We must preach what is practiced

In teaching non-specialists, we need to cover the procedures they will come across in their other studies and subsequent work. Since many courses and introductory texts do not do so, students are often disillusioned with them and regard such courses as irrelevant. I now describe some topics which we often teach unnecessarily (or with the wrong emphasis) and some which we mostly do not teach at all even though they are often needed. They range from deep matters like statistical inference or causation to more hum-drum ( but important) ones like means, medians and modes.