I talked about a study that looked at dropouts and recently came across the full text. The authors used a ‘cocktail’ of three remedies proposed in the ‘pre-Mturk’ world and found that these remedies can decrease dropout rates.
“Reips (2000) proposed three other nearly costless strategies that
could nudge participants who have consented from backing out:
(a) prewarning (e.g., telling participants that “you will be writing
about the goals you want to accomplish in your life”); (b) person-
alization (e.g., asking for individuating information such as email);
and (c) appealing to conscience (e.g., telling participants that
dropping out could affect the quality of data and that science needs
good data). However, given this proposal was not tested, here we
evaluate its efficacy. Because from a practical point of view,
gauging the relative merits of the three strategies Reips proposed
is of little value, we combined all three strategies to create a
Here’s the Reips cite:
Reips, U.-D. (2000). The Web experiment method: Advantages, disadvan-
tages, and solutions.
Psychological Experiments on the Internet,