aikyam school

Payoff-Maximizing Task Selection

Individuals choosing between easy low-reward options and challenging high-reward options frequently make sub-optimal decisions due to poorly calibrated beliefs regarding their own success probabilities and effort efficiency.

Picture this

Imagine standing in front of two arcade games: one pays out 1 ticket every single time, while the other pays out 4 tickets but requires timing a button press. If players underestimate their own timing ability, they waste hours on the 1-ticket machine even though practicing for 5 minutes on the 4-ticket machine yields far more total tickets.

What the evidence says

Treated students were 14.1 percentage points more likely in Sample B during Visit 1 (p < 0.01) and 9.1 to 9.6 percentage points more likely across both samples during Visit 2 (p < 0.05) to select the task difficulty that maximized expected material payoffs.

Who
N = 2,894 fourth-grade students across two field cohorts in Istanbul, Turkey (Sample A: N = 1,704 Visit 1, N = 1,578 Visit 2; Sample B: N = 1,190 Visit 1, N = 1,129 Visit 2).
How
Empirical distribution logit modeling predicting individual success probabilities based on baseline traits, comparing actual choice against expected payoff-maximizing choice across two field visits.

What to do

Calculate expected success probabilities based on historical baseline metrics before task selection to ensure participants choose difficulty levels that maximize total expected yield.

From the source

"In Sample B, students are more likely to choose the payoff-maximizing task in both visits. In particular, treated students are 14 percentage points more likely to make the payoff-maximizing choice in visit 1, and 10 percentage points more likely to make the payoff-maximizing choice in visit 2."

Ever Failed, Try Again, Succeed Better: Results from a Randomized Educational Intervention on Grit

Tagged

  • expected payoff
  • rational choice
  • risk reward calibration

Nearby findings