Change Blindness

1884 Words8 Pages
The Effect Of Central And Marginal Interest Changes On Change Blindness Abstract Change blindness is a phenomenon describing the failure of people to notice large changes to a visual scene. This study examined the effect of central and marginal interest changes in images on change blindness. It found that changes in central interest objects were perceived significantly faster than changes in marginal interest regardless of the type of change made. This provides support for Rensink et al’s research in 1997 and the theory that change perception is dependant on attention. This challenges traditional models of perception based on memory of internal visual representations. This research also found that different types of changes seemed to…show more content…
Currie et al, (1995) found that when participants focused on an object that moved between displays in a variety of visual stimuli they were only able to detect changes in the saccade target, not the surrounding image. This change blindness was attributed to saccade specific mechanisms but other evidence suggests many incidences of change blindness (and saccade-contingent change blindness) can be attributed to reduced attention. Transient motion signals are normally a huge visual cue by attracting an observers attention (Klein et all 1992) and absence of these signals for example retinal blurring in a saccade or modified picture presented at intervals may be responsible for change blindness when attention can not easily be directed to the change. Rensink et al suggested that people never form a completely accurate, detailed mental image of scenes they observe, making it difficult to perceive changes in images or their environment unless full attention is directed. Their research on change blindness used real world images and modified versions presented very quickly, intercepted by blank fields in a “flicker” sequence. They found that response times were faster if the change took place in an object of central interest compared to changes in objects of marginal interest, for all 3 different types of change…show more content…
There does not seem to be a particular change type that is easiest or most difficult to spot in both conditions, however in MI, location appears to take a lot longer to spot than colour and presence (mean location=16293.25 compared to colour= 9486.44 and presence= 113.96.13) Mauchleys test of sphericity was found to be non-significant (F=0.931p>0.05) therefore spherificity was assumed. The ANOVA test shows a significant effect of the Interest (F= 83.079 p<0.001), Change Type (F= 5.406 p <0.05) and a significant effect of Interest and Change Type (F=13.094 p<0.001). The dependant T-Tests shows a significant effect of interest (central vs marginal) on response times for colour (t = -3.25, df = 15, p< 0.05) for location, (t = -9.9, df=15, p<0.01) and for presence, (t = -6.23, df = 15, p<0.01). Therefore the hypothesis must be accepted and the null hypothesis rejected as there response times were faster in the CI than the MI with all 3 types of change. Figure 1: Chart showing the total number of inaccurate results for each change type in each interest
Open Document