You working on statistical models is a scary thought, given that you seemed unable to understand my criticism of my primitive analysis of lofeatgoogle. I read

You working on statistical models is a scary thought, given that you seemed unable to understand my criticism of my primitive analysis of lofeatgoogle. I read a book in German (specifically on how to lie with statistics), don't know which ones to recommend in English. There is the obvious one with the name "how to lie with statistics", but it is very old. Otoh, available for free (but I haven't read it). Feynman, too, don't forget about Feynman. http://www.ar-tiste.com/feynman-on-honesty.html http://www.ar-tiste.com/feynman-on-honesty.html Thinking, Fast and Slow was a lot about bias, but not sure if it really gets the point across effectively. Seeing how you interpret my comments, maybe starting with statistics is actually backwards. Perhaps start with language and communication first? Sorry if my comment sounds mean, but it really seems a bit crazy how you try to bend my comments into something outrageous.