" The hardest part is trying to influence the nature of the measurements obtained, so that the key information can be obtained."
As quoted from his own website: http://www.columbia.edu/~jeh1/interests.shtml