Qualtrics Validation Tutorial: reduce random responses from MTurk
Go to LacePadilla.com for more tutorials!
Qualtrics best practices for experiments using Amazon’s Mechanical Turk as a population.
While many studies have found Mechanical Turk responses are generally reliable, researchers know that in practice participants can respond randomly or don’t fully participate to the best of their abilities. The majority of MTurk’s population are good participants, and the trick is to insure that they are allowed to participate in your studies.
The primary way we have found to reduce random responses from 30% to 2% is to add a response validation question at the start of any survey. This tutorial will detail the step by step process of adding a validation question in Qualtrics.