FTA: "The same bias was exhibited by students in a laboratory setting."

The GP was talking specifically about the experiment design. Surveys are also notoriously unreliable as sources of meaningful data. Not that either point discounts the value of the research, of course.