Experimental Design Details

Part 1 of the study is designed to observe procrastination in an unobtrusive and naturalistic setting, and to test which practical interventions help reduce it.

We use a math-related real-effort task that is designed to demand a sustained commitment of time and attention, that is not strongly related to individual math ability, and that cannot be solved easily through cheating. The task is based on the so-called matrix task (Mazar, Amir and Ariely 2008 ), in which students are given a sequence of 3x4 matrices filled with 1- or 2-decimal numbers between 0 and 10. The task is then to solve a fixed number of matrices by selecting the two numbers that add up to exactly 10 in each matrix. Each matrix is randomly generated and can be unambiguously evaluated. Whether the numbers have one or two decimals is randomly generated with equal probability. To complete the entire homework task and obtain the monetary incentive, the students must solve 120 matrices in one single sitting, whereby inactivity of up to 15 minutes is permissible, before the end of the deadline. The monetary incentive for the completion of the homework task consists of AU$20 (US$13.20), which is added to an initial balance of AU$5 (US$3.30), in the form of a gift voucher of a store of their choice.

The advantage of this task is that it gives us the flexibility to adjust the difficulty to an appropriate level for senior secondary students. We have trialled the task with a small number of students in a pilot study in the laboratory. Based on the pilot results, we calculate that completion of 120 matrices requires approximately one hour of sustained effort by students in this age range. We also measure the students’ ability at solving the matrix task by incentivising them to solve as many matrices as they can within 5 minutes at our initial school visit. This ensures that the students get to know the matrix task and gives us a rough idea of their relative ability in solving the matrices. Each correctly solved matrix from this practice round adds AU$0.25 (US$0.17) to the student’s gift voucher.

In Part 2 of the study, we build upon Augenblick et al. (2015) to measure impatience and present bias. The time preference elicitation requires each student to make two sets of decisions over the allocation of work effort between two work dates (call them “Week 1” and “Week 2”), with one decision randomly selected to be implemented for real. To identify the discount rate (impatience), the rate at which work on one date may be substituted for work on the other is varied from one decision to the next. To identify present bias, the same choices are elicited at two points in time: once in Week 0 when both work dates are in the future, and again in Week 1 when sooner work must be performed immediately. To facilitate the allocation of work over time, the effort task for this stage of the project must be divisible in nature. Following Augenblick et al.’s procedures, each task consists of transcribing one line of Greek text. On average, this corresponds to one minute of work effort, as trialled in our laboratory pilot with students in the relevant age range. Depending on the selected allocation, students will be required to complete 15 to 60 such tasks in total, distributed over the work dates in Weeks 1 and 2. Successful task completion on both work dates qualifies for a AU$20 (US$13.20) incentive in the form of a (second) gift voucher at a store of the student’s choice.