventral fronto-parietal network

IS there evidence for endogenous syste?

Using brain imaging Corbetta and Shulman, 2002 identified a goal directed top down att. system (dorsal fronto-parietal network) resembling Posners endogenous system. Influenced by expectations, knowledge, goals. Cues to location, or motion or forthcoming visual stimulus activate this system.

Slippage or leakage?

VonWright et al (1975) found evidence for slippage- heightened GSRs to shock associated words on unattended channel. Lachter et al, (2004) tested slippage account- lexical decision task where pps decided whether a letter string was a word (a visual task). String was primed by word that was same as or unrelated to target word presented for lexical decision. Prime word was presented for 55ms, 110ms, 165ms to unattended location. PPs would need to shift att. to unattended prime to show priming effect. There was priming when prime word was presented for 110 and 165ms but not at 55ms as it takes 50ms for an attentional shift to occur therefore there was no evidence of unattended prome word processing when steps were taken to prevent slippage but not leakage.

Is BBs filter theory redundant?

We know there are sensory buffers for auditory (echoic memory) and visual modaility (iconic memory) which is in line wiht BBs proposed sensory buffer for immediate memory that briefly holds unprocessed information. BB believed it took 500ms to shift attention between buffers and so ruled out rapid switching on attention between buffers. We now know that involuntary shifts only take 50ms so that shifting attention in sensory buffer is almost as effective as shifting att to actual object. This finding means we have another explanation for semantic processing of unattended stimuli. The leaky filter of Treisman is explained by Lachter et al as slippage in BBs modified Th.

What happens to unattended visual stimuli

Wojciolik et al (1998) same different task applied to blocks of faces and houses with other stimulus type being unattended. FFA activation stronger when faces were attended to than not McGlinchlly et al (1993) Neglect patients (who ignore left visual field) decide which o ftwo drawings match previously presented drawing presented to either left or right field. patients performed well when drawing was presented to right visual field but not left. However when presenetd with letter strings that patients decided were words or not, they pperfomed better regardgless of whic visual field info was presented to. indicates that there is some semantic processing of left field visual stimuliby neglect patients

How is divided attention studied?

by presenting at least 2 stimulus inputs at the same time with instructions that individuals must repond to all stimulus inputs- multi tasking

LImitations of cental capacity

critcised by 1.response selection can cause interference-Hegarty et al (2000)- response selection is more important than task demands on central capacity (identical figures vs paper folding task coupled with random number gen. distractor task). 2. circularity- redescribes finding to say that dual task interference is caused by resources exceeding cc, and lack of interference due to 2 tasks not exceeeding those resources. 3. evidence for CC does not clarify nature of CC (Bourke et al) 4. assumes all pps use same strategies in all dual task situations whereas HUbner 2009 trained pps to use either serial or parallel processing when performing dual tasks and found that serial processing pps performed better but they found the tasks more effortful (EEG)s). 3

Multiple target visual search

focuses not on single target detection (which assumes lengthening of detection times with increasing distractor number), instead focuses on multiple target detection

Why does multiple spotlights theory provide a better account of visual attention?

it is more flexible than zoom lens- we can show 'split attention' to two or more regions of space not adjacent to eachother. saves processing resources because we avoid attending to irrelevant spatial regions. Awl and Pasher (2000)-pps shown 5by5 visual display containing 23 letters and 2 digits and reported identity of 2 digits. prior to display pps were given two cues inicating the probable locations of the 2 digits (cues were invalid 20% of trials)- on an invalid cue with one digit presented between cued locations- performance was lower for digits betwen cued locations than for digits at cued locations- thus attention can be doughnut shaped with nothing inbetween. Also imaging studies show cortical action in and close to primary visual cortex, with no activation corresponding to region in between (Morawetz et al., 2007).

Corbetta and Shulman's (2002) identification of brain areas for bottom up and top down systems?

metaanalysis of studies where pps detected low-frequency targets using ventral netweok. Brain areas in this attentional network inc. temporo-parietal junction (TPJ), intraparietal sulcus (IPs), frontal eye field (FEF) and middle frontal gyrus (Mfg).

cognitive neuroscience evidence for cross modal effects?

multimodal neurons etected that respond to stimuli in various moailities- responding strongly to multimodal stim. at given location but showing reduced responding when there is mm stim involving more than one location (Stein & MEredith, 1993), these neurons are found in close proximity to several brain areas (mid brain and ccortex). ERPS in task irrelevant sensory modaility when pps asked to attend to visual or auditory features on an object (Molholm, 2007)

What other studies did Cherry conduct?

one auditory message was shadowed while a second was played to the other ear and little information was extracted from the second non attended ear (foreign language, reversed speech). Physical changes were always detected however (e.g. a pure tone).

cognitive neuroscience evidence for executive fundtion.Why is it odd that no previous studies found activation higher activation of prefrontal cortex for dual over single task performance (ass. with executive function) when exec function clearly controls and coordinates task management via attentional control and should thus be more activated under dual task conditions?

paradoxical findings that activation in prefrontal cortex was only same level in dual tasks as single tasks, led to assumption that single tasks must promote high levels of prefrontal activation, and that if two tasks individually made minimal demands on this functioning, evidence would be obtained. JOhnson and Zatorre (2006) did this- Findings? 1. only divided att. ass. with act. of dorsolateral prefrontal cortex suggesting brain area (inv. in exec processes) is needed to handle demands of coordinating two tasks at same time, but not required for selective att. Further evidence that dorsolateral prefrontal cortex is actually REQUIRED for DT performance (johnson et al, 2007) TMS ued to disrupt dorsolateral PF cortex function ipaired ability of pps to divide att, between two previous tasks of johnson and zatorre 2006

what did Eimer and Schroger 1998 find re endogenous spatial attention

pps are presented with 2 streams of light- one to left and other to right eye, whilst simultaneously presenting 2 streams of sound one to each ear. cond 1. pps asked to detect deviant visual events (longer than usual stim) presented to one side only. cond. 2. pps have to detect deviant auditory events in only one of two streams. Result- when pps had to detect visual targets on left side, ERPs to deviant auditory stim were greater on left than right side (demosntrating a cross modal effect in which endogenous allocation of visual att, affects allocation of auditory att. when pps had to detect auditory targets on one side, ERPs to deviant visual stim on same side were greater so allocation of auditory attention influenced visual attention too.

Evidence (Treisman and GElade) ?

pps seached for a terget in dispaly of one-30 items. target was either object based on a conjunction of features (Green letter T) or a single feature (blue letter or an S). when ntarget was a green T all non targets shared one feature with target. set or display size affected etection speed when target was a conjunction of features because FOcussed Attention was required. (lack of focused att produces illusory conjunctions!).

What is inhibition of return? Does it apply to locations or objects?

reduced perceptual priority for information in a region that recently enjoyed higher priority (Samuel & Kat, 2003). IOR is location based- WHen time interval exceeds 300ms in targets in cued location they are responded to more slowly than those in non-cued location. Uner 300ms this doesn't happen (Posner & COhen (!984) List and Robertson (2007) IOR is location based- object based inhibition of return slow to emerge and small in magnitude. Location based IOR rapidly, greater magnitude, and consistently found. Leek (2003) IOR is locationand object based- in standard condition (location based plus object based inhibition) inhibitory effect was stronger than in condition where the object was absent.

exogenous spatial attention- cross modal effects

spence and driver (1996)- pps fixated straight ahead with hands uncrossed holding a small cube in each hand. 2 light emitting diodes with on elight at top and bottom. cond 1. loudspeaker placed above and below each hand close to light source. sounds from loudpeaker shortly before one of the four lights was illuminated made visual judgments more accurate when loudspeaker/auditory cue was was on asme side as subsequent visual target even though cue did not predict which light would be illuminated- thus exogenous auditory att. influenced allocation of visual att. also worked when visual and auditory modalities were reversed so that exogenous visual attention influenced alocation of auditory attention.

Visual Search

studies in which a specific target in a visual display must be detected ASAP.

How did Schneider && Schiffrin (!977) experimetnally demonstrate their distinction?

task- pps memorised up to 4 letters (memory set) and were shown visual display containing up to four letters. pps decided rapidly if any items in visual display were the same as anyu of items in memory set. Crucial manipulation was type of mapping used (consistent mapping- only consonant were used as members of memory set and numbers used as distractors in visual display or vice versa). So pps given only consonants knew that any consonant detected in visual display must be an item from memory set. varied mapping- mix of numbers and consonats was used ot form the memory set/provide distractros in visual display. RESULTS- numnber of items in visual display and memory set only affected decision speed greatly in varied mapping. which Schneider and Schiffrin interpret as a controlled process inolving serial comparisons between each item in memory set and each in visual display until a match is achieved or every comparison made. Consistent mapping on the other hand involves automatic processes operating independently in parallel, they evolve through years on practice with distinguishing numbers and letters. they are inflexible however, as although there was great improvement in a task involving 2100 trials in consistant mapping with B-L consonants (set 1) and Q-Z consonatn (set 2), where one set are always used in construction of memory set and distractors in visual isplay were all selected from other, there a further 2100 trials of reverse mappng resulted in still greatly disrupted performance levels.

What did this evidence lead JOhnson et al to speculate>?

that Dorsolateral PF cortex is needed ot manipulate info in WM.

Object based-

visual attenton is often directed to objects rather than particular area of space- pps watching two superimposed moving scenes could easily attend to one and ignore the other (Neisser and BEcklen (1973). brain imaging- O'craven et al (1999)- pps presented with face and house transparently overlapping at same location with on eobject moving slightly. pps atended to direction of motion of moving stim or position of stationary stim. if attention is location based, pps would have to attend to both stim. if object based, attended stim. should be processed more thorughtly. USing fMRI to detect activity in FFA and parahippocampal, there was more activity in FFA when face stim. was attended than unattended, comparable results for houses- thus attention was object based. neuropsychological- Persistent neglect patients (Marshall & Halligan, 1994). patient presented with ambiguous displays of white or black shapes on white or black backgrounds with jagge edge dividing 2 shapes at the centre of each display. THe patient copied this jagged edge when drawing shape on left sie, but not when drawing shape on right side sujesting patient attended to objects in space rather than region.