A Model-Based Analysis of Anxiety and Biased Processing of Threatening Information

Please use this identifier to cite or link to this item: http://hdl.handle.net/1811/44499

Show full item record

Files Size Format View
hayes_coreywhite.pdf 541.9Kb PDF View/Open

Title: A Model-Based Analysis of Anxiety and Biased Processing of Threatening Information
Creators: White, Corey
Advisor: Ratcliff, Roger
Issue Date: 2009-04
Abstract: The present study employed a decision model to explore processing biases in anxiety. Anxious individuals show preferential attention to threatening compared to neutral information, which can serve to maintain or increase levels of anxiety. Several models of anxiety predict biased processing of threat under most circumstances, but previous research has failed to demonstrate a threat bias in certain tasks involving word recognition and memory. This has been taken by some researchers as evidence that the threat bias is dependent on contextual factors. An alternative explanation is that threat bias is present in these tasks, but analytical methods are not sensitive enough to detect small differences. The present study was designed to determine if threat bias could be shown in word recognition and memory tasks. Individuals with high and low levels of anxiety were compared in their responses to threatening words. Behavioral data were analyzed with the diffusion model, a quantitative model of decision processes that provides a more sensitive measure of processing than traditional measures, response times and accuracy. The diffusion model analysis showed that the high-anxiety participants had consistent advantages for early processing of threatening words, but disadvantages for later memory of threatening words. The results from these experiments are consistent with several models of anxiety, with anxious individuals showing initial vigilance in the detection of threat, but avoidance of deeper processing of threatening information. The improved sensitivity of the model-based analysis suggests that previous failures to show threat bias might have been due to use of imprecise dependent measures. By demonstrating threat bias in these tasks, there is no longer a need for ad hoc explanations of why the bias does not occur in certain conditions, allowing for more parsimonious models of anxiety.
Embargo: No embargo
Series/Report no.: 2009 Edward F. Hayes Graduate Research Forum. 23rd
Keywords: Anxiety
Cognitive Model
Description: Social and Behavioral Sciences; Social Work; Law: 1st Place (The Ohio State University Edward F. Hayes Graduate Research Forum)
URI: http://hdl.handle.net/1811/44499
Bookmark and Share