bit(atom(bit)) empirical tom class

[hyunjimoon]

description of knowledge creation journey of PhD seems similar to operations for startups (my research topic). Hence, I already learned a lot by translating it with Charlie's NSS framework. Your goal12 can help me complete nail stage, goal34 can help me scale. Once the class ends, I'd like to ask, if I may, some advice on sailing.

To be specific, if I could answer myself what data to collect, who to collaborate with, which journal to publish my first work, after taking your course, it'd be rewarding. Can you help me achieve the three goals?

nail stage

• test desirability and technical feasibility of our research question
• iterate between phenomena, theory, measurement to nail our value proposition that satisfies us, customer (paper reader or our future students), investor (advisor)

scale stage

• test operational feasibility and viability of our research before investing significant time and energy
• collaborate with equipment providers (data and tool providers), sampled customers (trusted colleagues and faculty), investor (referee), using and to develop our profession
• segment customer and set beachhead (research outlet's audience)
• build community both from supply and demand side and acculturate our vision, share progress in the language of aligned evaluation metric

sail stage

• processify the above that helps us automate, replicate, platformize, capitalize the knowledge creation itself

[hyunjimoon]

syllabus

Empirical Technology and Operations Management

INFORMATION
Faculty: Kris Ferreira
Office: Morgan 413
Email: [email protected]
Faculty Support Specialist: Liat Shapiro ([email protected])
Faculty: Toni Moreno
Office: Morgan 493
Email: [email protected]
Faculty Support Specialist: Tamsin Priestley ([email protected])
Day/Time: Fridays, 12:30-3:30pm. January 26 – April 26; no class on March 15 or April 19.
Class location: Morgan 450.

PREREQUISITES

You should be familiar with basic linear algebra and basic statistics and econometrics concepts. The course does not assume that you have taken a graduate-level econometrics course, but having taken one (or taking one concurrently) will be helpful. The pre-requisites are similar to the ones of a graduate-level econometrics course. You should be able to read materials at the level of Jeffrey Wooldridge’s Econometric Analysis of Cross-Section and Panel Data, but we do not expect you to have covered those materials.
The primary audience of this course is doctoral students in technology and operations management. Please contact the instructors with any questions.

GOALS

This course has four main goals. First, this course is designed to equip you with a basic skill set of empirical methods for use in technology and operations management (and interdisciplinary research of operations management and information systems, marketing and other areas). The skill set refers to a variety of methods commonly employed in empirical research. The sessions are designed to make you feel comfortable with the basic aspects of the methodologies used in the papers, even if you have limited background in econometrics, machine learning, statistics, or other methodologies. However, if you plan to do empirical research, you should definitely take econometric and other methods courses and applied empirical courses from other departments as well.
Second, the course aims to provide you with a working knowledge of some of the most well-known and recent empirical research in technology and operations management. The goal is to expose you to some well-implemented and well-known examples for each method, and to the frontiers of empirical technology and operations management research. You will be required to read and lead some paper discussions. This exercise will reinforce the understanding of the material and provide opportunities to practice presenting to an audience.
The third (and probably the most important) goal of this course is to help you develop taste in research and develop your ability to conduct interesting research, which is also the goal of doctoral training. During your courses, you will build a strong methodological foundation if you work hard enough and we teach carefully. However, a good taste in research cannot be taught in a lecture, and we aim to help you develop it in this seminar-style course. You are in a doctoral program not to learn knowledge, but to create knowledge. In this course we will try to advance in this dimension by systematically discussing research ideas and aspects related to building research taste and developing the “craft” of research, and by giving feedback on a research project that you have to develop during the course (ideally, this could evolve to eventually become one of your dissertation essays). You should also develop the habit of consulting trusted audiences (your colleagues and faculty) once you come up with a research idea before you choose to invest significant time and energy into it.
Finally, along the way we will try to discuss some institutional details of the academic community in technology and operations management. For example, what type of research different outlets publish, how the refereeing process work, the main conferences, job market, etc. (of course, this does not substitute the valuable information that you can obtain from your advisors and doctoral coordinator!).
What this course is NOT: This course is not intended to replace the training you should be receiving from standard econometric courses. Due to the limited time we have, we will move rather quickly through the material.

COURSE STRUCTURE

Most classes will be centered around an empirical methodology commonly used in TOM research – e.g., machine learning, instrumental variables, field experiments, lab experiments, differences in differences, regression discontinuity, matching, etc. – and additional assignments, activities, and discussions will be had to achieve the learning objectives outlined above that are beyond specific methodologies. We expect you to have read the materials made available to you (book chapters and articles) prior to coming to each class.
Most sessions comprise three parts aligned with the aforementioned goals. In the first part we will discuss the fundamental ideas related to the method. The second part will be paper discussions led by students (typically two to three presentations related to the ). The third part will involve some “taste-building”, development of individual research projects, and discussion about the academic community in operations management.

REQUIREMENTS

• Problem sets and assignments: 15%. We will have 3 or 4 short assignments or problem sets.
• Paper discussions: 15%. Each student will have to lead the discussion for several papers.
• Final paper/presentation: 40%. Each student will have to produce a substantial research paper on a topic approved by the instructors. There will be several opportunities to get feedback on potential ideas and also a midterm and final presentation.
• Class participation: 30%. We expect everyone to come prepared to class and actively participate.

MATERIALS

There is no required text but we will make note of the appropriate references for each module. Some of the texts to which we refer are:
• Wooldridge, Jeffrey M., 2010, Econometric Analysis of Cross-Section and Panel Data, Second Edition, MIT Press, Massachusetts.
• Angrist, Joshua D., and Jorn-Steffen Pischke, 2009, Mostly Harmless Econometrics, Princeton University Press, New Jersey.
• Greene, William, 2011, Econometric Analysis, Seventh Edition, Prentice-Hall.
• Cameron, A. Colin, and Pravin K. Trivedi, 2005, Microeconometrics: methods and applications.Cambridge university press.
• Imbens and Rubin, 2015. Causal Inference in Statistics, Social and Biomedical Sciences. An Introduction. Cambridge University Press.
• Rosenbaum, 2017. Observation and Experiment. An Introduction to Causal Inference.Harvard University Press.

COURSE LIST

1. Introduction to Empirical TOM

INTRODUCTION
In this session we will cover the following topics:
• We will discuss the historic evolution of the Technology and Operations Management field
• We will discuss the taxonomy of empirical/data-driven approaches to research in Technology and Operations Management
• We will discuss a paper in detail
MATERIALS
Reading List: Empirical/data-driven research in OM
• *Fisher, M. (2007) Strengthening the Empirical Base of Operations Management. Manufacturing and Service Operations Management. 9(4) 368-382. [Link]
• Van Mieghem, J. A. (2013). OM Forum-Three Rs of Operations Management: Research, Relevance, and Rewards. Manufacturing & Service Operations Management, 15(1), 2-5. [Link]
• Simchi-Levi, D. (2014). OM Forum--OM Research: From Problem-Driven to Data-driven. Manufacturing & Service Operations Management, 16(1), 2-10. [Link]
• *Christian Terwiesch (2019). Empirical Research in Operations Management: From Field Studies to Analyzing Digital Exhaust, 21(4), 713-722 [LinkLinks to an external site.]
• *Terwiesch, Christian, Marcelo Olivares, Bradley R. Staats, and Vishal Gaur. (2020). OM Forum—A Review of Empirical Operations Management over the Last Two Decades. Manufacturing & Service Operations Management 22(4), 656-668. [Link]
• *Hummy Song, Anita L. Tucker, Karen L. Murrell, David R. Vinson (2017), Closing the Productivity Gap: Improving Worker Productivity through Public Relative Performance Feedback and Validation of Best Practices, Management Science [Link]
ASSIGNMENTS

1. Identify two papers in the technology and operations management or related areas that you consider particularly good or interesting (they do not need to be empirical, although it is preferrable). They can be classic papers or more recent ones. Be ready to discuss why you chose those papers and what you like about them. Please upload a slide about each of the papers by 11:00am on Friday 1/26/24.
2. Read the papers in the reading list. We expect you to read all the papers. The ones with * should be read in enough detail that you would be comfortable being call-called to discuss them.

2. Machine Learning

Introduction

Machine learning is a powerful tool used for prediction, yet this "prediction" objective does not necessarily guarantee causality - another common objective that we have as researchers. We'll start the class with a nuanced discussion of these two objectives, and then we'll discuss examples of how machine learning is used in various ways in TOM.

Materials
Shmueli, Galit. (2010). "To Explain or To Predict?" Statistical Science, 25(3): 289-310. linkLinks to an external site.

Ferreira, Kris J., Bin Hong Alex Lee, and David Simchi-Levi. (2016). "Analytics for an Online Retailer: Demand Forecasting and Price Optimization." Manufacturing & Service Operations Management, 18(1): 69–88. link

Mejia, Jorge, Shawn Mankad, and Anandasivam Gopal. (2019). "A for effort? Using the crowd to identify moral hazard in New York City restaurant hygiene inspections." Information Systems Research, 30(4): 1363-1386. link

Ferreira, Kris J., Sunanda Parthasarathy, and Shreyas Sekar. (2022). "Learning to rank an assortment of products." Management Science, 68(3): 1828-1848. link

Bertsimas, Dimitris, Jean Pauphilet, Jennifer Stevens, and Manu Tandon. (2022). "Predicting inpatient flow at a major hospital using interpretable analytics." Manufacturing & Service Operations Management, 24(6): 2809-2824. link

Assignment
Please read each of the five papers listed in the Materials section above, and come prepared to discuss. For the paper titled "To Explain or To Predict?", I'd like you to keep in mind that the "to predict" objective described in the paper is the objective most commonly used in machine learning models in research and practice in our field, and the "to explain" objective described in the paper is the causal objective commonly used in our field. For the remaining four papers, please focus primarily on the machine learning aspects of the papers, and consider the role that machine learning plays in the overall research process and contribution. Please don't worry about understanding all the technical details of each paper, and you're certainly welcome to skip appendices of proofs!

3. Observational Studies & Instrumental Variables

4. Presenting Research Ideas

5. Field Experiments

6. Lab Experiments

7. Simulations

8. Differences in Differences

9. Regression Discontinuity & Matching

10. Faculty Presentations & Community Knowledge Sharing Pt. 1

11. Faculty Presentations & Community Knowledge Sharing Pt. II

12. Student Presentations

[hyunjimoon]

angie's course-taking goals:
from above, goal12 can help me complete nail stage, goal34 can help me scale. hope to get some advice on sailing.

To be specific, if I could answer myself what data to collect, who to collaborate with, which journal to publish my first work, after taking your course, it'd be rewarding.

nail stage

• test desirability and technical feasibility of our research question
• iterate between phenomena, theory, measurement to nail our value proposition that satisfies us, customer (paper reader or our future students), investor (advisor)

scale stage

• test operational feasibility and viability of our research before investing significant time and energy
• collaborate with equipment providers (data and tool providers), sampled customers (trusted colleagues and faculty), investor (referee), using and to develop our profession
• segment customer and set beachhead (research outlet's audience)
• build community both from supply and demand side and acculturate our vision, share progress in the language of aligned evaluation metric

sail stage

• processify the above that helps us automate, replicate, platformize, capitalize the knowledge creation itself