Purpose
Judgment, not trivia
The cases surface the habits senior analysts need: metric humility, causal caution, operational risk, fairness burden, and uncertainty communication.
Data Science Judgment Lab
This is a 25-case room for practicing the hard part of data science: deciding what messy evidence can support, what it cannot support, and how confident you should be when pressure is in the room.
Purpose
The cases surface the habits senior analysts need: metric humility, causal caution, operational risk, fairness burden, and uncertainty communication.
Method
You work the evidence before the explanation appears. The struggle is the point: tempting wrong stories become visible enough to learn from.
Record
Completion creates a local judgment record after reviewed cases. It is a learning artifact, not a proctored credential.
Case library
A launch-week chart jumps, a deck is due, and several teams have reasons to claim the movement.
A checkout readout lands just before planning closes, and different artifacts point toward different launch stories.
A polished retention model arrives with a renewal deadline, a crowded outreach queue, and a promise that the save team can act sooner.
A city inspection team has a new routing screen, a long backlog, and one week to decide how much authority the score should have.
A district impact brief is headed to a funding vote after students who used a tutoring platform show stronger spring gains.
A city housing office must set winter overflow capacity from a forecast that fits ordinary nights better than pressure weeks.
A state benefits agency wants to use a verification score to cut backlog, but the burden may land unevenly on applicants with messier administrative records.
A benefits agency chatbot demos smoothly before a filing surge, and the launch packet must decide what evidence is enough for live claimant use.
The same benefits agency must choose a payment-hold threshold that catches fraud without turning suspicion into broad payment delay.
A modernization dashboard gets a cleaner headline metric, and leadership wants to use it as proof that service is improving.
A customer survey produces a clean majority for a support redesign, and tomorrow's slide asks how far that voice can carry.
A familiar hospital operations field powers a clean improvement story while source systems leave conflicting traces.
A clinical risk report looks steady after an analyst trims the file, and the committee wants to shift attention toward treatment timing.
A de-identified public health export clears a checklist, but linkage, consent scope, and lifecycle controls make the release less simple.
A board packet turns an early operational shift into a dramatic story, and the chart frame is doing more work than it first appears.
A regional media test arrives just before a national buying window, with a confident lift estimate and a deadline to scale.
A workforce pilot briefing shows a promising post-launch gap, and the budget office wants a statewide recommendation.
A housing navigator pilot produces a sharp estimate near an eligibility line, and the agency wants to carry it into a budget request.
A product experiment gets a fast no-go recommendation, but the exposure record and interval width leave more than one interpretation alive.
A subscription checkout test lifts paid starts, but refunds, retention, and support burden make the growth claim less settled.
A hospital readmission score looks unusually strong in validation, and the launch team wants it in the discharge workflow next month.
A trust-and-safety model clears a vendor benchmark before peak season, and leadership wants to turn high scores into automatic action.
An ETA model alarm appears while headline service levels still look acceptable, and no team is eager to slow the workflow.
A holiday replenishment model reports strong backtest accuracy, and operations wants to let it override store planners next week.
An enterprise assistant performs well on clean sales workflows, and revenue operations wants tool-enabled expansion before renewal season.