This page contains all of the information you need to know as a domain mentor. Current mentors and potential future mentors alike will find this useful (students probably won’t!)
- Action Items
- Program Overview
- Quarter 1 Preparation
- Quarter 1 Logistics
There are several action items mentioned throughout this page; they are summarized here.
- Check the enrollment page to see when and where your section is being held. If you’re teaching in-person, the location of the new HDSI building can be found here.
- Use this sheet to find your students’ names and emails. (Let Suraj know if you don’t have access to the sheet.)
- Send an email introducing yourself to your students, asking them to complete any pertinent tasks, and reminding them that the first meeting is the week of Monday, October 2nd.
- If you’re teaching a remote section, send your Zoom link to your students by this weekend.
- Choose when to hold an hour of weekly office hours (outside of your discussion time) and share it with your students. (Note that we have not booked physical spaces for office hours – if you’d like additional physical rooms booked, let Suraj know.)
- (Optional) Create a website or Drive folder for your section and share it with your students. If you’d like your website to be linked on the enrollment page, let Suraj know.
- (Optional) Create a Slack/Discord/Ed group for your section and share it with your students.
Remember that sections begin the week of Monday, October 2nd (Week 1). Friday sections will have to reschedule in Week 6 due to Veterans Day; plans for the week of Thanksgiving will be announced later.
The capstone program is a two-quarter sequence (Fall 2023 and Winter 2024) in which you, a domain expert, act as a mentor for student projects. You will host a single domain of inquiry, in which 4-10 students are enrolled, and within your domain, students will form project teams of 2-4. This means that ultimately your domain will consist of 1-5 teams, each of which are working on different projects that relate to your domain. We’ve structured the capstone program in this way to ensure that:
- there are adequate guard-rails for students new to independent work,
- students have material to follow and learn from when developing their own work, and
- the scope of projects remain manageable for the time frame and appropriate for the level of the students.
Students were shown the domain topics and descriptions that you provided here in advance of the enrollment period and have already enrolled at this point. They are free to change the domain they are enrolled in during Week 1, but space is limited.
To help you calibrate your expectations, you should look at last year’s projects if you haven’t already.
As mentioned above, the capstone program spans two quarters. In both quarters, there are two components to the capstone: domain-specific instruction (provided by you) and methodology-specific instruction (provided by Suraj and the DSC 180AB TAs).
|1 (Fall 2023)||Getting students up to speed in the domain by working on a guided project||Software development, reproducibility, environment independence|
|2 (Winter 2024)||Working on an original project||Project management, effective communication skills|
- Quarter 1 (Fall 2023) is designed to provide students an opportunity to acquaint themselves with a domain by working through a guided project that you prescribe. This constitutes their Quarter 1 Project. Often times, this involves replicating an existing paper, but you’re free to guide your students however you see fit. During Quarter 1, students also form project groups and write/present proposals for their final project, known as the Quarter 2 Project.
- Quarter 2 (Winter 2024) is when students execute their Quarter 2 Project.
Throughout both quarters, you will hold an hour per week of discussion section. Students are required to attend these synchronous section meetings.
- In Quarter 1, this hour should be spent answering questions students have on the readings and tasks you provide them with.
- In Quarter 2, this hour should be spent checking in with students on their projects.
You will also hold one hour per week of office hours throughout both quarters. More details and logistics on all of the above components are provided below.
The methodology component of the program is designed to equip students with the skills they’ll need to execute their projects. This includes accessing servers, executing long-running jobs, and project management. You can see the full schedule on the homepage of this website.
Aside from on the first Monday of the quarter, when Suraj will be holding a live introductory lecture, methodology content will be delivered asynchronously in the form of readings. Many of the methodology lectures will have accompanying methodology assignments (e.g. accessing DSMLP, creating a Docker image, writing a document in LaTeX).
The methodology team (instructor + TAs) will hold several weekly office hours. These are designed for students to get help with methodology lectures and assignments, and to ask infrastructure-related questions about their projects.
As mentioned above, your primary goal in Quarter 1 should be to get your students familiar with the types of problems you’d like them to be able to solve on their own. A common way this is done is to choose a result or paper in your field around which to structure students’ introduction to the domain. While not strictly necessary, having known results to replicate helps students know they are on the right track. (After they start working on their Quarter 1 Project, they’ll submit proposals for their more independent Quarter 2 Project, which will be within the realm of your domain.)
To structure student learning, you should prepare a weekly schedule (see examples below). Each week should consist of three parts:
- A list of 1-3 topics they are working on for the week.
- Readings and/or tasks for the week. This is typically “read section XX from the paper and replicate their analysis on your data”.
- A list of 1-3 participation questions they should answer before discussion. These may be conceptual or computational, depending on what is appropriate for the tasks they were given. These may be formulated the week before, given your current understanding of student progress. More details are in the Participation Questions section below.
The outline below follows a reasonable cadence for Quarter 1:
|Topic||Number of Weeks|
|Intro + Data||2-3|
|Elevator Pitch for Quarter 2 Project||1|
We will not “enforce” the above outline; feel free to move at a pace that is appropriate for your domain.
You should plan to be flexible; for instance, it’s quite possible that not all of your students completed the summer tasks you outlined for them here, so you may need to spend more time on fundamentals than you initially plan. This is fine.
One way to share your weekly schedule, readings, and tasks with your students is through a website. You can see example websites from domains in previous years below. If you’ve never run a domain before, it’s highly recommended to look at these websites to understand what a typical schedule looks like.
- Aaron Fraenkel: Understanding Code Through Graphs (Malware Detection)
- Jingbo Shang: Text Mining and NLP
- Tiffany Amariuta: Genetic risk prediction of gene expression, complex traits, and polygenic disease
- Justin Eldridge: The Spread of Misinformation Online
- Alex Cloninger and Deloitte: Activity Based Travel Models and Feature Selection
You can see more examples in Column I of this sheet from 2022-23.
The easiest way to create a website of your own is to clone one of the websites above.
- Many follow Aaron Fraenkel’s site; you can fork it from here. Make sure to enable GitHub Pages.
- Kyle Shannon’s site this year is also excellent; you can fork it from here.
Note that it’s not required to create a website; you may choose to share information with students in other ways, like a Google Drive folder. If you create a website and want to link it at dsc-capstone.org/enrollment for posterity or visibility, let Suraj know.
Here, you’ll find details on how each relevant component of the capstone works. Each week, you will receive an email from Suraj with a brief summary of what students should be working on that week and with any relevant updates (deadlines, changes to the schedule, new action items, etc).
Weekly discussions are not lectures, but rather a time that students can ask clarifying questions about their weekly reading/tasks. You may begin by asking students how they approached the tasks and readings in the weekly schedule. (Another goal of Quarter 1 is to have students learn best practices for communication in your domain, so it’s worthwhile to discuss how the readings were organized, too.)
One piece of advice to encourage participation in discussion is to let students know that orienting oneself in a new area of research is difficult and that it’s totally normal to get lost and misunderstand the material. Students may be quiet because they’re not confident in the work they’ve attempted. Try to reinforce to students that the purpose of discussion is to get feedback and guidance from domain mentors. The weekly schedule tries to facilitate that.
Remember, discussions begin the week of Monday, October 2nd. Check dsc-capstone.org/enrollment to see where and when your section is. If you’re running your sections through Zoom, send your students your Zoom link via email. (Find your students here).
In addition to the readings and tasks students are asked to complete each week, students must submit answers to participation questions 24 hours in advance of discussion, starting in Week 2. The questions are due well before discussion in order to allow you to use students’ answers to structure your section, if you wish. We’ve created a “default” set of participation questions, linked here, that should apply to all domains. However, you’re free to devise more specific questions for your domain. You’re encouraged to look at past domain websites for examples of such questions.
You’re encouraged to set up some type of communication channel for the students in your domain, so that they can more easily discuss the readings and tasks with one another under your guidance. Such forums are also useful for fostering a sense of community amongst students in your section.
The main purpose of office hours are to provide students with help and guidance on their current work and to provide a meeting time for feedback on submitted work.
You should require that students meet with you in office hours every ~3 weeks in Quarter 1, both to gauge their progress and to form stronger interpersonal connections (one of the implicit goals of the capstone is to improve students’ communication and networking skills).
Plan to be available for at least one hour per week outside of your regularly-scheduled discussion section. This can be a synchronous hour, either in-person* or on Zoom, or can be an hour where you promise students you’ll be immediately available via email/Slack/Zoom to answer questions. If students are unable to meet with you during your regularly scheduled office hours, you may need to meet with them by appointment.
*Note that we have not booked physical spaces for office hours – if you’d like additional physical rooms booked, let Suraj know.
All assignments in the course are generic, designed to work for most domains out-of-the-box. If any of the assignments (other than the methodology assignments) don’t fit with your domain, you can modify it to work for you; let staff and students know of these criteria.
|Quarter 1 Project||Students work on a guided project (e.g. replication) in your domain||Individual or groups (up to domain mentors)||Week 5 (checkpoint), Week 10 (final submission)||Report is graded by domain mentors on Gradescope; code feedback given by methodology staff|
|Quarter 2 Project Proposal||Proposal for final project||Groups||Week 10||Domain mentor|
|Participation Questions (default questions)||Weekly questions to keep students engaged with the domain material||Individual||Weekly, 24 hours before discussion||Methodology staff (completion only); domain mentors should read to inform how to run discussion|
|Methodology Assignments||Assignments that develop students’ software development and project management skills.||Individual||Bi-weekly||Methodology staff|
- Students’ Quarter 2 (i.e. final) Projects must be completed in groups of 2-4. Individual projects will be allowed only by exception; to work alone, a student must meet with both their mentor and Suraj to justify why they should be permitted to work outside of a group.
- How the Quarter 2 groups are formed is up to you; you can divide your students into groups yourself or have them self-organize. Quarter 2 Project groups will need to be finalized around Week 8 of Quarter 1.
- With that said, students’ Quarter 1 Projects can either be done individually or in groups; this is up to you. Possible strategies:
- Have all students work on their Quarter 1 Projects individually. (If you do this, you should stress that each one of their submissions should look different.) When it is time to form Quarter 2 Project groups, assign students to groups.
- Allow students to organize into groups at the start of Quarter 1 and, assuming the pairings are amicable, have them remain in the same groups in Quarter 2.
- Randomly organize students into groups at the start of Quarter 1. When it is time to form Quarter 2 Project groups, allow students to self-organize into groups. (The idea here is that they know how they work with their Quarter 1 group and can either choose to stick with them or experiment with another group.)
- Ideally, students formulate their own project closely related to their Quarter 1 Project. Formulating questions is a valuable skill, and staying close to the Quarter 1 Project will allow students to reuse their code.
- However, you may approve any project you are comfortable advising as a domain expert. Note that other students in the domain will be required to follow along and peer review their classmates, so a project too far afield may hinder those interactions.
- Fixing the project ahead of time (even, say, at the start of Quarter 1) and requiring students to work on a specific coordinated task in the area is fine. In fact, typically students want such guidance. You should make these expectations clear at the beginning of the quarter.
It is important to provide students with regular feedback on their work to ensure their progress towards a successful Quarter 2 Project. We’ve designed the capstone program to give students frequent feedback with minimal overhead for the mentor and minimal stress for students.
In order to ensure consistent grading across such a diverse array of domains, we will utilize a coarse grading scheme with a clear rubric. This scheme will reflect broad checkpoints that students meet, and should help maintain focus on large, impactful things that students can improve on while reducing grading disagreements.
The grading scheme we will use for individual assignments follows an A/B/C/F scale (without plus/minus), developed by Shannon Ellis:
|A (4.0)||Accomplishes the task accurately, completely, and clearly. Code is clear, effective, and efficient. Written component is concise, at the appropriate level, and correct. Oral component (when applicable) is effective and within the time limit.|
|B (3.0)||Accomplishes the task well, but lacks some completeness or clarity. Code runs but lacks some aspect of clarity, effectiveness, and or efficiency. Written component is logical and generally correct, but lacks either clarity or accuracy. Oral component (when applicable) is moderately effective and/or slightly outside the time window.|
|C (2.0)||The task is somewhat accomplished, but lacks significantly when it comes to completeness and clarity. Code present but does not accomplish the task up to the standards of a data science graduating senior. Written component lacks substantial clarity/correctness. Oral component (when applicable) significantly lacks effectiveness/clarity.|
|F (0.0)||The task largely remains unaccomplished. Code lacks completeness, structure, and is unclear. Written component lacks required information to understand the work done. Oral component (when applicable) is nonsensical/unclear.|
Hopefully, it is “obvious” what grade an assignment should receive. With that said, the focus is not the grade, but rather actionable feedback. Whenever you are responsible for grading something, you will need to enter your letter grades on Gradescope, but you can provide students with feedback either in writing or in office hours.
If any changes are necessary to the above rubric for a particular assignment, we will communicate that to you before it is time to grade.
Final grades will be computed using the grade-points above, using the proportions given in the course components table. Letter grades will be assigned using the standard university cutoffs. At the end of Fall 2023 students will receive a grade in DSC 180A, and at the end of Winter 2024 they will receive a grade in DSC 180B; each course is worth 4 units.
There are a few platforms to be aware of.
- Gradescope: This is where students will submit all of their work and where you will do any necessary grading. We are not using Canvas at all this quarter. You should have already been added to the DSC 180A Gradescope; let Suraj know if not. Any time you need to grade something, you’ll be sent a direct link to the assignment you need to grade, with instructions on how to access the work of just your students.
- Ed: This is the message board that methodology staff will use to communicate with students. You have been added as well so that you’re aware of the announcements that students receive each week. You don’t need to post anything there.
- Slack: There is a Slack channel in which mentors can ask questions about how to run their domains; ask Suraj if you’d like to be added.
As far as compute goes – alongside personal laptops, students will use the campus ITS DSMLP Cluster. This is a Kubernetes cluster where students can specify their compute resources and a data science environment in which to work. Later in the first quarter, students will create and use a Dockerfile for deploying their data science environment on the DSMLP cluster. Unless your project has very specialized computing needs, you likely don’t need to provide your own computing resources for your students.