Sprint Review โ Looking Back at TR2
Before planning a new sprint, Scrum reviews the last one. This isn't a formality โ it's why several things about this trimester look different from TR2, and it's worth understanding what changed and why before you start.
What Worked โ Keep Doing This
Teaching approach
Multiple independent students cited the Socratic "why" method and story-driven teaching as what most changed their confidence over the term โ several went from not speaking up on day one to presenting without notes by the end. Corroborated separately by several students on presentation confidence specifically.
Real-world project framing
A stock-market project felt more meaningful than a made-up example to more than one student โ the applied, practical structure was repeatedly preferred over theory-only assignments.
Assessment feedback quality
Detailed, actionable feedback on assessments and presentations was called out as a standout strength, independently, more than once.
What's Changing This Trimester
The weekend crunch
The most substantive critique: with data collection realistically starting Friday and wrapping by Sunday, teams were forced into a two-day rush instead of the daily iteration Scrum is meant to teach โ some went through the motions of standups without practicing the real rhythm.
Figma wasn't built for this
Three independent, distinct critiques: no notifications or task management (became a static whiteboard, not a working board), access/onboarding friction disrupting early momentum, and required updates feeling like disconnected admin overhead rather than something supporting real progress.
Assessment integrity โ commit history can be gamed
Repos with no protection allow rewritten or backdated commit history, which undermines any grading that treats commit timestamps as evidence of sustained work.
Team formation and accountability
Random assignment carried real risk โ some teams had members who never participated from the start, putting sustained pressure on the rest. A separate but related theme: individual grades leaning too heavily on group outcome meant low-contributing members received similar outcomes to those doing most of the work.
AI fluency gaps
Not every student was equally skilled at using AI for programming problem-solving โ observed failure modes included sending screenshots instead of typed problems, letting conversations lose context, and not giving the AI real task details.