Change Blindness

What is Change blindness?

Change blindness is the phenomenon due to which a person observing a particular scene fails to detect large changes induced in the scene. This phenomenon is attracting an increasing amount of attention from experimental psychologists and philosophers alike.

Many experiments have been conducted that show a surprising lack of ability to observe striking dissimilarities in two successive scenes.

Can you detect any change in following image?

Image courtesy: http://nivea.psycho.univ-paris5.fr/

Most of us would have noticed the appearance of a bag. We are good at change detection when the change signals are easily visible, i.e when the shift from the original version to the changed version is instantaneous and clearly visible; as well as when there are no other distractions[1].

But, experiments have shown that people are quite unable to notice any change when they are momentarily distracted by a "flicker", "mudsplats" or when the eye blinks. Let us see an example of a "flicker" test, which was developed by Ron Rensink in 1990.

Do you see anything odd about this image?

Image courtesy: http://csc.ncsu.edu

No matter how hard we stare at this photograph, most of us are unable to figure out anything odd about it. Look more closely. The engine of the airplane disappears and reappears from under the wing!!

Most people observing such photographs cannot believe that they have missed such a large change in the scene. We conducted an experiment on 16 of our wingmates and asked them to observe the above photograph.

Most of them did not find anything odd at all. Even when they were informed that this scene consisted of 2 distinct photographs, only 1 was correctly able to identify the change.
1 out of our 16 friends said that he was "feeling" that the 2 photographs were different. He even pointed out the region of change but was unable to pinpoint the exact nature of it. A researcher in this field, Rensink, calls this gut feeling, "mindsight" and we shall discuss this concept subsequently.

When do we experience Change Blindness?

Change blindness can happen due to many reasons:

Due to diversion of attention from the change signal
Changes in visually more distinct objects are easier to detect than changes in non-prominent regions
Attention to the changing object is necessary for change detection
Sometimes, attention to the changing object also may not be sufficient as has been found in some studies[2]

Objects in an image that are preferentially given attention are more likely to be encoded and compared. Faster detection of the change can be attributed to earlier attention to the changing object, which can be studied closely to provide insight into personality and expertise differences. Although attention can be divided among 4-5 objects at a time, only a single change can be detected at a time.

When a blank screen is introduced between 2 images, observers fail to make a distinction between the 2 photographs. The blank screen functions to produce a luminance change everywhere in the image which obscures the visibility of the changing signal. Changes made during eye blinks, 'mudsplats' and brief occlusions also function in a similar manner.

Let us see an example of a mudsplat test:

Image courtesy: http://nivea.psycho.univ-paris5.fr

Changes in attended objects often go unnoticed due to its unexpected nature. Awareness of the change can bring about an attempt at conscious change detection.

Person Swap Experiment:

The following experiment is done at a hotel on the customers who come there. While filling the form for the room acquisition the person at the reception bends down for something a new person comes in place of him. Most people (>50% [1]) are unable to notice this change.

video courtesy: http://www.vids4students.com

Why do we experience Change Blindness?

There are many theories that try to explain the phenomenon of change blindness.

The mechanisms that generate eye movement actively suppress perception during this time. This phenomenon is called saccadic suppression[1].
Some researchers have argued that our internal visual representations are completely absent due to which we are not able to detect change (We shall examine this in detail in the next section)
It is possible that this failure to detect change is due to our inability to compare information from before and after the change, rather than failure to represent the pre-changed scene
Another argument suggests the possibility of the preserved information being in a form or format that is not usable for conscious change perception

Are our internal visual representations absent?

Visual representation is the mental projection of any scene as we see it through our eyes. This representation should have content (what is being represented) and format (nature of the representation)[3]. Some researchers argue that change blindness implies that our internal visual representations are absent or sparse which was strongly supported by blindness to large changes in scene. But for this to happen there are following requirements which must be met as outlined in Simons' and Rensink's paper [2] :

Evidence must eliminate the possibility that detailed and complete representations exist, but decay or are replaced before change perception occurs. Experiments on visual search and iconic memory indicate that detailed representations do exist, but are volatile, lasting only 0.5 s or until some new stimulus replaces them.
Evidence must eliminate the possibility that representations of the pre-change stimulus exist, but are at locations or in pathways inaccessible to the mechanisms used for change perception.
Evidence must eliminate the possibility that such representations exist, but are in a format that cannot be used for change perception.
Evidence must eliminate the possibility that such representations exist in an appropriate format and pathway, but the comparison operation is not applied (even though it could be).

Mysterious "Mindsight"?

Change detection research has also developed ways to explore possible non-attentional (and perhaps nonconscious)
contributions to visual perception. Such processes might register the presence of a change before conscious perception, either via the possible operation of implicit processes or via other mechanisms such as ‘mindsight’ – the ability of some observers to experience a ‘gut feeling’ as to the presence of a change before they visually identify it [2].

Mindsight is the phenomenon that a person can sense change without actually seeing them. For example, in the flicker test some people say that they sense something changing without being able to visually identify it. This implies that the observer represents information from the scene and compares it, perhaps without focused attention to the changing region. True mindsight has been considered by some to be similar like a sixth sense.

However there are arguments against this. It is possible that the feeling of sensing some change is due to the demands of the task itself.

Can we draw any conclusions?

Change blindness is an amazing phenomenon. There are a lot of arguments regarding change blindness which are not yet proved or disproved, though it is sure that this phenomenon can be immensly helpful in our undersatnding of the human visual perception system and its working. It makes us aware of our limits in detecting relatively large changes and tells us that inspite of our sincere attempt to be attentive for small events, we will not necessarily succeed. Most of us will believe that our vision can be impaired as demonstrated by change blindness - this counterintuitiveness called "change blindness blindness," which is also garnering a lot of scientific attention. It is a direct proof for those people who firmly believe that they are fully vigilant during driving while talking on a phone. We see that when we are unable to detect changes consciously then how can we be confirm regarding the same when our attention is divided into multiple tasks. These studies are helpful for determining the effect of personality and expertise on the percetion. Also change detection techniques can help us identify training methods that can be adopted for more attention seeking jobs.

References:

[1]Simons, D.J., & Ambinder M.S. (2005). Change blindness: Theory and consequences. Current Directions in Psychological Science 2005 14: 44 . It article provides a good overview of change blindness, its current state/theories and future directions.

[2] Simons, D.J., & Rensink R.A. (2005). Change blindness: Past, present and future: TRENDS in Cognitive Sciences Vol.9 No.1 January 2005 . This paper also talks about the change blindness related concepts and the inferences drawn from them. Specially it give some strong arguments agains some erroneous inferences (as they say it) like absence of visual representation.

[3] http://www.jimdavies.org/research/visual-analogy/proposal/node32.html : Gives a description of the concept of visual representation