Paul Part 2 Flashcards
(30 cards)
Automatic processing
Fast, parallel (number can happen at once), little effort, no capacity demons (make task harder but doesn’t matter), some through practice, born with others, hard to control, stim driven (react inevitably, once start can’t stop)
Controlled processing
Slower, serial, effortful, depends on capacity but can be changed as control behaviour, need focused attention, dependent on control processes
Schneider and shiffrin background
Ps own number of items than asked to search for them in number of frames w distrsctors, say present or not. Two conditions: variable mapping for controlled and consistent for auto. Consistent is target items were never distractors e.g. numbers vs consonants, variable was target in one trial could be distractors in next. Also varied set size( total number of items), frame size (items per frame) and frame length presented (time)
Schneider and shopfront initial results
Consistent mapping had no effect on number of items, frame size - no increase in rt so no capacity demands. For variable: as no. Items and frames increased, as does rt- affected by memory and perceptual load . consistent quicker than variable so may have fast and slow processing
Schneider and shiffrin training
Consonants and numbers clearly diff so split into two groups: B-L, Q-Z. After many trials, performance became automatic, no longer influenced by task difficulty. When reversed groups, decrease in performance as slower to build back up (auto hard to learn, harder to unlearn), changes can be altered by instruction. Controlled open to conscious control but auto not , need training to auto but controlled astbalished in few trials . Despite practice, variable mapping performance reflects controlled processes
Posners spatial cueing task background
Ps watched fixation then they are cued, either central cuing which is arrow pointing to where the target will appear or peripheral which is cross exactly where target would appear. Cue either valid and informative or invalid (false) and not. Also had cue target delay (time between cue going and target coming. Dev is validity effect- diff between invalid RT and valid RT. + means invalid slowe so using cues, if 0 no diff and not following cues, if - invalid faster so focused on opposite
Posner results
Informative:valid: for peripheral, responds fast and consistent between 0-100ms, validity effect early. For central, see effect slower at 100ms to follow cues. Uninformstive: invalid: central: no validity effect, not listening to cue, for peripheral, fast, validity effect present but cues not useful so instant. Means peripheral automatic and central controlled
Posner model
Attentional spotlight focused on fixation cross, when cueing happens, spotlight/attention disengages and moves to cues spot. If valid, spotlight in target so fast rt but if not, then stages have to be related. Central slower as cue needs to be interpreted, control spotlight so can ignore but peripheral stim driven. Endogenous: controlled, effort, too down, vounjtsry, mod by goals. Exogenous is bottom up, auto attention, external events, invol, attention
Posner parading
Inhibition of return: as increase diff between cue and target, for auto see reverse validity effect. If cue invalid, quicker than valid, means cued targets longer than uncuedauto. Attention drifts from cue as invalid. For peripheral invalid, see validity immediately but longer at longer delay when uncued are quicker as attention has drifted and slower to come back. But don’t know why, some say attention leaves when doesn’t serve you, auto response
Other cues- frisesen and kingston 98
Valid or invalid cues using emoji faces w eyes gazing in different sides, all central. Same effect- eye gaze auto cue even when invalid, quicker to cue valid tho (early validity effect). Driver- used real faces, ps had to say if target T/L, validity effect only later on/700ms but harder task and faces may capture attention so primed face first then did gaze, and found early validity effect. When switched to 80% invalid, still validity effect early then counter at 700ms, ps could then control focus away . Auto first then controlled
Risti 2002
Validity effects for eye gaze and arrows later on (central) , even w young children, evidence of auto for central cue. Arrows have strong social sig, but no inhibition of return for arrows or eyes so not peripheral
Dual tasks background
Diff in performance between task on own and with another is dual task decrement. Structural early selection say due to task switching. Capacity saybcapaicity
Dual task in car - talking
Strayer 2003: car simulation: first just driving then hands free convo w confed. Also either high or low density traffic. In dual: ps slower to break, breaker for longer and took longer to slow in both traffic, also less id of items in memory test. Real car convo: 12% fail parking compared to 50% when talking hands free
Dual task in car- texting
Drews 2003: when driving and texting, slower breaking, more varied following distances and failure to maintain lane. Worse than hands free e.g. dangerous compared to poor driving. May bc hands free and driving diff modalities but texting also using hands
Hyman 2010- dual task walking
Compared walking and talking on phone w walking when listening to music, alone and walking in pairs. On phone more changing direction and weaving, don’t acknowledge others and missed big events e.g. unicylcing clown
Reason for multitasking fail - workload models- from kahneman capacity model
Focus on demands of task/resources. E.g. piloting has multitasking in hierarchy as first maintain stable then navigation then routine checks then listening to coms
Wicken model of multitasking - background
Separate resources split into different areas. Also stages of processing e.g. how far down cog the process is. Bandwidth split into diff processes: perception, thinking and responding. Also diff codes processing e.g. spatial vs verbal. Also diff modalities like visual and audio. E.g. if two tasks both visual do worse than one visual one auditory
Wickens 2003
Either had normal instruments in front of pilot w verbal comm from ATC or instruments plus text updates from ATC. Found better sim flight in 1, scanned info about environment more but still read back errors. Dual task better in diff modalities
Learning and multitasking
Mendoza 2018: students watched 20m ted talk, with mcq test after. Half kept phone and experimenters texted students. Test had diff quarters for the lecture. Those w phones scored worse, last quarter spec- linked to nomophobia- fear of being away from phone
Radiologists
Van der gijp 2017: expertsnfixate quicker on abnormalities, spend less time in non salient, diff search patterns, more systematic and global/focal. Kok 2012: used expert vs resident vs student: viewed chest x rays. Used diff patterns of issues either diffuse, focal or none. For focal: focus more narrow w expereince. Diffuse more efficient searches, less area looked and same for none. Students also less accurate w diagnosis when there was abnormality
Radiologist part 2
Dreq 2013: radiologists can detect abnormalities in 200ms, use gist, even before eye movement and at 70% accuracy. But localisation judgements after not accurate e.g. id problem fast but can’t say where/what-a use gist first to guide search. Novice all over the place: example of non selective and selective pathways
Baggage search mccalrey 2004
Mccarley 2004: diff to radiologists as doing complex searches under time pressure in disorganised space, hard to get gist. Ps searched for target over no. Trials . Accuracy and speed improved but then decreased when new target so more id then foraging.
Baggage search Wolfe 2005
In lab experiments target often there 50% time but for radiologist, only 0.3% of what radiologists see have abnormality. So did target present in 1, 10 or 50% in degraded stim w random objects- reward for correct answers. At 1% there was only 41% accuracy, even in 4000 trials. As more items, usually harder but here made no diff- worryingly low
Sport
Bard and fleury 76: expert vs novice basketball player shown pics of play and had to make choice: shoot, dribble, pass or stay. Experts didn’t make faster choices but focused more on empty space and teammates markers. Savelsbergh 2002: expert vs novice goal keepers- viewed penalty kicks and used stick to predict direction and height. Experts more accurate but waited longer to make decision, fixated on less area- kicking, non kicking leg and ball- novice more on top half and hips of kicker
