Getting Started

EMSE 6035: Marketing Analytics for Design Decisions

John Paul Helveston

August 26, 2026

Week 1: Getting Started

1. Course orientation

2. Intro to conjoint analysis

3. Introductions

4. Getting started with R & Positron

Week 1: Getting Started

1. Course orientation

2. Intro to conjoint analysis

3. Introductions

4. Getting started with R & Positron

Meet your instructor!

John Helveston, Ph.D.

2025 - Present: Associate Professor, EMSE
2018 - 2025: Assistant Professor, EMSE
2016-2018: Postdoc at Institute for Sustainable Energy, Boston University
2016: PhD in Engineering & Public Policy at Carnegie Mellon University
2015: MS in Engineering & Public Policy at Carnegie Mellon University
2010: BS in Engineering Science & Mechanics at Virginia Tech
Website: www.jhelvy.com

Tools

Course website: https://madd.seas.gwu.edu/2026-Fall/

Course slack: https://emse-madd-f26.slack.com

& Positron: Course Software Page

Why ?

Wait, why aren’t we using Python?

The Marketing field out of Statistics, not Machine Learning

Python is a general purpose language developed by Guido van Rossum, a computer scientist.
Unlike R, Python was not originally developed for data analysis.
Both languages are extremely useful, and you should probably learn python too.

Learning Objectives

After this class, you will know how to…

…work with data in
…design effective surveys to get rich data
…analyze consumer choice data to model consumer preferences
…effectively communicate insights
…use Git and GitHub as everyday version control and collaboration tools
…direct agentic AI tools to write and audit code that supports analyses

Course prerequisites

This course requires prior exposure to:

Probability theory
Multivariable calculus
Linear algebra
Regression

Not sure?

Take this self assessment

Weekly HW (30% of grade)

Do some readings, recorded lectures, practice problems

Write a short reflection

~Every week (13 total)

Due 11:59pm Monday before class

Graded for completion (looking for engagement)

Quizzes (10% of grade)

At the start of class every other week-ish. Make ups only for excused absences (i.e. don’t be late).

6 total, lowest dropped

10 minutes

Why quiz at all? The “retrieval effect” - basically, you have to practice remembering things, otherwise your brain won’t remember them (see the book “Make It Stick: The Science of Successful Learning”)

Exam (10% of grade)

In-person, first 45 minutes of class on 12/2

All hand-written (like a long quiz)

Semester Project (45% of grade)

Teams of 3-4 students

Goals:

Assess market viability of a new technology or design
Recommend best design choices for target market or application

Key deliverables:

Item	Weight	Due
Project Proposal	5 %	Sep. 21
Survey Plan	5 %	Sep. 30
Pilot Survey	5 %	Oct. 14
Pilot Analysis	5 %	Nov. 02
Final Survey	5 %	Nov. 16
Final Analysis Report	15 %	Dec. 07
Final Presentation	5 %	Dec. 09

Grades

Item	Weight	Notes
Participation / Attendance	5%	(Yes, I take attendance)
Homework	30 %	13 assignments (12 x 2.5%, lowest dropped)
Quizzes	10 %	6 quizzes, lowest dropped
Final Exam	10 %	In class, closed book (12/2)
Project Proposal	5 %	Teams of 3-4 students
Survey Plan	5 %
Pilot Survey	5 %
Pilot Analysis	5 %
Final Survey	5 %
Final Analysis Report	15 %
Final Presentation	5 %

Course policies

BE NICE
BE HONEST
DON’T CHEAT

Copying is good, stealing is bad

“Plagiarism is trying to pass someone else’s work off as your own. Copying is about reverse-engineering.”

– Austin Kleon, from Steal Like An Artist

Use of AI

We will heavily use AI tools this semester

Large language models (LLMs) are pretty good
Sometimes they suck.
I will grade whatever you submit. It should not suck.

Ways to not have your work suck:

Don’t submit code that doesn’t run (actually run it before submitting it).
Actually read what the AI generates, and don’t submit something you don’t understand. (Ask the LLM to explain why something works or doesn’t work)

Use the approach I teach: agentic workflows (starting next week)

Late submissions

- 3 late days - use them anytime, no questions asked

- No more than 2 late days on any one assignment

- Contact me for special cases

How to succeed in this class

Participate during class!

Start assignments early and read carefully!

Get sleep and take breaks often!

Ask for help!

Getting Help

Use Slack to ask questions.

Schedule a meeting w/Prof. Helveston:

Mondays from 8:00-4:30pm
Tuesdays from 8:00-4:30pm
Fridays from 8:00-4:00pm

GW Coders

Week 1: Getting Started

1. Course orientation

2. Intro to conjoint analysis

3. Introductions

4. Getting started with R & Positron

Engineers often design things nobody wants!

We want to answers to questions like…

- Higher prices decrease demand, but by how much?

- How much more is a consumer willing to pay for increased performance in X?

- How will my product compete against competitors in the market?

Answers depend on knowing what people want

Directly asking people what they want isn’t always helpful

(People want everything)

Which feature do you care more about?

Battery Life?

Brand?

Signal quality?

Conjoint approach:
Use consumer choice data to model preferences

Use random utility framework to predict probability of choosing phone j

1. $u_j = \beta_1\mathrm{price}_j + \beta_2\mathrm{brand}_j + \beta_3\mathrm{battery}_j + \beta_4\mathrm{signal}_j + \varepsilon_j$

2. Assume $\varepsilon_j \sim$ iid extreme value

3. Probability of choosing phone j: $P_j = \frac{e^{\beta'x_j}}{\sum_k^J e^{\beta'x_k}}$

4. Estimate $\beta_1$, $\beta_2$, $\beta_3$, $\beta_4$ by minimizing $-L = - \sum_n^N \sum_j^J y_{nj} \ln P_{nj}$

Willingness to Pay

$u_j = \beta'x_j + \alpha p_j + \varepsilon_j$

$\omega = \frac{\beta}{-\alpha}$

Respondents on average are willing to pay $XX to improve battery life by XX%

Make predictions

$P_j = \frac{e^{\hat{\beta}'x_j}}{\sum_k^J e^{\hat{\beta}'x_k}}$

Example: Pocket Charge

A Flexible, Portable Solar Charger

Example survey choice question

Your project starts now!

Click here to view project ideas

Week 1: Getting Started

1. Course orientation

2. Intro to conjoint analysis

3. Introductions

4. Getting started with R & Positron

Introduce yourself

- Preferred name

- Degree program

- Prior experience

- What do you hope to gain from this class?

- Project interests?

Break

If you haven’t already, install everything on the software page
Add your GitHub username to this sheet (you need this to get your course repo!)
Stand up, meet each other, (maybe form teams?…use this sheet)

05:00

Week 1: Getting Started

1. Course orientation

2. Intro to conjoint analysis

3. Introductions

4. Getting started with R & Positron

This semester we use Positron

(not RStudio)

Positron Orientation

Open `intro-to-R.R` file and follow along

Staying organized

1) Save your code in .R files

3) Always open your folder from Positron

10:00

Your turn

A. Practice getting organized

Make a new folder called week1, then open it in Positron (File > Open Folder…).
Create a new R script and save it as practice.R.
Open the practice.R file and write your answers to these questions in it.

B. Creating & working with objects

1). Create objects to store the values in this table:

City	Area (sq. mi.)	Population (thousands)
San Francisco, CA	47	884
Chicago, IL	228	2,716
Washington, DC	61	694

Using the objects you created, answer the following questions:

Which city has the highest density?
How many more people would need to live in DC for it to have the same population density as San Francisco?

>20,000 packages on the CRAN

Installing packages

`install.packages("packagename")`

(The package name must be in quotes)

install.packages("packagename") # This works
install.packages(packagename) # This doesn't work

You only need to install a package once!

Loading packages

`library(packagename)`: Loads all the functions in a package

(The package name doesn’t need to be in quotes)

library("packagename") # This works
library(packagename) # This also works

**You need to load the package every time you use it!**

Installing vs. Loading

Example: wikifacts

Install the Wikifacts package, by Keith McNulty:

install.packages("wikifacts")

Load the package:

library(wikifacts) # Load the library

Use one of the package functions

wiki_randomfact()

#> [1] "Did you know that the Romans built a temple possibly dedicated to a boar cult within the prehistoric Chanctonbury Ring hillfort? (Courtesy of Wikipedia)"

Example: wikifacts

Now, restart your R session in Positron: > Session -> Restart R

Try using the package function again:

wiki_randomfact()

#> Error in `wiki_randomfact()`:
#> ! could not find function "wiki_randomfact"

Using only some package functions

You don’t always have to load the whole library.

Functions can be accessed with this pattern:

packagename::functionname()

wikifacts::wiki_randomfact()

#> [1] "Here's some news from 05 May 2026. The Antigua and Barbuda Labour Party, led by Prime Minister Gaston Browne (pictured), wins an increased majority in the general election. (Courtesy of Wikipedia)"

If you haven’t yet, install these packages

Back `intro-to-R.R` for the rest of class!

Week 1: Getting Started

1. Course orientation

2. Intro to conjoint analysis

3. Introductions

4. Getting started with R & Positron

Week 1: Getting Started

1. Course orientation

2. Intro to conjoint analysis

3. Introductions

4. Getting started with R & Positron

Meet your instructor!

Tools

Why ?

Wait, why aren’t we using Python?

Learning Objectives

After this class, you will know how to…

…work with data in

…design effective surveys to get rich data

…analyze consumer choice data to model consumer preferences

…effectively communicate insights

…use Git and GitHub as everyday version control and collaboration tools

…direct agentic AI tools to write and audit code that supports analyses

Course prerequisites

Probability theory

Multivariable calculus

Linear algebra

Regression

Weekly HW (30% of grade)

Do some readings, recorded lectures, practice problems

Write a short reflection

Quizzes (10% of grade)

Exam (10% of grade)

In-person, first 45 minutes of class on 12/2

All hand-written (like a long quiz)

Semester Project (45% of grade)

Grades

Grades

Course policies

BE NICE

BE HONEST

DON’T CHEAT

Use of AI

We will heavily use AI tools this semester

Ways to not have your work suck:

Late submissions

- 3 late days - use them anytime, no questions asked

- No more than 2 late days on any one assignment

- Contact me for special cases

How to succeed in this class

Getting Help

Week 1: Getting Started

1. Course orientation

2. Intro to conjoint analysis

3. Introductions

4. Getting started with R & Positron

We want to answers to questions like…

- Higher prices decrease demand, but by how much?

- How much more is a consumer willing to pay for increased performance in X?

- How will my product compete against competitors in the market?

Answers depend on knowing what people want

Directly asking people what they want isn’t always helpful

(People want everything)

Which feature do you care more about?

Conjoint approach:Use consumer choice data to model preferences

Use random utility framework to predict probability of choosing phone j

Example: Pocket Charge

A Flexible, Portable Solar Charger

Example survey choice question

Your project starts now!

Click here to view project ideas

Week 1: Getting Started

1. Course orientation

2. Intro to conjoint analysis

3. Introductions

4. Getting started with R & Positron

Introduce yourself

- Preferred name

- Degree program

- Prior experience

- What do you hope to gain from this class?

Conjoint approach:
Use consumer choice data to model preferences

Open `intro-to-R.R` file and follow along

`install.packages("packagename")`

`library(packagename)`: Loads all the functions in a package

**You need to load the package every time you use it!**

Back `intro-to-R.R` for the rest of class!