Searching the visual scene - A2 Flashcards
What did original ideas about how we search the visual scene suggest?
searches that rely on more than one element of the target (e.g. red crosses) are less efficient than ones that rely on one element (e.g. red)
What factors are usually varied in visual search tasks?
- number of distractors
- presence or absence of targets (positive or negative)
What are feature searches and conjunction searches?
- feature = target is distinct from distractors by one feature
- conjunction = target is distinct by a combination of features
What are the findings for visual search tasks for feature searches?
- positive trials = size of the array doesn’t affect RT
- negative trials = size of array does affect RT (more to search to check it isn’t there)
What are the findings for visual search tasks for conjunction searches?
- the size of the array increasing always increases RT
- this increase in RT is greater for negative trials
When looking at visual search task results, what does m represent in the equation?
the increase in RT for every new item added to the array
What are the 2 processing stages in Feature integration Theory? What are some characteristics of them?
- feature detection - fast, parallel, automatic and efficient - single features pop out
- feature integration - gluing all the features together (attentional focus is the glue) - slow, inefficient search
What did Wolfe find with regards to the predictions made by feature integration theory?
there is no real pattern in the results between feature and conjunction searches
What does Wolfe’s model of guided search suggest?
- there is a top-down part that constrains how you search and a bottom-up part that interact to create a more efficient search
What is an example that supports Wolfe’s model of guided search?
When looking for a red item of specific orientation it is easier to find it when some of the shapes are green than them all being red as the top-down process can make you ignore all the green items, reducing the search
What things drive bottom-up search? (2)
- salience (target is more or less different to distractors)
- attributes (elements that capture attention)
What drives top-down search? (2)
- scene properties
- values
What does attentional engagement theory suggest?
the efficiency of the search is based on aspects of the task
- target/non-target similarity and non-target/non-target similarity
How does target/non-target similarity affect search?
more similar = target less likely to pop out so search is less efficient
How does non-target/non-target similarity affect search? What implications does this have?
- if they are heterogeneous (all different) then search is less efficient, even though they still only differ by one dimension from the target
- doesn’t fit with FIT as that suggests that one dimension different = should always pop out