INSH 6406 Analyzing Complex Digitized Data
Northeastern University, Fall 2024, Fall 2025, Fall 2026
Instructor
This course trains graduate researchers to use computational methods for analyzing large-scale text, image, and multimodal data as part of their own social science research. The goal is not to train model builders: it is to build sophisticated, critical directors and evaluators of these methods, since generative AI now makes it possible to have a machine gather, structure, and analyze web-scale data that no research team could code by hand. Digitized data such as news articles, social media posts, and images is increasingly the raw material of empirical social science, and this course treats AI-assisted analysis as the default way researchers now work with it, not an advanced add-on. For every computational method covered, from text preprocessing through topic modeling, classification, embeddings, and multimodal analysis, students learn three things: (i) what the method is and what it assumes about the data, so that it can be recognized and understood in a published methods section; (ii) where and why the method commonly fails or misleads, so that its output can be judged critically; and (iii) how to direct an AI system to carry out the method correctly and how to verify, using concrete evaluation standards such as precision, recall, and inter-rater reliability, that the resulting output is trustworthy enough to support a research claim. Students will apply these skills across the semester to a real, instructor-provided dataset (a multilingual news corpus and social media data), working individually at first and then joining a small collaborative research team for an original project, culminating in a team-authored paper and presentation aimed at conference or journal submission. By the end of the course, you will be able to read a computational social science paper and critically evaluate its methods, direct AI tools to carry out complex text and image analysis responsibly, and judge whether a given computational measure is a valid and defensible way to answer your research question.
COMM 2105 Social Networks
Northeastern University, Spring 2024, Spring 2025, Spring 2026, Spring 2027
Instructor
In this course, we explore the fascinating world of social networks, going beyond social media to understand the expansive networks that shape our lives—from personal relationships to professional interactions. Through social network theories and analytical methods, this course aims to decode the intricate web of connections that orchestrate the world we live in. You'll learn to look at the world like a network, which will help you see how everything is connected. As we navigate these themes together, we will uncover the dynamics of network formation and explore the profound impact that these structures have on individual behavior, belief systems, and access to opportunities. By the end of the course, you will gain insights into optimizing personal networks, understanding what makes influential figures like Steve Jobs impactful, and comprehending phenomena such as the spread of pandemics, the rise of social movements, and Internet polarization.
INSH 6500 Statitscial Analysis
Northeastern University, Fall 2023, Spring 2024, Spring 2025, Spring 2027
Instructor
INSH 6500 is an introductory course in probability and statistics, specifically tailored for graduate students at the College of Social Sciences and Humanities (CSSH). The main objective of this course is how social ‘scientists’ see and leverage data to understand human behaviors and society. We will cover a range of topics including data visualization through tables and graphs, descriptive statistics, probability, sampling distributions, estimation, hypothesis testing, and the analysis of relationships among variables, such as regression analysis. This class will help you establish a solid groundwork for comprehensive understanding and proficiency in statistical inference and regression, which will be explored extensively in INSH 7500 or an equivalent graduate-level statistics course. The course focuses on teaching you when and how to apply various statistical methods, while also helping you become skilled in computer-based analysis and interpretation. Students who take this course will acquire a comprehensive understanding and a set of skills in statistical analysis and empirical data handling across various social dimensions.
POLS 2400 Quantitative Techniques
Northeastern University, Spring 2023, Summer2 2024, Summer2 2025, Spring 2026, Summer1&2 2026
Instructor
POLS2400 serves as an introduction to quantitative methods in political science. This course introduces students how political ‘scientists’ see and leverage data to understand human behaviors and society. It covers a range of topics including data visualization through tables and graphs, descriptive statistics, probability, sampling distributions, estimation, hypothesis testing, and the analysis of relationships among variables, such as regression analysis. There are various methods to teaching statistics. Some instructors focus heavily on formulas and calculations, while others concentrate on interpreting computer-generated results. Some emphasize procedural techniques, whereas others prioritize conceptual comprehension. In this course, I aim to integrate and balance all these approaches. Performing certain calculations is essential for understanding the logic behind statistical tools, although you won't be required to memorize any formulas. The course emphasizes understanding when and how to use different statistical procedures and ensures you become proficient in computer-based analysis and interpretation.
POLITSC 7552 Quantitative Political Analysis II
Ohio State University, Spring 2022
Instructor
This course is doctoral-level course of quantitative methods to learn a family of the linear and generalized linear modeling using maximum likelihood estimation (MLE) and Bayesian estimation. The primary purpose of the course is to build on statistical foundations taught in PoliSci 7551. We will perform regression analysis with the following types of outcome variables: continuous, counts, dichotomous outcomes, ordered categorical outcomes, unordered categorical outcomes, bounded variables, and multilevel variables.