Urgh ugh ugh ugh

This evening I discovered what could very well be a disaster for my study at work. The study has a delayed treatment control – after finishing the second of two surveys, 7.5 months after signing up for the study, the participants in the control condition get access to the intervention (the website and a mailed booklet). Tonight, though, I discovered that about 27% of the control condition participants have been using the website. Someone somewhere coded something that lets them in. This is pretty ghastly. I cannot possibly afford to start the recruitment all over again. I probably can’t even just drop those 68 “contaminated” people, either, because they’re not in there at random. My subcontractor is investigating it, and I sent an e-mail to my project team – hopefully one of them will have some valid shortcut to fix the thing, because the way it looks right now, someone has ruined my $3.7 million study.

At least my car works now.

Update – I've looked at it some more, and now I’m thinking that maybe I’ll only have to replace the 68, or some fraction of them. I’ll feel better when my project team weighs in with more than sympathy.