Previously we faced problems when the workers did not follow our instructions producing incorrect data. Data quality and variability is of high importance to us, therefore we want to be sure people who work with us understand this. Now we are creating a pool of permanent mturkers whom we can trust and who will work on our tasks going forward. We are going to raise the price per hit for these workers exclusively as we understand you need more incentives to work on the hits. We kindly ask you to try our hits and propose your friends doing the same. Any feedback from your side on how we can improve is welcomed.
These are the hits launched by us recently:
1) Restaurant Reservations Survey - 1 hit/person
2) [Restaurant] Find the robot's bad responses - Bot Message Bonus - Follow Scenarios (There are 1-6 scenarios) - 1 hit/person for each scenario
3) Your first message to the assistant - scenario based (1) (There are 1-4 scenarios) - 1 hit/person for each scenario
4) [Restaurant] Ask bot to make restaurant reservations in US - we run these on daily basis, 1 person is allowed to work on 3 hits per day; workers receive bonus for each correctly identified robot's bad response
5) [Movie] Ask automatic bot to buy movie ticket in Seattle area - we run these on daily basis, 1 person is allowed to work on 3 hits per day; workers receive bonus for each correctly identified robot's bad response
Some of the tasks you can do once (1 hit per run) and some are available to the US residents only. These restrictions are to ensure language proficiency and variability violations. This and more information is available in the task instructions, so please read them before accepting the work. The restrictions are temporary. Hope to see you working on our tasks.