I'll Finish it This Week and Other Lies - Why PhD students lack productivity in time management

Weekly goal setting in academia often fails in predictable ways: researchers systematically underestimate how long tasks will take, complete only about half of their planned weekly assignments, and then carry unfinished work forward—creating a cycle of frustration rather than progress. A short study led by Kaylee Brower, an astrophysics PhD student at MIT, tracked a weekly goal group over nine months and found that expected finish times rarely matched actual completion times, especially for tasks that are hard to scope in advance.

Across 559 tracked tasks, 17 were still incomplete by the end of the nine-month period. On average, participants completed 53% of the weekly tasks they assigned themselves. The bigger mismatch was timing: tasks took 1.7 times longer than anticipated. When tasks were grouped by type—coding, writing, reading, administrative work, talk preparation, and problem sets—coding and writing consistently ran longest. Reading, email, and administrative items were comparatively easier to estimate because their inputs are more bounded (for example, a known-length paper or routine messages). Writing and coding, by contrast, can balloon when unknown problems surface: a “simple” coding task can drag for weeks, while something expected to be lengthy can finish quickly.

The study also tested whether time-estimation improves with career stage. The cohort included undergraduates, graduate students, and postdocs, but the sample size was too small to draw strong differences. Still, a pattern emerged: undergraduates finished tasks more on time than higher-level researchers, likely because undergraduate work is often tied to externally imposed deadlines and clearer time budgets. Postdocs were slightly better than graduate students, suggesting some improvement with experience, though not in a clean, linear way. Participants’ own estimation skills showed a similar non-monotonic trend: accuracy improved somewhat through the process, then dipped later.

Task categories shifted with seniority. Homework-like problem set work decreased as participants moved up the academic ladder, while writing and administrative tasks increased; coding stayed roughly constant. Another practical finding: tasks with deadlines were more likely to be completed on time, reinforcing the idea that time-based goals can outperform open-ended “do these tasks this week” lists.

The takeaway for PhD productivity is less about working harder and more about changing how work is planned. The study’s results align with a common failure mode—constantly pushing unfinished items into the next week—turning planning into a recurring record of missed targets. The creator of the productivity templates and the upcoming part two plan frames the next step as moving away from weekly task lists toward a system designed to handle uncertainty more realistically, drawing on David Allen’s Getting Things Done. In parallel, Notion templates are being developed to support time-blocked scheduling, project-based views, and routine checklists, with a “new week” button to reset planning without carrying forward the emotional weight of perpetual backlog.