Your data may be tangled, but beware that when you've wrangled, the signal isn't mangled.
March Results:
The cartoon for our 106th cartoon caption contest in March is shown above. Amongst the 27 submissions for the contest, the judges found the winning caption for the cartoon to be “Your data may be tangled, but beware that when you've wrangled, the signal isn't mangled.” submitted by Jerry Nedelman from the TB Alliance. Jerry’s caption can be used to illustrate the idea that over-cleaning or over-smoothing messy data may risk the loss of important information and yield a misleading characterization of the data. An honorable mention this contest goes to an anonymous submission getting at a similar point: “Reducing dimensionality reduces variability - but where'd the motorcycles go?” A second honorable mention goes to Nathan Shiosaki, a student at University of Colorado Boulder for the caption “Using principal car analysis, we were able to merge the congestion of vehicles into a single lane of cars.” Nathan’s caption is designed to encourage students to think about methods like principal components to reduce the dimensionality of data.
Thanks to everyone who submitted a caption and congratulations to our winners!