Announcements

QSS offers course in Quantitative Literary Criticism

The Program in Quantitative Social Science at Dartmouth College offers students the opportunity to combine modern methodological training with enduring questions in the liberal arts. This is exemplified by Neukom Fellow Joseph Dexter's course in Quantitative Literary Criticism. This course, cross-listed in QSS and the Department of Classics, is described as follows:

Digitization of vast numbers of texts and rapid advances in computational methods are enabling new forms of criticism in all areas of literary study. Classics was an early adopter of digital technologies, and computation is now pervasive throughout the field, as illustrated by flagship projects such as the Perseus Digital Library. Beyond the familiar examples of digitized texts and simple word searches, scholars and students also benefit from an ever-growing array of sophisticated quantitative tools, and from increasing engagement with diverse technical disciplines – natural language processing, data science, even bioinformatics. Through a survey of recent research at the intersection of Latin literature and the digital humanities, this course will introduce you to the state of the art in quantitative literary criticism. To ground our methodological investigations, we will explore a diverse selection of Latin poetry, including epic (Vergil, Lucan, and Catullus), elegy (Catullus), and comedy (Plautus), and sample some less famous later authors, such as Paul the Deacon and Vitalis of Blois, who were influenced by classical antecedents. At each turn, we will examine the interplay between traditional (close reading, philology, theory) and data-driven analyses of Latin literature and consider how quantitative methods can support humanistic inquiry. Along the way, you will gain hands-on experience with powerful computational tools and be introduced to now ubiquitous critical approaches, such as intertextuality and reception studies. Assigned readings will be in English translation using bilingual Latin-English editions; in addition to reading all of the English, students with Latin will be responsible for understanding and translating "micro samples" of the original texts.

Joseph Dexter received his doctorate in Systems Biology from Harvard University after having studied chemistry as an undergraduate at Princeton University Dr. Dexter is one of the founders of the Quantitative Criticism Lab, and his current research focuses on the use of computational methods from natural language processing, machine learning, and bioinformatics to understand large-scale changes in literature and culture.

Dr. Dexter taught Quantitative Literary Criticism last spring. During the term, students had the opportunity to participate in the Digital Humanities Beyond Modern English  conference, which brought 14 experts on text analysis for premodern languages to campus. Teams of students also completed independent research projects of their own design, including application of the Tesserae search tool to trace patterns of influence in Latin literature and contributing to the open-source Classical Language Toolkit.

Jennifer Wu '19 presents thesis at Japanese politics workshop

Jennifer Wu, who graduated with a degree in Quantitative Social Science from Dartmouth College in 2019, presented her undergraduate honors thesis at the inaugural meeting of the Northeast Workshop on Japanese Politics.  This workshop was organized by Professor Yusaku Horiuchi of Dartmouth's Department of Government and funded by a conference grant from the College. The speakers at, and co-organizers of, the workshop included Frances Rosenbluth  at Yale University, Christina Davis  and Daniel M. Smith at Harvard University, and Amy Catalinac at New York University.  According the Professor Horiuchi, it was one of the largest workshops ever organized specifically on Japanese politics.Jennifer is currently a pre-doctoral fellow at Stanford University, where she is supervised by Professors of Political Science Andrew Hall and Justin Grimmer.  Jennifer is currently working on a project studying Youtube and political polarization, and she is taking Stanford's quantitative methods sequence for political science.  While at Stanford, Jennifer plans to continue researching political communication and behavior, and with Professor Horiuchi she is working on revising her QSS honors thesis for submission to an academic journal.

QSS scholars publish article on racial inequalities in drug policy

Three scholars associated with the Program in Quantitative Social Science at Dartmouth College recently published a paper in the Journal of Health Politics, Policy and Law. Their paper, titled "Treatment versus Punishment: Understanding Racial Inequalities in Drug Policy," analyzes racial inequalities in policy responses to the ongoing opioid crisis and the crack scare of the 1980s and 1990s. Many observers believe that policy responses to the opioid crisis are less punitive than those associated with the crack scare and that the reason is that victims of the former are (stereotypically) white. The study's authors compare policy responses to these twin health crisis and show that legislators across the United States have introduced more drug treatment-related bills during the opioid crisis than punitive bills.  This was not the case during the crack scare.  However, the study's author show as well  that legislators seeking to response to the opioid crisis have been more responsive to white deaths than black deaths in their legislative activities. This result suggests that the recent shift toward treatment-oriented responses to the opioid crisis is driven by white victims, and it constitutes evidence that racial inequalities in American drug policy are persistent.

The study's authors are Jin Woo Kim, Evan Morgan, and Brendan Nyhan. Kim, a post-doctoral fellow at the University of Pennsylvania, was a post-doctoral fellow in QSS during 2017-19 and is currently working on research projects on how partisans respond to new information and the effects of political rumors on public opinion. Morgan graduated from Dartmouth with a degree in QSS in June 2019 and is currently working as a data engineer at Mastercard. Nyhan is Professor of Government at Dartmouth College and serves on the QSS Steering Committee.

QSS alumni work with data and each other

Three former Dartmouth College students, Franklin Dickinson '16, Clara Wang '17, and Junghye Kim '19, who studied in the Program in Quantitative Social Science, recently crossed paths as employees in the Washington, DC and Chicago offices of Civis Analytics, a data science consultancy and technology firm founded by the former chief analytics officer for Barack Obama's re-election campaign.  Clara Wang was the first Dartmouth student to complete a minor in QSS, and Junghye Kim was a QSS major.

These three students all took Brendan Nyhan's Experiments in Politics seminar, which is cross-listed in QSS and Dartmouth's Department of Government.  They are also co-authors on a manuscript reporting the results of the randomized controlled trial they conducted in the class, which they hope to publish in the coming year. Clara Wang recently left Civis after two years to pursue graduate study in China after being selected as a Yenching Scholar.

The Applied Data Scientist role at Civis involves coding in R, SQL, and Python to analyze, visualize, and communicate data, with a particular focus on survey data and predictive models. On the Political team at Civis, data of interest often pertain to which races are most important when considering resource allocation, the demographics and preferences of the electorate within such races, and which messages matter most to electorates.

Civis is Franklin's second job since graduating Dartmouth; he first worked for a year and a half as a Revenue Intelligence Analyst for a search engine marketing technology startup in Buenos Aires, Argentina.  After this job, Franklin attended a three month data science bootcamp in New York City. He distinctly credits his time learning experimental design and online surveys in Nyhan's Experiments in Politics in preparing him for his work now.

Civis is Junghye's first job post-graduation. She has been able to use skills she learned in her QSS classes, such as tidyverse from both Data Visualization and Race, Incarceration, and Politics; survey research techniques from Experiments in Politics and the Washington, DC, Foreign Study Program; and, machine learning concepts from Computational Text Analysis. Overall, Junghye feels fortunate to have been a part of QSS at Dartmouth and to have had the opportunity to think about social phenomena and apply quantitative skills to learn more about them-something she does daily in her job.  She is always happy to talk more about her experience in QSS and/or life at Civis and interested students should feel free to reach out at jkim2@civisanalytics.com.

QSS graduate on Joe Biden's ability to attract black and racially resentful voters

Alexander Agadjanian, a 2018 graduate of Dartmouth College's Program in Quantitative Social Science, who currently works at the MIT Election Lab, recently published an article in the Washington Post examining what's shaping candidate preferences in the Democratic presidential primary. He finds significant roles for voter prejudices, like racial resentment and sexism, in dividing Democratic voters and an interesting paradox as well: current leader and former vice president Joe Biden enjoys huge advantages among both African American voters and also among those high in anti-black prejudice. In this recent work, Agadjanian applied the data analysis and visualization skills he learned in QSS 17 Data Visualization, a course he took his third year at Dartmouth.

Since leaving New Hampshire, Agadjanian, who double-majored in QSS and Government, has been working on election data projects at his lab at MIT. He's also involved in numerous other research projects, including a recent academic publication with Dartmouth Professor of Government Brendan Nyhan and other seminar classmates, and a resently presented a paper at the 2019 Annual Meeting of the American Political Science Association in Washington, DC.  This paper uses cast vote records—also known as ballot logs---to study rates of ticket splitting, e.g., voting for a Democratic presidential candidate and a Repulican United States Senate candidate. More of Agadjanian's work is available on his website. In addition, Agadjanian has maintained a politics and data blog, where he puts his data analysis skills to use.

Agadjanian wrote an Honors Thesis in QSS, titled "Party Leader Influence and Conflicting Signals in Opinion Formation." In this work, Agadjanian explores the ability of political party leaders to influence public opinion. He finds that intra-party elite division — for example, when Republicans oppose Donald Trump on policy — significantly weakens a leader's (here, Trump's) persuasion ability. Policy information, on the other hand, is not particularly effective in remedying blind leader adherence: partisans still mostly "follow the leader" even after they're better informed about the policy they're evaluating. Agadjanian currently is in the process of revising his thesis for publication, and the latest version of it can be found here.