I dunno a youtube video, Ken Rockwell's site goes over color histograms so search that. Basically you have independent Red, Green, and Blue channels so if only one of those maxes out (typically Red) you don't see it on the white histogram.
Converting from 8 bit to 16 is called upsampling, and it won't do anything to improve your image (It can make editing easier.) The way to improve dynamic range is to capture at the highest possible bit depth and to expose as far to the right as possible (google ETTR). All of this in raw of course, jpeg severely limits your postprocessing.