|
|
|
Publications
Assessing Fit Quality and Testing for Misspecification in Binary Dependent Variable Models
In this paper, we present a technique and critical test statistic for assessing the fi t of a binary dependent variable model (e.g., a logit or probit). We examine how closely a model's predicted probabilities match the observed frequency of events in the dataset, and whether these deviations are systematic or merely noise. Our technique allows researchers to detect problems with a model's specifi cation that obscure substantive understanding of the underlying data generating process, such as missing interaction terms or unmodeled non-linearities. We also show that these problems go undetected by the t statistics most commonly used in political science. Coauthored with Justin Esarey.
In this paper, we present a technique and critical test statistic for assessing the fi t of a binary dependent variable model (e.g., a logit or probit). We examine how closely a model's predicted probabilities match the observed frequency of events in the dataset, and whether these deviations are systematic or merely noise. Our technique allows researchers to detect problems with a model's specifi cation that obscure substantive understanding of the underlying data generating process, such as missing interaction terms or unmodeled non-linearities. We also show that these problems go undetected by the t statistics most commonly used in political science. Coauthored with Justin Esarey.
Older, Younger, or More Similar? The Use of Age as a Voting Heuristic
Framing our analysis within the descriptive representation literature, we examine the use of a candidate’s age as a voting heuristic for members of the electorate across three electoral contexts (House, Senate, and gubernatorial elections) and two election cycles. Utilizing data from the 2010 and 2012 Cooperative Congressional Election Study (CCES), along with independently collected information on candidates’ ages, we argue that voters prefer to vote for candidates who are closest to them in age. Our analyses suggest that a candidate’s age can and does act as a voting heuristic. However, the strength of these findings are dependent upon the electoral context, individuals’ education levels, and the political party with which an individual affiliates. Coauthored with Steven Webster.
Working Papers
Gender and Twitter Behavior in the Russian Disinformation Campaign in the United States, 2012-2018
One of the major goals of disinformation campaigns is to ignite negative emotions in citizens that might exacerbate tension in a target country. Evoking anger is a common goal of disinformation campaigns, and campaigns can exploit gender norms as a possible cleavage in society. In this paper, we theorize how gendered disinformation campaigns use gender signals to promote their objectives. We answer the question, is gendered differentiated in disinformation campaigns, by analyzing over 850,000 tweets released by Twitter as part of the investigation into the activities of the Russian Internet Research Agency (IRA). We find that gendered patterns of disinformation do exist both in the levels of Twitter activity but also in the types of content released. These results suggest that understanding how gender norms can be exploited is an important component of combating disinformation campaigns.
One of the major goals of disinformation campaigns is to ignite negative emotions in citizens that might exacerbate tension in a target country. Evoking anger is a common goal of disinformation campaigns, and campaigns can exploit gender norms as a possible cleavage in society. In this paper, we theorize how gendered disinformation campaigns use gender signals to promote their objectives. We answer the question, is gendered differentiated in disinformation campaigns, by analyzing over 850,000 tweets released by Twitter as part of the investigation into the activities of the Russian Internet Research Agency (IRA). We find that gendered patterns of disinformation do exist both in the levels of Twitter activity but also in the types of content released. These results suggest that understanding how gender norms can be exploited is an important component of combating disinformation campaigns.
Sorted and Polarized: The Foundations of Partisan Ideological Change in the American Electorate
One of the most significant changes in the American electorate over the past several decades has been the increase in partisan sorting and partisan-ideological polarization. While several studies have examined the causes of this shift in partisan and ideological beliefs, they have been limited by the types of public opinion measurement available. I overcome the traditional limitations on measurements of ideology by using Bayesian Item Response modeling. Using data from the American National Election Studies, I show that shifts in partisan-ideological behavior began much earlier than traditional estimates project. My estimates suggest that shifts in partisanship and ideology began as early as the 1950s. Additionally, I present a novel, mass-centric explanation for partisan-ideological change in the electorate: rising levels of education. My results suggest that rising levels of education kick-started partisan ideological change in the 1950s.
One of the most significant changes in the American electorate over the past several decades has been the increase in partisan sorting and partisan-ideological polarization. While several studies have examined the causes of this shift in partisan and ideological beliefs, they have been limited by the types of public opinion measurement available. I overcome the traditional limitations on measurements of ideology by using Bayesian Item Response modeling. Using data from the American National Election Studies, I show that shifts in partisan-ideological behavior began much earlier than traditional estimates project. My estimates suggest that shifts in partisanship and ideology began as early as the 1950s. Additionally, I present a novel, mass-centric explanation for partisan-ideological change in the electorate: rising levels of education. My results suggest that rising levels of education kick-started partisan ideological change in the 1950s.
Partisan Polarization and Gender Socialization in the United States
Since the 1950s, voters have become more strongly identified with the two major parties, and within those parties, have become more ideologically extreme. This polarization has arguably led to greater policy deadlock but also greater mass mobilization. Given the significance of its consequences, understanding the causes of polarization is an important step in forecasting the durability of its effects. I present a novel argument about the causes of polarization with strong implications for the future of polarization. I argue that much of the polarization we see in current electorate is a consequence of generational replacement and generational differences in political socialization between men and women. I support this argument with data from the American National Election Studies. My findings have strong implications about the causes and durability of mass polarization in the American electorate.
Since the 1950s, voters have become more strongly identified with the two major parties, and within those parties, have become more ideologically extreme. This polarization has arguably led to greater policy deadlock but also greater mass mobilization. Given the significance of its consequences, understanding the causes of polarization is an important step in forecasting the durability of its effects. I present a novel argument about the causes of polarization with strong implications for the future of polarization. I argue that much of the polarization we see in current electorate is a consequence of generational replacement and generational differences in political socialization between men and women. I support this argument with data from the American National Election Studies. My findings have strong implications about the causes and durability of mass polarization in the American electorate.
