Stat 470/670: Exploratory Data Analysis

Logistics

Meetings: Tuesdays and Thursdays, 12:45-2pm, WH 111

Website: http://jfukuyama.github.io/teaching/stat670

Instructor: Prof. Julia Fukuyama		jfukuyam at iu dot edu
Office hours: Tuesdays 2:30-3:30pm, Wednesdays 1-2pm Swain East 225
Associate Instructor: Mr. Yue Yu		yyu3 at iu dot edu
Office hours: Mondays and Fridays 1-3pm		Swain East 200

Course Overview

Graphical and modeling techniques for exploring data, with an emphasis on visualization, interpretation, and clear communication of findings. Use of modern software tools for data manipulation and visualization. Connections to traditional statistical methods.

Textbooks

We will be drawing heavily on Cleveland’s Visualizing Data and Hadley Wickham’s ggplot2: Elegant Graphics for Data Analysis. Both of these are available online through the IU library.

Also useful will be R for Data Science by Wickham and Grolemund, available online.

Readings and notes for topics not covered in the textbooks will be posted to the course website and to canvas.

Class Structure

Classes will be a combination of lecture and tool demonstration. It will generally be helpful for you to have an R session open to follow along wth the code. Slides or notes, with R code, will be posted to the class website before each lecture.

We will also have a regular set of in-class exercises where you will try out some of the sorts of analyses we have recently covered in class, think through why we would want to perform such an analysis, and think through what the implications of the analysis might be.

Assessment

Grades will be assigned based on:

In-class exercises and presentations, worth 30% of the grade. By the second week of class, you should identify a dataset. A couple of times over the course of the semester, we will take some time in class for you to (1) apply some of the techniques we have talked about to your dataset, and (2) think about why someone would want to do that analysis and what the implications of that analysis would be. You will be required to upload your analysis and a response to the question of ``why would you do this analysis’’ to canvas. You will also be required to present your results at least one time over the course of the semester.
Mini project, worth 30% of the grade. This will involve more substantial data analysis and a writeup.
Final project, 40% of the grade, on a dataset and question of your choice.

There will be no final exam; the last responsibility for the course will be the report for the final project due on the last day of class.

The projects will be graded on how well the material is presented in addition to accuracy. This means there should be no extraneous material, plots should be readable, and text and figures should be formatted nicely.