Meaghan Winder

Introduction to Data Science (formerly MATH 0118)
In this course students will gain exposure to the entire data science pipeline: forming a statistical question, collecting and cleaning data sets, performing exploratory data analyses, identifying appropriate statistical techniques, and communicating the results, all the while leaning heavily on open source computational tools, in particular the R statistical software language. We will focus on analyzing real, messy, and large data sets, requiring the use of advanced data manipulation/wrangling and data visualization packages. Students will be required to bring a laptop (owned or college-loaned) to class as many lectures will involve in-class computational activities. (formerly MATH 0216) 3 hrs lect./disc. (Not open to students who have taken BIOL 1230, ECON 1230, ENVS 1230, FMMC 1230, GEOG 1230, HARC 1230, JAPN 1230, LNGT 1230, MATH 1230, NSCI 1230, PSCI 1230, SOCI 1230, or WRPR 1230.)

Regression Theory and Applications (formerly MATH 0211)
Regression is a popular statistical technique for making predictions and for modeling relationships between variables. In this course we will discuss the theory and practical applications of linear, log-linear, and logistic regression models. Topics include least squares estimation, coding for categorical predictors, analysis of variance, and model diagnostics. We will apply these concepts to real datasets using R, a statistical programming language. (Concurrent or prior MATH 0200, and STAT 0116 or STAT 0201 or PSYC 0201 or ECON 0111) (Not open to students who have taken ECON 0211.) 3 hrs lect./disc.

Statistical Inference
An introduction to the mathematical methods and applications of statistical inference using both classical methods and modern resampling techniques. Topics will include: permutation tests, parametric and nonparametric problems, estimation, efficiency and the Neyman-Pearsons lemma. Classical tests within the normal theory such as F-test, t-test, and chi-square test will also be considered. Methods of linear least squares are used for the study of analysis of variance and regression. There will be some emphasis on applications to other disciplines. This course is taught using R. (Concurrent or prior MATH 0200, and MATH/STAT 0310) 3 hrs. lect./disc.

Data Science Across Disciplines
In this course, we will gain exposure to the entire data science pipeline—obtaining and cleaning large and messy data sets, exploring these data and creating engaging visualizations, and communicating insights from the data in a meaningful manner. During morning sessions, we will learn the tools and techniques required to explore new and exciting data sets. During afternoon sessions, students will work in small groups with one of several faculty members on domain-specific research projects in Biology, Interdepartmental, Political Science, and Statistics. This course will use the R programming language. No prior experience with programming is necessary. (Not open to students who have taken STAT 0118, STAT 0201, ANTH 1230, BIOL 1230, CLAS 1230, ECON 1230, ECSC 1230, ENVS 1230, FMMC 1230, GEOG 1230, HARC 1230, HIST 1230, INTD 1230, JAPN 1230, LNGT 1230, MATH 1230, MUSC 1230, NSCI 1230, PSCI 1230, SOCI 1230, STAT 1230, or WRPR 1230.)

BIOL 1230: Why does one island in the Gulf of Maine have red foxes and its neighbor none? In this section we will work with historical museum records and present-day community-science observations of mammals collected on Maine’s coastal islands. Islands are ecology’s classic natural experiment: their size and their distances from the mainland can shape which species arrive, which persist, and which blink out. Students will organize and visualize these collected records to test predictions first made in the 1980s about what influences species richness, colonization, and extinction. Beyond answering interesting ecological questions, these analyses will contribute to community engaged work with regional conservation organizations. Final projects will follow a question of each student’s own choosing, from a single species to a whole archipelago.

INTD 1230: How does artificial intelligence shape the way we learn and how can we understand those impacts? Drawing on research from the learning sciences and critical technology studies, we will explore the questions of what we gain and what we lose when we rely on AI to help us think, write, study, and create. We will explore how data science can help us understand the impacts of AI on learning and where it might fall short. Students will analyze data related to the use of AI in learning at Middlebury, including patterns of use and beliefs about AI in learning, and human vs. AI writing.

MUSC 1230: Music in its various forms (listening, playing, writing) establishes a series of psychological expectations within the human mind that, when challenged or affirmed, can invoke emotions, encourage movement, and even trigger latent memories. In this course, we will leverage the power of data science to unearth music’s expectational features in large datasets — corpora — of digitally encoded music. Students will learn how music is translated into machine-readable formats, practice parsing databases for categorical information, and generate accessible visualizations of musical-feature distributions. Such activities will culminate in a poster project where students investigate a specific research question using course corpora. In doing so, students will come to better understand the nature of psychological expectations as they shape everyday musical experiences.

PSCI 1230: What do young Americans think about democracy, political institutions, and public issues? How polarized are these views? How do these views compare with those of older generations in the United States and people in other countries? In this session, we will use the tools of data science and cross-national survey data to explore these and other questions about young people's political attitudes. The session will also introduce students to the basics of survey research and the study of public opinion. Students will complete a final project showcasing the concepts and tools learned in class.

STAT 1230: In this course, students will experience the world of data science by working with species abundance data. Species abundance data is crucial to ecologists for tracking biodiversity, developing tools for community outreach, and informing conservation planning; however, perfect data collection is expensive and often impossible in ecology. Students will thus learn the tools and techniques required to process this data. Specifically, students will gain experience with all steps of the data science pipeline from data scraping and wrangling to visualizing and performing basic statistical inference. Statistical topics we will work with may include clustering, regression, and experimental design. Throughout these topics, communicating results to all audiences clearly and honestly will be an overarching theme.