Risk register

What could move a deadline, how we would know, and what we do about it

Ten risks, scored likelihood times impact on a three-point scale. Each one names the trigger that tells us it is happening and the response that is already in the activity list. The register lives in data/risks.csv; close a risk by setting its status, and add one when a sync surfaces a new way to lose a deadline.

The matrix

Low impact

Medium impact

High impact

High likelihood

2R02, R07

1R01

Medium likelihood

1R10

3R05, R09, R11

4R03, R04, R06, R08

Low likelihood

The register

ID	Phase	Risk	Likelihood	Impact	Score	Trigger	Response	Owner	Activities	Status
R01	Proposal	Fifth member is not on the official repo and the course guideline says teams of three to four; a five-name proposal submitted from a repo he cannot access raises questions	High	High	9	Still missing from the official repo on Sep 18	Ask the professor at the workshop meeting to add him and to confirm the team size	Team	S05	Open
R02	Proposal	Draft carries AI paste artifacts and no references; the professor spots them at once and the Organization and Writing points go	High	Medium	6	Artifact string or chatgpt link still in the .qmd at render	Blocking-fix list in the Sep 16 review; a second reader checks the rendered HTML before the push	Team	P06, P07	Open
R03	Proposal	The professor challenges the topic on the emerging and pre-commercial criteria (smart locks and thermostats are common in new Class A buildings) and asks the team to narrow scope	Medium	High	6	Pushback at the workshop meeting	Lead with the design-decision argument: which features to bundle at what premium for mid-market and retrofit stock; no public tradeoff-based WTP exists; be ready to drop lighting or EV charging	Team	P05, P06	Open
R04	Survey plan	Price attribute mis-specified: a zero-dollar premium makes a free smart package a valid profile and there is no base-rent anchor, so the WTP estimates are biased	Medium	High	6	Design file generated with the draft levels	Fix in SP01: reserve zero for the standard-apartment reference or shift the range; show a base rent in every choice task	Team	SP01, SP03	Open
R06	Pilot survey	Survey hosting breaks: Supabase not connected, shinyapps deploy fails, or a key leaks into the repo	Medium	High	6	Responses not stored in a test run; a key in a diff	Assign the account owner by Oct 2 (S07); keys in .Renviron only; end-to-end test on Oct 12 at the latest	Account owner	S07, PS04, PS05	Open
R07	Final survey	Weeks 9 to 13 stack HW8 to HW12, three quizzes and three deliverables; the pilot-to-final handoff has one day of float	High	Medium	6	Any pilot-analysis activity slips past its planned finish	Start cleaning on partial pilot data (PA02); triage feedback in parallel (PA05); owners agree a weekly time budget before Oct 21	Team	PA02, PA05, FS01	Open
R08	Final analysis	Prolific fielding starts late or runs long over Thanksgiving, leaving days rather than weeks between the final data and the Dec 6 report, with the Dec 2 exam in between	Medium	High	6	Launch not confirmed by Nov 11; fewer than the target responses by Nov 25	Confirm the launch date at the Nov 11 class; scaffold every script on the pilot data (FA01) so the final run is a re-run	Team	FS04, FA01, FA02	Open
R05	Pilot analysis	Six attributes at three to four levels each overloads respondents and the pilot standard errors are too wide to size the final sample	Medium	Medium	4	Pilot coefficient standard errors dwarf the estimates	Pilot with the full set; drop or merge the weakest attribute (lighting is the likely candidate) in FS01	Team	PA03, PA04, FS01	Open
R09	Ongoing	Uneven contribution; the professor grades non-contributors differently and the Attribution section has to be honest	Medium	Medium	4	CONTRIBUTIONS.md shows one or two names for a deliverable	Owner and backup per deliverable (S06); contributions logged weekly (O03); raise it at the sync, not at submission	Team	S06, O03	Open
R11	Survey plan	The survey plan repeats the proposal's GWU-network distribution while the course expects Prolific with an eligibility pre-screen; the final survey then lacks a fielding plan the professor can run	Medium	Medium	4	SP02 drafted without naming Prolific and the exact eligibility text	Decision D07: state Prolific as the strategy for the final survey, classmates for the pilot, networks as a supplement; write the eligibility question text in the plan	Team	SP02, FS03	Open
R10	Ongoing	Decisions made on WhatsApp never reach the repo and the team re-argues them	Medium	Low	2	A decision cannot be found in decisions/DECISIONS.md	Same-day copy rule (O01); the tracker maintainer checks the log at each sync	Harshit	O01	Open