Randomised experiment: If you’re truly not sure whether or not to stop your task or split up, then chances are you most likely should

Simply how much can we trust the ‘quit job’ and that is‘break-up? Regarding the plus part:

  • This is certainly a nearly-experimental outcome (using the coin toss as sort of ‘intention to treat’).
  • The specification is consequently transparent and simple.
  • The outcome are statistically significant and pass some robustness checks.
  • The sign of the total result(positive) is plausible on its face, being explained by status-quo bias and risk-aversion. But, the magnitudes are unexpectedly large, so much more likely than perhaps maybe not the possibility overestimate.
  • Levitt actively seeks indications of some types of bias ( e.g. individuals being inclined to overstate their pleasure if they obeyed the coin flip, or those whom benefitted through the modification being almost certainly going to fill in follow-up studies) and discovers evidence that is little them. Continue reading