Please enable JavaScript.

Coggle requires JavaScript to display documents.

Change Blindness discussion - Coggle Diagram

- - - - could our expectations lead us to disregard incongruent changes as 'irrelevant?' i.e., not 'relevant' to the scene
  - - - the 'blank screen' interrupts sustained attention
        
        the blank screen places higher demand on working memory by forcing participants to store two potentially different images to working memory as opposed to if the images were continuous/unbroken and therefore processed as one image p69
        
        'visuospatial scratchpad' is the specific component of working memory which is overloaded
        
        Psychological refractory period (PRP) - "delay in responding to the second of two stimuli presented very close together in time, assumed to occur because information processing systems are still preoccupied with processing the first stimulus"
        
        "it takes longer to respond to the second stimulus when the interval between the stimuli is very short (typically, about half a second" - the blank screen's presense for 0.25 seconds qualifies
        
        'Attentional blink' week 13
        
        however, a similar effect was observed using 'mushsplashes' which did not obscure the entirety of the scene p121
        
        this effect was believed to be due to a diversion of attention, as opposed to the ISI of the blank screen p122
        
        "abrupt changes in the enviroment can capture attention"
  - - - extraction of low level physical properties (eg size, shape, colour)
    - - concerned with "more complex psychological properties (Such as meaning)" p.65
        
        limited capacity, "this second stage operated in a serial fashion (i.e, processing one peace of information at a time)
        
        this could produce a 'bottleneck in processing', 'restricting the speed or amount of information that can be processed. p65
  - - - detection of change requires a 'conjunction search' which employs greater cognitive resources through identifying an object (or change to an object) through processing multiple features at once (size, shape, colour)
      - higher liklihood of error on account of complexity
        
        "failing to attend to an object, the withdrawal of attention from a critical part of the scene, or an impoverished representation in visual memory can all account for failures [in perception]" p 125
      - "to integrate features to form a lasting representation requires us to focus on the object... to establish a representation known as an object 'file' p124
        
        introduction of an ISI and therefore a PRP may hinder formation of this 'object file' and therefore storing of the image to short term memory, necessary for making a comparison to the subsequent modified image
        
        however, "fixating on an object does not guarantee that an individual has the experience of seeing it - is possible to look without seeing" (65%, using eye tracking)
    - - automatic 'scanning' of the environment for basic attributes such as shape or colour
        
        the blank screen inhibits direct comparison of the before and after images, is this due to a failure to store information to working memory?
        
        "we have n abiltiy to detect basic visual features in the environment with little effort" p124
        
        this would suggest that we are most proficient at detecting 'within-category' changes as this typically requires locating the change of a singular dimension (colour, pattern, shape) however our results contradict this
        
        does this utilise a broader mode of attention, whereby multiple stimuli are processed in tandem?
  - - - "In similar fashion, Rensink, O’Regan, and Clark (1997) demonstrated how the global properties of a scene can constrain semantic points of interest. In this experiment, participants detected changes relatively efficiently in areas of the scene that had been deemed to be of high interest by an independent group of raters. Taken in combination, these results show the strong influence of prior experience, expertise, and scene schematics on
        the deployment of focused attention."
    - - "The detectability of a change depends on the physical saliency of the change as well as on observers' expectations as to which objects are most likely to change... Such expectations may be based on object semantics (e.g., clouds are more likely to change than buildings), guesses about the experimenters intentions, or guesses regarding objects that lend themselves to a manipulation."
        
        "in their original experiment they found it took 11 seconds to notice a marginal interest change, and 5 seconds to notice a central interest change" p121
        
        evidence for the position of the change as a confounding variable, include in limitations of study
      - "Target objects that were contextually congruent with the background scene were found faster and fixated earlier than target objects that were contextually incongruent with the background scene."
      - "Responses were significantly faster for the incongruent context (M4833 ms) than for the congruent context (M5745 ms),
        t(19)7.27, pB.001."
      - "when the ISI was reduced to 0 ms, change detection performance was faster for congruent targets than for incongruent targets, and there were fewer identification errors for congruent targets than for incongruent targets"
        
        this suggests that the ISI plays a significant role in allocation of attention within change detection tasks
        
        change detection in the abscence of an ISI relies on 'transient visuals [look up to elaborate]
        
        changes between scenes including an ISI used 'gist based attention capture'
        
        why do these findings contradict the findings of our experiment?
        
        "there does not appear to be any one model to explain all the phenomena... this is because there are a multitude of methodologies used and also a range of cognitive mechanisms that fully, or partly, determine any effect (e.g. attention, perception and visual short-term memory) p124
  - - - contextual cues within the scene 'triggers' schemas of that scene which then shapes our expectations about objects and their relationship to that enviroment
        
        stimuli that do not conform to this context and therefore expectations may therefore be more likely to be missed
      - "lots of different, and often competing, schemas to choose between. An unconscious 'contention scheduling' system select which schema to perform, in a fairly automatic way" p.72
        
        elaborate on how expectations/previous experience shape perception
        
        Gorilla/basketball inattentional blindness study (simon and chabris 1999) p129
        
        what role do semantics play in this failure of perception? is it because a gorilla walking through the scene 'doesn't make sense?'
- - - - as we shift our attention from one object/stimuli to the next, we are blind to incoming sensory information such as the changes that may be happening in our peripheral
        
        this phenomenon in conjunction with the 'blinding' of the blank screen and blinking causes a significant reduction in sensory input
        
        "Posner's (1980) 'spotlight' model of attention suggests that attentional acuity declines the further away a stimulus is presented from the central fixation point" p.135
  - - - "given the limitations of short-term memory... and attention, it makes sense that we are not highly sensitive to changes in the environment because we do not have capacity to represent them all" p116
  - - - this demonstrates the limitations of our peripheral vision for noticing change
      - this study also demonstrated our tendency to miss stimuli in our centre of our visual field if attention is quickly diverted to the periphery
        
        this may occur as we switch attention between objects at the pace of the blank screen (ISI = 250ms), we may therefore look at the changing object but fail to notice it, moving on to other areas of the visual of the scene with each new ISI
        
        similarly, the rapid re-directing of focal attention involved in the task of following a bouncing ball around the screen may blind us to other changes occurring in the center of our visual field
        
        this in combination with the blinding effect of making continuous shifts, resulting in saccadic suppression
        
        "50% of people failed to detect the gorilla, even when presented at fixation" p.131
        
        Eriksen and st james (1986 p136 'zoom lens' theory
        
        "information from a wide range can be processed in a general way, but a smaller area can be selected for analysis in finer detail" - this 'finer detail' likely includes the attribution of semantics, key to situating an object within its context
  - - - this could be argued to demonstrate the role of expectation in perception
        
        Similarly to how participants did not 'expect' a cross shape to enter the scene and therefore may be 'filtered out' by attentional allocational strategies, incongruent objects may also 'filtered out' as a result of expectational effects on perception
      - "we tend to adopt stategies to reduce cognitive workload.... we tend to rely upon our expectations... and therefore adopt a visual attention strategy based on these expectations. This allow us to ignore irrelevant information, and attend to pertinent information"
        
        such efforts to reduce cognitive workload could be amplified by participant's apathy toward to the task and low consequentiality regarding task performance
        
        "Such changes are only detected if there is high-level interest in the change. Therefore, although you focused your attention on the task of detecting a change, there was no real, internalised need to detect the changes" week 13 ou
      - "when the cross was the same colour as the set of objects attended to, 94% of participants noticed it, when the cross was a different colour, only 3% noticed it" p133
        
        stimuli congruent with the task (noticing a red cross when monitoring the movements of red 'L's) are noticed more than stimuli incongruent with the task
        
        in a similar way, we may be more inclined to look for congruent changes as a result of the contextual cues inherent in the semantics of the scene