Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: Whereas many studies argue that roll call voting in the House of Representatives is increasingly partisan, this paper focuses on the fact that Congressmen are casting their vote based on the position of both the president and leaders simultaneously. Another objective of this paper is to test whether the strength of presidential influence is different across the type of policy. The analyses, using a path model and the CQ key vote for the 1953-1996 period, demonstrate that 1) presidents are more likely to take their position in foreign and defense policy or a policy with a national constituency, 2) presidents’ position taking generally makes voting results partisan, and 3) presidents wield a stronger positive influence on their party in bills having a national constituency than those without a national constituency.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: This paper explores notions of the American dream ideology, particularly its embracing of a degree of inequality in principle, and argues that it is not the static constant that academics and popular commentators have often taken it as. Rather, I argue that it is a dynamic concept that is continually questioned or reaffirmed and renegotiated with economic and historical shifts, and I document some examples of such shifts and their effect on the dominant ideology. I then explore changes in the political realm, particularly high levels of inequality in forms of political participation, what implications they hold for the American dream, and if such changes have led to challenges to or renegotiations of the ideology. Although it seems that high levels of inequality in both the economic and political realms should challenge certain tenets of the American dream, in many cases this has not occurred. I identify five major barriers that inhibit actual inequality from challenging the benefits of inequality in principle, which reveal the contradictions inherent in the American dream and the complicated ways in which Americans attempt to resolve these contradictions. However, despite these barriers, ideas do sometimes change, and I explore the possibilities that would enable such change.

Supporting Publications:
Supporting Document

Publication Type: Paper/Poster Proposal
Abstract: Epidemiological research relies on population estimates of medical conditions from health surveys. Despite a number of studies over the past fifty years, no consensus has emerged about the validity and reliability of survey estimates of chronic disease.

The Army Chemical Corps Health Study was designed as a long-term health assessment of a sample of approximately 4,000 veterans who served in the Chemical Corps during the Vietnam War era. Health and fertility outcomes for cases and controls were assessed in a 45-minute telephone interview. Since the validity of self-reported medical conditions could not be determined from medical records for the entire sample, a study was undertaken to assess the reliability of the survey reports of medical conditions as a potential silver standard for comparative disease prevalence in the two populations.

A sample of 495 cases was drawn from 2500 interviews that had been completed by the end of November 1999. This sample was stratified by case/control status (Vietnam/not Vietnam) and by consent for medical records (no medical conditions requiring consent/medical conditions with consent/medical conditions without consent).
The retest survey that was initiated approximately eight months after the initial interview was limited to questions about the lifetime prevalence of eleven health conditions from the baseline survey and the date of diagnosis if a condition was reported. Ultimately, 449 out of the sample of 495 subjects (91%) were re-interviewed in the retest survey.

This paper reports the reliability of chronic condition reports in health surveys based on repeated measures of the same conditions by the same respondents less than one year apart. In addition to the overall reliability of reporting in health surveys, the paper explores differences in reliability by condition and type of respondent that might explain the sources of variability in reporting of medical conditions.

Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: Panel studies are of crucial importance to our understanding of the complex, interacting nature of causal processes in the social world. A limitation to valid and reliable inference from panel studies, however, derives from the reflexive nature of humans as research subjects; the act of observation often serves to transform the behaviour of those being observed (Kalton and Citro 2000). In panel studies this source of error falls under the general heading of ‘panel conditioning’ bias and relates to the fact that responses to questions in later rounds of the panel may be influenced by those given in earlier waves. Factors such as ‘satisficing’, cognitive consistency and self-presentation are likely to play important roles in influencing responses between waves. However, while conditioning effects are widely believed to be pervasive in panel survey research, the general consensus appears to be that their impact on marginal distributions is, at most, negligible (Holt 1989).
In this paper, we take a somewhat different approach and examine the effect of panel membership on the internal consistency or reliability of attitude items. Using data from the British Household Panel Study, we evaluate the hypothesis that simply administering political attitude questions serves to increase the apparent consistency or ‘constraint’ of respondents’ Political Belief Systems (Sturgis 2001). Such an effect is argued to arise because the interview prompts respondents to reflect on the subject matter of the questions and to think more deeply about how the issues covered may be inter-related (Jagodzinski et al 1987). Our results show that estimates of internal reliability increase between the first two administrations of the attitude items examined but not between subsequent waves. These increases in scale reliability are, furthermore, predominantly concentrated amongst the least politically engaged respondents in the sample.

Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: Citizens employ a range of heuristic devices in reaching judgments and making evaluations regarding political candidates and issues. This paper examines the circumstances that lead to the use of multiple and potentially competing heuristics in making complex decisions. In our analysis, subjects demonstrate confusion regarding ideological and partisan heuristics - an inability to disentangle partisan and ideological criteria in the evaluation of candidates. We argue that the political environment alters both the accessibility of particular heuristic devices, as well as the strength of association in memory among alternative heuristic devices. Hence the ability of citizens to select among heuristics is conditioned on the larger environment of incoming information. As a practical matter, this means that the ability of citizens to employ various judgmental criteria depends on the larger political debate, and accessibility provides the connecting link between environmentally supplied political messages and the terms and criteria of individual political choice.