What dot clusters and bar graphs reveal: subitizing is fast counting and subtraction

Area of Science:

• Cognitive Psychology
• Visual Perception
• Numerical Cognition

Background:

• Subitizing and counting are distinct processes for number enumeration.
• Subitizing typically applies to small quantities (1-4 items), while counting is used for larger quantities (5+ items).
• The underlying mechanisms of subitizing and its relation to other numerical tasks remain under investigation.

Purpose of the Study:

• To investigate whether subitizing is involved in extracting numerical information from bar graphs.
• To compare the cognitive processes underlying dot enumeration and bar graph number extraction.
• To examine the flexibility and characteristics of the subitizing process.

Main Methods:

• Two studies were conducted using dot enumeration tasks and bar graph number extraction tasks.
• Response time (RT) functions were analyzed for different set sizes of dots and bar graph values.
• Factor analysis was employed to identify underlying numerical processing factors.
• Correlations between factor scores and task performance were examined.

Main Results:

• Dot enumeration showed distinct RT function slopes for 1-4 dots (subitizing) and 5-8 dots (counting).
• Bar graph recognition yielded a shallow-sloped, scallop-shaped RT function, distinct from dot counting.
• Factor analysis revealed two factors for dots but a unitary factor for bar graphs.
• Bar graph recognition correlated with dot subitizing, but not with dot counting.

Conclusions:

• Subitizing is employed in both the enumeration of small numbers of dots and the recognition of bar graph values.
• Subitizing appears to be a nonverbal process that flexibly operates on analog quantities with spatial extent.
• These findings support a fast-counting model of subitizing, suggesting it is more than just a simple recognition process.