How does the Delay Discounting Test work?
Each question offers some money today or $100 later, and you pick the one you would really rather have. The money is imaginary and there are no right answers. There are 30 choices: 6 at each of five delays, 1 week, 1 month, 6 months, 1 year and 5 years, asked in that order.
At each delay the offer starts at $50 today. Take it, and the next offer is lower; wait, and it is higher, by $25, $12, $6, $3, $2 and $1, one step after each choice. Six choices bring the offer to within a dollar of the amount where today and later feel the same to you: your indifference point. The two options swap sides at random, so tapping one side out of habit doesn't pass for a preference.
How is the Delay Discounting Test scored?
Your result is your indifference point at one year: what $100 in a year is worth to you today. Beside it are three summaries of all five points. The discount rate k comes from fitting James Mazur's hyperbola (1987), value = amount ÷ (1 + k × days), to your points: the larger k, the faster waiting eats value. One over k is the wait at which a reward is worth half, shown as “worth half after”.
The area under the curve (Joel Myerson, Leonard Green and Missaka Warusawitharana, 2001) is the area under your five points with money and delay both scaled from 0 to 1: 1 means waiting cost you nothing, near 0 means you wanted almost everything now. It needs no model, which is why it is often reported alongside k. How the hyperbola plays out for a few values of k:
| Discount rate k | $100 in 1 week | $100 in 1 month | $100 in 1 year | $100 in 5 years | Worth half after |
|---|---|---|---|---|---|
| 0.001 | $99 | $97 | $73 | $35 | 2.7 years |
| 0.01 | $93 | $77 | $22 | $5 | 3 months |
| 0.1 | $59 | $25 | $3 | $1 | 10 days |
What is delay discounting?
It is the way a reward loses value the longer you have to wait for it. The loss is steep for the first delays and gentler after, which is the hyperbola's shape: a week can matter more than the difference between four and five years. The questions here follow the adjusting-amount procedure of Wenjie Du, Leonard Green and Joel Myerson (2002), who used it to compare how people from different countries discount.
Answers can also be inconsistent: a longer wait worth more than a shorter one, or no change at all. Matthew Johnson and Warren Bickel (2008) proposed two checks for that: no point more than 20% of the amount above the one before, and the last point at least 10% below the first. If yours fail one, the result says so and the curve is a rough guide.
Is a low discount rate in the Delay Discounting Test better?
Not in itself. Taking money now can be the sensible choice: when you need it, when you doubt the later payment will come, or when prices are rising. The test measures how you trade time for money in these questions, not whether that trade is right for your life, and imaginary money may not feel like your own.
What can affect your Delay Discounting result?
Your situation as much as your temperament: how much $100 means to you, whether you trust a promise to pay later, your mood today. The order is fixed, from the shortest wait to the longest, so each answer is made with the last in mind. To see how well you know what you don't know, try the Overconfidence Test; to see how random your own choices are, the How Random Are You? test.