R

Visual search

Visual search is a perception task used to see how long it takes people to discriminate between objects based on their features when many similar objects are shown on a screen together.

Feature intergration theory says that there is a spotlight of attention and within this space all of the features (eg size, colour and orientation) are glued together by attention.

When looking for objects with a single unique feature people appear to search all the objects in view in parallel (at the same time) which means they search quickly.

When discriminating between objects on the basis of multiple features (conjunction search) people have to process them in series (one by one) which is slower.

Search slopes can be drawn to illustrate the increase in time for each additional object searched in serial search. Each item adds around 50 milliseconds to the search.

There is a 2:1 ratio for the length of time taken to complete a series search when the object is absent versus when it is present because in the absent condition the participant has to process every object, whereas in the present condition on average they will only have processed half of the objects when they find the target.

During a visual search if the features of the object are basic they can be searched in parallel so it is quick. Steep search slopes are seen for discriminating on high-level features (like facial features) suggesting that attention is paid to each object in turn (parallel search).