Week 1: Course Overview

Welcome to D\(^3\)! About me.

Image 1
I have my PhD in Agricultural Economics.
Image 2
My research focuses on food policy questions.
Image 3
My meditation practice is baking.
Image 4
I am a data junkie.

I’d like to get to know you, too.

Your first assignment (15 points):

Schedule a 15-minute one-on-one meeting with me. All meetings must take place before Feb 7th.

Book your one-on-one here.

Data is not the same as information. It is merely a collection of numbers and facts until we give it meaning. (Tim Harford, The Data Detective)

Let’s start with some examples.

Aldi leaves town ‘without a single grocery store’

Addressing Food Deserts with Data

How can communities improve access to affordable, healthy food? Data!

Using data-driven analysis, researchers and policymakers identify and address food deserts:

  • Map underserved areas by analyzing geographic and transportation data alongside income levels.
  • Predict how adding a new grocery store or market impacts household food access, shopping behaviors, and local competition.

For example, the abrupt closure of an ALDI store in West Pullman, Chicago in November 2024 left the neighborhood without a grocery store, exacerbating food desert conditions. Using data to proactively identify such risks can help communities and retailers address food insecurity before it worsens. (Source)

Explore Food Access in the U.S.

The Food Access Research Atlas allows users to visualize food access indicators across the United States, helping identify food deserts and areas with limited access to healthy foods.

Understanding food retail geography allows policymakers and research to answer questions like:

  • Which areas of the country are most affected by food deserts?

  • How might policy interventions or new retail locations address access challenges?

  • What demographic or economic factors correlate with low food access?

Empty shelves cost retailers billions

Preventing Stockouts with Inventory Management

Stockouts don’t just frustrate customers—they can cost businesses their loyalty.

Research shows that by the third time a product is out of stock, 70% of customers will not return to the store. (Source)

Retailers use data analytics to minimize stockouts by:

  • Accurately forecasting demand to avoid overstock or understock.
  • Tracking real-time inventory levels to ensure timely restocking.
  • Analyzing sales patterns to predict peak demand periods.

Effective inventory management is an application of D\(^3\) that helps businesses retain customers and maintain profitability.

Data-driven decision making is a process in which decisions are based on data and analysis rather than on intuition or personal experience.

The objective of this course is to introduce this process of transforming data into actionable insights.

This process requires understanding how to read figures (e.g., graphs, charts, and maps), discern between “good” figures and “bad” figures, and tell effective stories using data-derived figures.

Let’s discuss these figures.

  1. What does the figure tell you?

  2. Is it clear what the author’s did to produce the figure?

  3. Was the research finding a shock or exactly as you expected?

 

Example 1: Good Data Project

Example 2: Sustainable Happiness

Example 3: Mapping Segregation

Sharpening your data visualization tools

By the end of this course, you will learn skills to produce high-quality figures by processing large datasets using R and Tableau.

Example 1: Student 1

Example 2: Student 2

Agenda

 

Syllabus and course orientation

 

Introduction to D\(^3\)M

 

Introduction to R

 

Introduction to Tableau

How to succeed in this course

  1. Attend class lectures (W) and labs (F) (5% of grade).
  2. Don’t miss the quizzes (20% of grade). *Lowest quiz grade is dropped.
  3. Review project assignments in advance. Each project is broken up into weekly problem sets. There are four projects (75% of grade).

Generative AI and Coding

  • You may use generative AI tools (e.g., ChatGPT) as coding assistants to help solve problems or to help you understand the software used in this course.
  • However, if you choose to use such tools, you are claiming the resulting work as your own, and you bear full responsibility for its accuracy and appropriateness.
  • You must disclose if AI tools were used and how they were applied.
    • Example: “Used ChatGPT to debug this code and suggest ways to improve its readability.”
  • Disclosure will not harm your grade and helps me guide you in using AI effectively.

Reminder: Generative AI cannot advocate for itself or attribute content appropriately. Integrity in your submissions is your responsibility.

Office Hours

Dr. Chenarides’ Office Hours: Wednesday 2-3 pm, Rm 201

Diya’s (TA) Office Hours: Friday 2:45-3:45 pm, Rm 165

D\(^3\)M

 

Data-driven decision making is a process in which decisions are based on data and analysis rather than on intuition or personal experience. It involves collecting data, analyzing it, and using the insights gained from the analysis to inform the decision-making process. The goal of data-driven decision making is to use data to make more informed, objective decisions that are likely to lead to better outcomes. This approach is based on the idea that data can provide a more accurate and complete picture of a situation than personal experience or intuition alone. Data-driven decision making is often used in business and government to improve efficiency, reduce costs, and make more effective decisions.

— ChatGPT

D\(^3\)M Introduction Exercise

Turn to a neighbor and…

  1. Introduce yourself - name, major, fun fact.
  2. Find some common ground - find one shared characteristic. This must be something you did not previously know and something that is not visible.
  3. Discuss real-world examples of D\(^3\)M with your neighbor.
  4. Share one example related to ag or natural resources with the class.

Google Human Resources

Google’s Project Oxygen: Google surveys employees about their managers and connects the responses to team output to figure out what attributes or management practices lead to more productive teams.

 

Google’s PiLab: Google conducts applied experiments to determine the most effective approaches for managing people, including increasing productivity, increasing happiness, and improving health.

 

Diversity at Google: Google’s people analytics team conducts analyses to identify the root causes of weak diversity recruiting, retention, and promotions. These analyses have led to efforts to reduce bias in job descriptions, develop diversity-oriented mentorship programs, and produce annual diversity reports.

Amazon

Amazon uses a dizzying array of metrics to track your behavior with their platform in order to recommend products that they think you will buy.

Amazon Poll 1

     

What information does Amazon use to collect data on users?

Amazon Poll 1

What information does Amazon collect from users?

  • Purchases

  • Wishlist items

  • Shopping cart saves

  • IP address, login credentials, computer location

  • Clicked URLs; Mouse hovering

  • Timing of scrolling and clicking

  • Alexa

  • Kindle highlighting

Amazon Poll 2

     

How does Amazon use the data it collects? (in other words: what decisions does data inform?)

Amazon Poll 2

     

How does Amazon use the data it collects?

  • Personalized recommendation system

  • Anticipatory shipping model

  • Book recommendations

  • Prices (Amazon changes their prices an average of 2.5 million times a day)

  • Suggested add-ons

  • More scrolling!

Summary and Reminders

 

What is D\(^3\)M?

 

What are the course objectives?

 

Up Next: What is R? What is Tableau?

What is R?

What is Tableau?