So let's just ignore the fact that I haven't posted in six months and skip to the good stuff. Wednesday is April 1st, the day when friends become enemies and enemies become meaner enemies. All of a sudden, everyone you know is lying to you, trying to convince you that your car was hit by a meteorite, or that there are free dilly bars in the office freezer, or that the tic-tacs that they are offering you *don't* taste like habanero peppers. And these, of course, are the relatively innocuous pranks that people can do. Remember, a good prank is something that gets entirely resolved by saying "April Fool's". Hiding around a corner and hitting a passerby with a 2x4 is not a good prank because it fails in to meet that criterion, and to a lesser extent so does telling somebody that their family was eaten by a whale. Of course, the second one also fails just about every other criteria too. So let's reiterate: don't attempt either of those.
But, for as much distress as it causes me, I still participate every year. I'm part of the problem, I know. But anyway, I thought I'd give a sneak peek (also known as completely explain) what I'm planning for this year, because A) nobody who reads this knows who I'm bepranking, and B) it might inspire others to bigger and better things.
So, some context: As part of my job, I'm to be heavily involved in the planning stages for a lot of health-related studies, and a lot of them use pigs as test subjects for emergency medicine studies that you could never do with people (like overdosing on beta-blockers and randomly assigning them to current or experimental new detoxifying treatments, for instance). But, as the statistician, I don't actually collect any of the data myself, so I have to stay in touch with the researchers, or as we in the biz call them, the investigators.
From: Christopher Anderson
Subject: Investigator needed for new study
Dr. ______,
I have an idea for an new study proposal, and I am looking for someone in the emergency department to take the lead as principal investigator. As you know, many retrospective studies of bicycle accidents have found an association of mortality and severe head injuries with not wearing a helmet. However, I think a prospective, controlled trial in needed to prove this relationship definitively.
By using pigs as test subjects, we can randomly assign them to either have a helmet or not, and then supply each with a traumatic impact. Considering the resources available to us at the Emergency department at Regions hospital, the quickest, most economical and reproducible way of supplying such an impact is to take test pigs up in the life link helicopter to an altitude of exactly forty feet, and push them out over the parking lot.
Outcomes of interest are (I) survival and (II) memory loss, as assessed by time-to-completion food pellet release puzzles. The null hypothesis to be tested is that there is no difference in outcomes between pigs that wore helmets and those that did not, against the alternative that there is a difference in survival and memory loss between groups. This sets up a straightforward statistical analysis via t-test.
Let me know what you think! Look forward to hearing back from you!
-Chris Anderson
