Curriculum

Quantitative research addressing societal issues increasingly relies on a combination of innovative statistical modeling, typically involving sophisticated computational methods, along with a deep understanding of social science, broadly conceived. To complete the requirements of the doctoral degree in Statistics and Computational Social Science, you will take two foundational courses, along with courses in statistics, computational methods, and the social sciences. 

Foundations:

• Data science for social impact
• Data ethics

Statistics; representative courses include:

• Probability
• Inference
• Experimental/Survey Design
• Generalized Linear and Mixed Models
• Causal Inference
• Specialized topics (e.g., Spatial Statistics, Networks, Multilevel Modeling)

Computational methods; representative courses include:

• Programming
• Machine Learning
• Data Structures & Algorithms
• Large/Messy Data
• Database Systems

Social sciences and related disciplines; courses chosen from areas including:

• Sociology
• Economics
• Political Science
• Psychology
• Education
• Health
• Public Policy 

Comprehensive Exams

You will take two written comprehensive examinations covering core areas of knowledge that underpin the application of statistics to research in the social sciences; it is expected that students pass both exams by the end of their second year in the program. The first exam will be a standard timed exam assessing knowledge of statistics (with specific topics including, e.g., causal inference, machine learning, probability, and inference) and computational methods (with specific topics including, e.g., optimization and analysis of algorithms). The second exam will take place over the course of a week, and will involve writing a hypothetical grant proposal outlining a research question, proposed data collection strategies, research design and analysis choices, and dissemination plans.

Qualifying Paper

To demonstrate depth of knowledge outside statistical and computational methodology, you will write a qualifying paper describing current knowledge and specific questions relevant to your “cognate discipline,” that is, a sub-area within a social science or related field, and pertaining to the planned dissertation topic. This paper will be evaluated by at least one professor outside the core SCSS faculty who is an expert in the cognate discipline, and must be completed before the dissertation proposal stage.

Research Activities

In addition to formal coursework, you will also engage in a variety of research-related activities each semester, beginning with the first semester on campus:

• Regular meetings with a faculty advisor to discuss ongoing research projects. Doctoral students are also welcome to work with faculty outside of the department with the approval of their advisor.
• Implementing a research plan to contribute to ongoing research projects
• Regular attendance in research seminars hosted by NYU’s PRIISM Center.
• Presentations of relevant literature, questions of interest, and ongoing research findings in research seminars to program faculty.
• Participation in a consulting project, such as a research-practice partnership, under the guidance of a faculty mentor.
• Preparation for comprehensive exams and qualifying paper.

Dissertation

The activities of research, coursework, seminars, comprehensive exams and the qualifying review paper will have exposed you to a wide range of faculty and their interests. By the third year in the program, you will have developed an independent research agenda that you can pursue with support from your advisor, and which will result in the completion of a dissertation.

The dissertation format for the Ph.D. in Statistics and Computational Social Science will follow a three-paper model, common in many social science disciplines, that codifies the interdisciplinary philosophy of this doctoral degree program. Each paper will address a different aspect of the same research topic — for example, one could be a review paper intended for the “cognate” discipline’s audience, the second could be a methods paper, incorporating both statistical and computational elements, and intended for a more technical audience, and the third could be an applied paper that demonstrates the utility of the method to practitioners. Together, the three papers should change the way in which we understand the world in a manner that was unattainable without the mixture of disciplines and related techniques. 

Committee

After completing comprehensive exams and the qualifying paper, you will choose three faculty members to serve on your dissertation committee, with one designated as committee chair. Given the interdisciplinary nature of the degree, at least one faculty member should represent the “cognate” discipline. Committee members will provide regular feedback on dissertations and dissertation proposals.

Dissertation Proposal

You will prepare, submit, and orally defend a manuscript research proposal, similar to a dissertation proposal.

Dissertation Defense

The manuscripts, taken together, must reflect a coherent and cohesive research line of inquiry and will be defended in a final oral defense for completion of the PhD. The presentation portion of the dissertation defense will be open to the public. By the date of the defense, at least one  first-authored manuscript must be under review at a peer-reviewed journal.