← Back to Mentoring

Mentee Compact

These are my ideal expectations for a mentoring relationship. They’re a starting point, not a rigid contract — the specifics are worked out jointly with each mentee at the start of a project.

Goals

The long-term goal of the working relationship should be clear and well defined from the start. The skills developed in research can support a number of outcomes, including (but not limited to) preparation for graduate research, preparation for a career in industry, or development of problem-solving skills to support academic success more broadly.

Throughout the project, mentor and mentee meet to establish and periodically review their shared goals, expectations, and milestones. Initial and ongoing project scope is defined jointly, to find topics that best match mutual interests and skills, with the big-picture goals revisited regularly over the course of the mentorship.

Mentor and mentee should also identify at least one additional mentor for support — such as a faculty career mentor, graduate student, postdoc, or mentor from a national society or affinity group — who can help in areas where the primary mentor isn’t able to.

Core Mentee Responsibilities

The mentee reports on their progress before each weekly meeting, following a shared template. Because the research is data-intensive, progress is often easiest to show through a new plot delivered each week — though the plot need not be the mentee’s own creation (for example, when the week was spent on a literature review, an informative plot from an existing paper works just as well). Reading and writing activities happen regularly and are documented in the same report.

The mentee maintains documentation of their research and an archive of research products somewhere mutually accessible (GitHub and a shared document, for example), and makes themselves available for meetings and research group events to the greatest extent possible.

Authorship and Publication Guidelines

The mentee participates in the publication and dissemination of research and in applications for funding to support it. “Publication” here is broad — it includes academic papers, but also public code, data releases, posted tutorials, internal training materials, and any other way of sharing work with the world. All student output is valuable and must be properly attributed to the mentee in all future use.

A paper is a common goal, but not always the outcome of a single semester or summer research experience — it depends on research progress, project scope, the mentee’s career goals, and circumstances that come up along the way. When an academic paper does result, authorship is defined as follows:

• If the mentee is lead on the project at the time of submission, they are guaranteed first authorship.
• If the mentee contributed the majority (or a significant share) of the analysis, regardless of their current participation in the project, they are guaranteed first or co-first authorship with the current project lead.
• If the mentee is no longer working with the group or on the project but contributed in some way, they are offered authorship, though not guaranteed first-author rights.

All other publications are credited similarly, based on who took the lead or made the most significant contribution.

Communication and Working Hours

For the mentor:

• Provides a set number of hours of face-to-face time per week for progress updates and planning, scheduled around both parties’ availability.
• Keeps individual meetings to roughly an agreed length, with extensions or reschedules handled case by case; if the mentor is late without notice, that time is made up.
• Is clear on both short- and long-term project plans, with achievable benchmarks and a plan for week-to-week progress, and provides direct help or resources if roadblocks come up between meetings.
• Is reachable by email or Slack, responds to the mentee within 48 hours (and ideally much sooner), and gives advance notice of any travel that may affect availability.
• Treats the mentee as a respected colleague — without discrimination or harassment on the basis of race, ethnicity, religion, country of origin, gender, gender expression, sexual orientation, or disability — and makes every effort to accommodate the mentee’s identity and needs.
• Provides the tools, professional development, and scientific knowledge needed to succeed, and publicizes and promotes the mentee’s work whenever appropriate.

For the mentee:

• Communicates problems as they arise, and lets the mentor know as soon as possible about personal conflicts that will affect their schedule.
• Reaches out to the mentor or a peer mentor when stuck, and speaks up if unhappy with the project’s progress or direction so the work can be rethought together.
• Lets the mentor know as soon as possible about lateness or absence so meetings can be rescheduled.
• Works roughly an agreed number of hours per week (more or less depending on schedule and obligations), distributed in a way discussed with the mentor in advance of any significant changes — aiming for meaningful progress without unnecessary pressure.
• Can do this work anywhere with an internet connection and a laptop — but is encouraged to block out dedicated time rather than work at all hours. Meetings, coding practice, literature review, group meetings, and seminars all count toward the agreed weekly hours, which are logged and shared with the mentor for reference.
• Meets with the mentor to find solutions if they are consistently spending significantly more than the agreed hours — especially if coursework is being affected.

Guidelines for Use of AI

In general, generative AI should not be used for research activities — writing, coding, or reading papers — since it can impact the mentee’s ability to develop strong research skills, and often produces misinterpretation, misattribution, or misleading results when applied uncritically. Using generative AI on someone else’s code or writing without their knowledge is also unethical, since it shares their intellectual property without consent.

Any use of generative AI must be transparent and agreed upon by the mentee, the mentor, and all other stakeholders on the project. This is addressed case by case, but should always follow current best-use practices, such as those outlined in arXiv:2510.22254.