A joke for discussing how transformations can make data more normal and stabilize variances across groups with different means (here the square root transformation for Poisson data). The joke was written in 2016 by Larry Lesser from The University of Texas at El Paso.