Design scenarios — defence cards
Eight settings. Cut along the dashed lines and deal one per team. Cards 1–4 suit meeting 7 (experiments, instruments, discontinuities); cards 5–8 suit meeting 8 (difference-in-differences, fixed effects, event studies). Page 2 is the defence sheet.
The savings nudge
A bank wants to know whether its monthly financial-literacy email makes customers save more. It has 400,000 customers on the list, it controls who receives what, and it can wait six months.
Defend: what design, and what does it cost them to run it properly?
The scholarship cutoff
A university awards a scholarship to every applicant scoring above 1200 on an entrance test, and to nobody below. Administrators want to know whether the money raises graduation rates.
Defend: what design — and who exactly does your estimate apply to?
The water filters
An NGO distributed water filters to the villages that asked for them. Diarrhoea rates fell in those villages and not in others. The donor wants to know whether the filters worked.
Defend: a design — or defend the claim that this cannot honestly be answered.
The local label
A food company is deciding whether to put "locally sourced" on a package. It wants to know what that claim is worth in cents per unit before it commits to the supply chain.
Defend: what design, and what is the gap between what it measures and what they want?
The self-checkout rollout
A grocery chain installed self-checkout in 340 of its 900 stores over eighteen months, region by region, in an order set by an old refurbishment schedule. Did basket size change?
Defend: what design — and what does "region by region" buy you?
The border wage
One state raised its minimum wage on 1 January; the state across the river did not. There are fast-food restaurants within two miles of the border on both sides.
Defend: what design, and what has to be true about the two sides?
The light rail
A city opened a light-rail line four years ago. House prices within half a mile of the new stations are up 22% since. The transit authority would like to claim credit.
Defend: what design — and why is the naive comparison almost certainly wrong?
The seller fee
A platform raised its commission for seller accounts created on or after 1 March, leaving older accounts on the old rate. It wants to know whether the higher fee drove sellers away.
Defend: what design, and what would make the two groups not comparable?
Defence sheet — team ______ · card ____
How this runs
Every team works its card at the same time. Then one team is drawn at random to defend at the front while everyone else attacks — and we draw again, and again. You will not know whether you are up, so be ready. Points for an attack that lands, and you have to name what is wrong, not just dislike it.
The graph — draw it
Our design, named
The assumption carrying it — the one we would have to defend
What data this requires that we have not been given
The attack we expect, and our answer to it
From the floor: the attack that actually landed