Science Friday interviewed a researcher who looked into how various bits of signage might increase the incidence of hand washing.  They were able to get a large sample by setting up shop in a rest stop.  A nice example of A/B testing.

Here is the key table from their paper.

The table has one section for men and another for women. What they put on the sign is shown in the first column and what lever they thought they were pulling is shown in the second column. In some cases I have my doubts that the statement was pulling the lever they think it was. The last column shows how effective the sign was compared to doing nothing.

I was bemused that the study authors thought it notable that what works on one group isn’t what works on another group (i.e. men v.s. women). I very much doubt the authors are unaware of the vast literature on the practices of discriminatory marketing. Presumably in a country like England which has deployed Orwellian social monitoring they should be able to customize the message based individual profiling. Web advertisers certainly try.

Experimental Pretesting of Hand-Washing Interventions
in a Natural Setting
Gaby Judah, BA, Robert Aunger, PhD, MSc, Wolf-Peter Schmidt MD, MSc, Susan Michie, DPhil, CPsychol, Stewart Granger, PhD, and
Val Curtis, PhD, MSc