type
Post
status
Published
date
Nov 6, 2025
slug
teaching
summary
tags
R
SAS
category
Teaching Experience
icon
password
It is my first time being a teaching assistant. It starts in September and ends in December.
I really enjoy the BUMK742-0501: Special Topics in Marketing; Marketing Analytics for Consulting with Professor Wedel in Fall 2024, and I sent an email to him to inquire about this opportunity. If you want to be a teaching assistant too, ask and connect with the professors.
📝 What I mainly do
Responsibilities for BUMK742 Marketing Analytics for Consulting:
1. Grading the memos, entering the grades on Canvas.2. Responding (together with me, mostly via Zoom) to questions that students have about the grades for each memo3. Tallying the grades for each memo, the presence/absence, and the peer evaluations, on Google Drive and Canvas.4. Calculating the final grades.5. Further, there may be specific questions from students that you can help with, often related to SAS.
Responsibilities for BMSO758E Advanced Marketing Analytics (online):
For BMSO758E (online), the tasks are similar:
- Grading the worksheets, entering the grades on Canvas.
- Responding (together with me) to questions that students have about the grades for each memo
- Tallying the grades for each memo, and the peer evaluations. We don't have discussions, there are more items in the gradebook than what I use in the course.
- Calculating the final grades.
- Further, there may be specific questions from students that you can help with, often related to R.
📎 Key Numbers
- 43 students in BUMK742 Marketing Analytics for Consulting
- 38 students in BMSO758E Advanced Marketing Analytics (online)
- An estimated 18-20 hours per week of work time
The course is about using Models
The BUMK742 starts in August, ends in December.
The BMSO758 starts in October, lasts 7 weeks, and ends in December.
project | Topic | Model | software | software (online) |
1 | Retail pricing decision
| Linear, sei-log, log-log regression models | SAS | R & Python
|
2 | Analysing print ad designs using eye-movement data | Generalized linear models: Poisson regression , logic model | SAS | R & Python |
3 | Evaluating the effectiveness of sales promotions based on scanner panel data | Models of incidence, choice and quantity: semilog regression, logic model, multi nominal model | SAS | R & Python |
4 | International market segmentation for global retailers | mixture regression models
multinomial logic models (MNL) and mixture MNL models | GLIMMIX | R & Python |
5 | New product development using choice-based conjoint analysis for coffee makers | Multinomial logit (MNL) models and mixture MNL models | GLIMMIX | R & Python |
Additionally, I had the opportunity to translate R scripts into Python for preparation for the fall class in 2026. One of the main challenges I encountered was implementing multinomial regression, as Python lacks readily available packages equivalent to those in R. Initially, I tried to compile the model with the help of CLAUDE, yet the process was time-consuming and computationally demanding due to the model’s iterative nature. After consulting a PhD student with a strong statistical background, I learned to call R packages efficiently from within a Python environment, which significantly streamlined the workflow.
Overall, the experience was rewarding, deepening my technical skills and problem-solving abilities while reinforcing the importance of collaboration and adaptability in data science projects.
Reflection
What have I learnt from this? Time management ability, interpersonal skills.
What can I do next? Seeking a TA and a TA-sponsored role.
I am still in this position and working on this blog
- Author:Yunzhu HUANG
- URL:/article/teaching
- Copyright:All articles in this blog, except for special statements, adopt BY-NC-SA agreement. Please indicate the source!
Relate Posts
