3  Academic Progress in the Statistical Science PhD Program

PhD journeys are very heterogeneous. Everyone has a different trajectory and there is usually no standard answer for any questions along the line of What do I need to do in order to achieve X. This makes the process of getting a PhD both fun and uncertain. While there is no general formula for most PhD things, some general guidelines and information on timeline can be useful to reduce the uncertainty along the way.

Disclaimer: please note the information presented below is for the Statistical Science PhD program at UCSC. Every institution and program has different expectations and pathways. Sometimes the difference can be significant, especially in the expected timeline.

3.1 First Year Exam and Start of Your Research Projects

PhD students in our department are expected to take core courses in their first year and complete the first year exam (FYE) early June. The FYE include a closed-book exam testing one question from each of the first year courses, and a take-home test that involve a guided data analysis report. It is common for students to focus primarily in preparing the exam during their first year (especially the Spring quarter). Some of past and current trainees in our group and the department join the program with extensive previous coursework in statistics already, and in some cases, they choose to start research projects during the first year. There is no right or wrong ways to approach the first year in the program.

The more important thing for the first year is to mentally transit from the mindset of a student to a graduate researcher. This can be achieved through coursework, reading, talking to peer and faculty members, exploring small research projects etc.

3.2 Advancement to Candidacy

After passing the first year exam, the next major milestone is advancement to candidacy (ATC). At UCSC, the normative time for ATC is before the end of the third year in the program (i.e., for student joining in the fall, the end of summer at year 3). There are some consequences with not meeting the timeline, primarily concerning funding related to tuition. So it should be the goal for all students to meet this timeline. There are some exceptions to the rule, e.g., if you take a leave of absence from the program, the clock can be paused during the time.

The goal of the ATC exam is to demonstrate that you know what you are going to research for your PhD. Usually, this can be shown by a combination of both what you have done and what you propose to do. A typical PhD dissertation usually consist of three projects (i.e., roughly three papers). For ATC with three projects, the usual expectation is

• One project that is completed or close to be completed.
• A detailed proposal of the second project, with strong motivations and background literature review, full derivations and sometimes, initial experiments.
• Clear description of motivation with a less developed proposal for the third project.

As each student’s PhD dissertation is different, the three-project scope does not apply to all cases. Sometimes if the first project does not have enough methodological novelty, e.g., an applied data analysis or software project, more will be expected from the other projects. Sometimes students have four projects where some of the projects are smaller in scope as well.

Instead of being concerned about what the norm is, it is more important to keep your focus on the big picture: the ATC is to demonstrate that you can conduct research at the PhD level and make contribution to the statistics and broader scientific community. This means you can recognize and describe research questions and why they are important, know and have a plan for how to choose the methods to address the question, and know how your work fits into the landscape of the existing literature. These quantities can be demonstrated in many different ways: Sometimes the proposal involve very technical aspect of proposing new statistical method/theory, and sometimes through careful explanation of an important applied problem and what the gaps are.

Logistically, each student is required to write an ATC document describing your work and proposal, and share with the ATC committee members at least 2 to 4 weeks prior to the exam. The document is typically around 30 pages in length (including tables and figures, but excluding references). However, there is no definitive guideline on the document’s length and there are a lot of variability among students. Providing committee members with more time is always preferred. The writing of the ATC document can take up a substantial amount of time, so it is highly recommended to start drafting it early (at least one or two quarters before the exam). For many students, the ATC document is their first major scientific writing assignment, and it is essential to allocate sufficient time for the writing process itself. Useful tips on writing from Dr. Li can be found at this link.

The ATC exam includes a 30-minute presentation from the student, an open Q&A session from the audience, and a closed-door session of further questions from the committee members.

Delivering the ATC presentation is also a good opportunity to practice scientific communication. More recommendations on presentation and answering questions can be found in the conference chapter.

The ATC exam is one of the rare occasions early in your career to have multiple faculty members read carefully what you write and engage deeply with your research and presentation. However much work it seems at the time, the payoff is usually great when looking back after a while.

More information about logistics can be found at this link. More recently, students typically advance to candidacy at the end of the third year (in Spring or Summer). But historically, there have been students in the past who advanced to candidacy in their second year.

3.3 Dissertation defense

After advancement to candidacy, our department typically expects the student to defend their PhD dissertation within the next two years. The official guideline from the Graduate Division at UCSC has a slightly longer timeline of three years after ATC. So students typically graduate from the program in 4 to 6 years, with the majority in recent years taking 5 years. Things are a bit trickier with taking leave of absence after ATC, and you should talk to Dr. Li and the graduate program director about it if there is a need for that.

Defending your PhD dissertation means that you have shown that you are an expert in your research area and is ready to embark upon the next journey in your career. Typically at defense, students will have completed three projects on statistics research, in either theory, method, or application. There is no requirement to have worked on all three aspects, but usually not all projects can be direct application of existing methods to data. As mentioned before, sometimes more or fewer projects is possible depending on the scope of each project.

The PhD dissertation is usually a combination of the papers and preprints you have written, with some reorganization (e.g., removing and grouping repeated introduction/background sections in the papers) and addition to connect the projects under a unifying theme. Past PhD dissertations from our program (and the majority of other universities in US and Canada) can be found on ProQuest. It is a good idea to read some of the dissertations in the past to get a sense of what they look like.

Similar to the ATC exam, the student is required to send the dissertation draft at least 4 weeks before the defense. The dissertation defense includes a 45-min presentation from the student, open Q&A session from the audience, and a closed door session of further questions from the committee members. Usually the students will get feedback and request for edits on their dissertation after the defense.

For students graduating around the end of any quarter, it is important to check the deadlines for the dissertation to be submitted in order for the degree to be awarded in time. For international students who are applying for OPT after graduation, it is also important to check with the international student office on when to file for what document. It is important to know that some of the immigration forms can be filed prior to dissertation submission, if that is the only requirement left for the degree. More information about logistics can be found at this link.

3.4 Publication

Typically PhD students are expected to have at least three first author papers when they graduate. The papers do not need to be published, as the cycle for publication is long in statistics, but they should be in a form that can be shared (e.g., ready to be posted as a preprint on arXiv).

What does first authorship mean in our Lab? Being the first author typically means that you are: (1) an intellectual lead on the project, (2) making major contributions conducting the scientific work (e.g., developing methodology, analyzing data), and (3) a main writer on the manuscript, usually crafting the first draft. For some projects, a first authorship role may also involve project management with collaborators and partners. All of these activities are done under the mentorship of Dr. Li as the PI with substantial feedback in an apprenticeship model.