Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


ucin1321372828.pdf (878.07 KB)

ETD Abstract Container

Abstract Header

Coordination of Local and Global Features: Fractal Patterns in a Categorization Task

Castillo Guevara, Ramon D.
http://rave.ohiolink.edu/etdc/view?acc_num=ucin1321372828

Abstract Details

2011, MA, University of Cincinnati, Arts and Sciences: Psychology.
To categorize an object successfully, one needs to attend to both local and global aspects to group it meaningfully. The open question pertains to how the mind coordinates local and global features of a display. Coordination of mental processes can be studied by looking at correlational patterns across many thousands of trials. The coordination among these time scales can be described according to a mathematical constructs called fractals. Even though fractals have been detected in trial series pertaining to motor and cognitive tasks, they have not been evaluated in categorization tasks in which are required attention to local and global features. The goal of this research was to fill this gap. Two experiments were implemented on undergraduate students in order to obtain a measure of the coordination of local and global aspects during a categorization task. Participants had to decide if two stimuli had the same contour around a string of elements, and if they shared an element. The assumption was that such competition might maximize the degree to which global and local attributes need to be coordinated. Therefore, this coordination might be characterized by fractal exponents, which in turn might be susceptible to constraints in task and participants. Experiment 1 (n = 32) tested the effect of two of those task constraints; the first being the predictability of the response of a trial from the answer of the previous trial. Depending on condition, trials were presented either in random order or in a predetermined sequence that allowed participants to anticipate the next trial. The idea was that predictability might increase the fractal exponent, while added randomness decreases the fractal exponent. Another factor manipulated pertained to the participants‘ skill. Here, it was assumed that a greater familiarity was associated with a greater skill in identifying the stimuli; and that fractal exponents close to 1/f pink noise must appear more clearly with familiar stimuli than novel stimuli. Experiment 2 (n = 64), was based on a new version of the categorization task. Here, the structure was modified to decrease the competition for attentional resources. In some conditions, participants were asked to make decisions by focusing their attention on elements, while in other conditions they were asked to focus on the global shapes. Thus, it was hypothesized that the fractal exponents obtained in the second experiment should be higher than those obtained in the first experiment. Reaction times were compared between familiarity of stimuli, the predictability of the next trial, and among types of slides. Also they were submitted to Spectral Analyzes to detect fractal organization. The results show that the predictability of the next trial was the most important factor. When trials were administered sequentially, exponents were higher than those obtained with random procedure. Additionally, when decision was centered on global aspects, exponents were higher when decision was centered on local elements and centered on both aspects. Finally, alpha exponents were higher in only one condition—where participants were focused on global shapes and responded to familiar stimuli presented sequentially.
Adelheid Kloos, PhD (Committee Chair)
John Holden, PhD (Committee Member)
Guy Van Orden, PhD (Committee Member)
49 p.
Cognitive Therapy
Categorization-Coordination; Perception-Attention; Fractal Analysis; 1/f noise; Self-Organized Criticality; Spectral Analysis

Recommended Citations

Citations

  • Castillo Guevara, R. D. (2011). Coordination of Local and Global Features: Fractal Patterns in a Categorization Task [Master's thesis, University of Cincinnati]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1321372828

    APA Style (7th edition)

  • Castillo Guevara, Ramon. Coordination of Local and Global Features: Fractal Patterns in a Categorization Task. 2011. University of Cincinnati, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=ucin1321372828.

    MLA Style (8th edition)

  • Castillo Guevara, Ramon. "Coordination of Local and Global Features: Fractal Patterns in a Categorization Task." Master's thesis, University of Cincinnati, 2011. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1321372828

    Chicago Manual of Style (17th edition)

ucin1321372828
448
© 2011, all rights reserved.
This open access ETD is published by University of Cincinnati and OhioLINK.