Survey Research And Methodology Program
Date of this Version
4-2010
Abstract
This study assesses the genetic heritability of various survey response styles using a classical twin design. The National Survey of Midlife Development in the United States (MIDUS) collected in 1995-96 included an oversample of twins with self-reported zygosity along with a large number of survey items that allowed for the assessment of acquiescent and extreme response style. The MIDUS singleton sample was used for the careful development of appropriate and reliable measures of these traits. The second wave of the MIDUS (2005-06), was used to assess the sources of survey response trait stability. Acquiescence appears to have a sizable and significant heritability component (~25% to 34%) and is not influenced by the environment co-twins share; extreme response has a smaller and nonsignificant heritable component (~20%) and a more sizable shared environmental effect (~32%). Additive genetic and common environmental effects drive the test-retest stability of traits. The variation explained by these two effects also wholly contributes to the covariation across waves as well. Controlling for income and education does not change results. Results suggest it is inappropriate to treat acquiescence and extreme response as a single survey response style phenomena and call for the exploration of specific genes in the case of acquiescence and social and familial environmental predictors of extreme response.
Comments
A Thesis Presented to the Faculty of the Graduate College at the University of Nebraska In Paertial Fulfillment of Requirements for the Degree of Master of Science, Major: Survey Research and Methodology, Under the Supervision of Professor Robert F. Belli. Lincoln, Nebraska: May, 2010.
Copyright 2010 Levente Littvay.