Conclusion Online digital labor markets are fraught with dishonesty, due to both the anonymity services Amazon MTurk provides its users and the incentive to complete as many tasks as possible in as little time as possible to ensure maximum revenue. This tendency to die was proven using a simple coin flip test. Users were asked to flip a coin ten times and to report the number of heads they flipped for each head, and an extra ten cents was added to the task payout. The test found that these MTurk users routinely lie and shirk their duties. It was also found the introduction of a voluntary oath of truth had a significant effect on the number of heads reported, with MTurk users who were asked to take the oath reporting numbers much closer to the true average than those who were not. The experiment also found that while the oath did have a marked effect on the number of truthful answers, it did not affect the likelihood of dishonestly lying about having completed the task and making up a number.
Response The efficacy of the voluntary truth-telling oath took me by surprise, but what I really found interesting about this research paper was the oath's failure to reduce shirking. A theory: could it be that the MTurk workers feel more obligated by the oath to answer the demographic survey more truthfully, compartmentalizing it and then disregarding the coin flip test?
The conclusion makes an interesting point about the MTurk worker placing significance on certain parts of the survey and neglecting others, and I feel like this could be studied further, perhaps in another, this time non-online, test of a similar nature.
“Do Truth-Telling Oaths Improve Honesty in Crowd-Working" is an interesting exercise in industrial and organizational psychology, especially in the fields of online, anonymous workplaces: a new industry.
