A TikTok video popularized a survey site for scientific studies: so much that it has skewed the data making it unusable
If someone publishes promotes your service for free because they genuinely appreciate it and as a result you get tens of thousands of new users, one sees it as a good thing in the first place. However, sometimes it can be so positive that you die of success (the Waze example is a classic). This is what, more or less, has happened to the Prolific survey site for scientific studies. In a matter of days they saw how their polls were corrupted by the skewed user base they got.
It all started on July 23, when Sarah Frank, a young TikTok user, posted a video about Prolific. Prolific is a survey website for scientific studies on behavioral research. In exchange for answering questions and tests in surveys, users receive a few dollars per survey (around 10 euros per hour). These surveys are subsequently analyzed by scientists to carry out their studies. A win-win mutually.
i’m excited for this series!! #sidehustles #sarahssidehustles #prolific #makingmoney #sidehustleideas
♬ original sound – sarah 🙂
According to Sarah Frank in the video, for her it is an easy way to earn some extra money every week conducting a series of surveys. He encouraged his followers to do the same if they wanted to earn some money easily. Said and done, with more than 4 million views, the video saw Prolific register around 30,000 new users.
11 TRICKS to MASTER TIK TOK
Women in their 20s
The problem that Prolific had with this flood of new users was that same: they were women in their 20s, like Sarah Frank. Her followers are mostly people of a similar age and gender, so most of the new users who came to Prolific were from a very defined segment of the population. This, for a survey website, is deadly.
If a survey for a scientific study is carried out with the idea of achieving the most accurate representation possible, the survey should be carried out on a sample that is as similar as possible. An analysis of hundreds of recent Prolific surveys found that 91% of the people who took the surveys were women. This, indeed, is not a representative sample of the population.
Due, many, many studies that were being carried out with Prolific surveys have had to be interrupted. Researchers fear that their data is biased and therefore useless in practice. Prolific took action and allowed researchers to conduct new surveys free of charge with demographic limits to avoid this bias.
The researchers and the platform were not the only ones affected, also the users themselves. At subreddit from Prolific many users began to complain that surveys are now paid less and less are available too. This is essentially due to a question of supply and demand, the more users sign up, the less money there is to distribute. What was a win-win for everyone it has become a lose-lose.
Via | The Verge
Image | Solen Feyissa