Study Finds Significant Number of Employees Spend Work Time Completing MTurk Tasks
More and more university researchers are turning to online services to recruit people to provide insights and answers for their projects and surveys. One of those services is Amazon’s Mechanical Turk, or MTurk, where people are paid for responding to surveys, among myriad other tasks.
Lynn Offermann, a professor of industrial and organizational psychology at George Washington University, has been watching the growth of MTurk since it was launched 10 years ago. She recently conducted a study with GWU graduate students Meredith Coats and Caroline Rausch to determine the extent that MTurk users were participating in the service while at their regular place of employment.
Their findings, presented at SIOP’s annual conference in Philadelphia last spring, suggest nearly half of MTurk users responding to a survey had used MTurk at work. Further, the group studied employee’s feelings regarding performing nonwork tasks during work hours, examining what is known as the “psychological contract.” A psychological contract is an employee’s belief about the expectations their organization holds of them through expressed or perceived messages about their job responsibilities.
“We examined the perceived psychological contract and job attitudes of MTurk workers who complete MTurk tasks during work hours while being paid by their employer as well as those who do not use MTurk and the implications of those actions on their work performance,” she said.
Nearly 46% of respondents admitted to completing MTurk tasks while being employed to do other work and, of those, 41% did it daily, 27% once a week and 26% rarely. Also, 33% acknowledged that “MTurking” reduced their productivity at work, a figure that Offermann found alarming.
The study suggested that employees are of two different minds as to what they owe their organizations in terms of how they use their time.
“Some feel that once they have completed an assignment there is no obligation to seek other work-related activities and any extra time is theirs to use as they see fit,” Offermann said. “We found that to be true of nearly half of the 551 respondents. They said it was acceptable to engage in MTurk between assignments and that the organization either did not know or care they were engaging in this behavior.”
Other employees view their responsibility as moving on to other useful activities that are beneficial to their job and the organization. Not surprisingly, those workers are least likely to participate in MTurk tasks.
MTurking can be an issue in the workplace. There is evidence that employers are not aware some workers are using work time and workplace computers to perform MTurk tasks.
“I suspect that lot of organizations would not be happy if they knew that MTurking is going on,” she said.
Though most organizations do not have a problem with employees doing some “cyber loafing,” like checking ball scores, sales, weather reports, and so on as long as it isn’t abused, and many people with smart phones are not using the organization’s computers and technology.
MTurk is popular with researchers because it is inexpensive, provides a large and diverse sample size, and the responses are immediate, something they don’t get with the traditional test cohorts of college students. It is estimated there are more than a half million “MTurkers” eagerly waiting to complete surveys and other tasks.
Pay is minimal, ranging from a few cents to a dollar per survey, depending upon the complexity and how long it takes to complete the survey. Completing MTurk tasks, called HITs, is certainly not a path to big money, Offermann said.
“However, money is a primary reason people engage in MTurking, despite being paid significantly less than minimum wage,” she added. “Also, many report they find the tasks interesting enough to fill their time between other work assignments.”
Nevertheless, MTurk has become a useful source of data for researchers conducting studies on a variety of topics, including I-O psychology. Previous research suggests that information collected from MTurkers is as valid as data from other sources, perhaps even slightly better than using samples of college students. To ensure greater reliability of data, I-O researchers often take steps to embed data quality assurance measures in their studies.
Offermann says a moderating factor in the use of MTurk is opportunity and whether or not completing MTurk tasks at work is a problem may depend upon what workers are supposed to be doing. For example, a cab driver at the airport may answer survey questions while waiting for a fare. A teacher or office worker could MTurk during lunch break. Those may be acceptable.
“There are a wide range of occupations that provide the opportunity to MTurk without hindering work,” she said.
Nevertheless, MTurking at work is clearly an issue organizations should address, Offermann said.
“The research suggests that employers should set expectations and make it clear what is appropriate and what is not in regard to the use of time at work,” she added.
Contact: Lynn Offermann at 202-994-8507 or email at firstname.lastname@example.org