Can MTurk be used for forensic science? That’s the question a new study asks because:
” For forensic psychologists, one important advantage of crowdsourcing is that crowdsourced samples might be more representative of people interacting with the legal system than typical college student samples.”
At first, I thought this meant that Turkers were more likely to be criminals, but then figured out it meant that Turkers are more likely than college students to be witnesses to crimes or to serve on juries.
This is basically another study that compares MTurk demographics to a student sample. This one found that MTurk was less ethnically diverse than some student samples. It all depends on the school, of course.
This study did have a very high attrition rate: only 69% of people who started the task finished it. The authors suggest that this was a 3-part task, where part 1 was challenging (solving anagrams) (really?) and there was a 24 hour wait between part 2 and part 3. I can definitely see how this waiting period would cause people to drop out.