Duke University
Explaining Variance in Affect Control Theory: Cultural Consensus, Deflection, and Redefinition
Abstract
dc:description.abstract<p>Affect Control Theory (ACT) conceives of affective sentiments as shared</p><p>meanings among individuals within a single culture. Recognizing the theory's potential</p><p>to explain cultural differences and behavior patterns, many researchers aim to test and</p><p>apply ACT's insights to within- and across-culture analysis. The growth of the theory's</p><p>popularity necessitates a review and exposition of the theory's fundamental</p><p>methodological assumptions and its causal mechanism, deflection. Using data from the</p><p>2003 Indiana EPA dictionary, I map the distribution of fundamental U.S. sentiments in</p><p>EPA space, define two new conceptions of deflection, map the universe of event</p><p>currently measureable deflections, and discuss the ramifications of these findings for</p><p>past and future research.</p><p>I critique ACT's operationalization of "shared meaning" as mean point estimates</p><p>calculated from individuals' numeric ratings on semantic differential scales. Past</p><p>research attributes variation in concept ratings to two sources: unsystematic error in the</p><p>measurement tool and imperfect cultural inculcation among respondents. By taking a</p><p>concept-focused approach, I show that variation between respondents is structured by</p><p>the institutional affiliation of identity concepts and concept labels' word difficulty. This</p><p>pattern exists even when controlling for individual-level characteristics, the traditionally</p><p>ascribed reason for variation in concept ratings.</p><p>I replicate a well-known ACT study that found support for the dynamic behavior</p><p>redefinition hypothesis and did not find support for ACT's redefinition hypothesis. I</p><p>make the design more robust and test both the original findings and my claims about</p><p>the role of institutions in ACT. I find support for the dynamic behavior hypothesis,</p><p>partial support for the ACT hypothesis, and support for the claim that individuals</p><p>depend on institutional information inherent in identity meanings.</p>
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Curdy, Brent Harrison
- Advisor dc:contributor.advisor
-
- Smith-Lovin, Lynn
Subjects
dc:subject × 4Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/10161/16952
- OAI identifier oai:identifier
- oai:dukespace.lib.duke.edu:10161/16952