http://scienceoftheinvisible.blogspot.com/2009/02/lies-damned-lies-and-frustratio

This term I will be mostly teaching statistics. To nearly 300 students split between first and second year. In two hours a week, which is where the problem lies.How can I say this diplomatically? As a population (SPSS has rotted my brain), biologists are not the most numerate of scientists. And yet - most think that statistics is the one bit of maths that they can do. So they get very upset when they find that they can't. "Can do" translates into English as "can work out the mode but have never used any statistics software before". Real world fail.

The conversation goes like this:

Dr Cann, the significance value is greater than 0.05, so I accept the null hypothesis, right?"The system" (gosh, does he mean schools?) has trained them to "do" statistics. But not to have any understanding of statistics. It has trained them to use a calculator, generate a number, write a report, without any clear rationale.

Good. What is the null hypothesis?

Tumbleweed.....

And then "we" come along and make things worse, by assuming that the piece of paper they have that says they have "done" statistics means they can "do" statistics. So we timetable statistics into our curriculum for two hours a week, but kill all hope by giving it the fatal "key skills" moniker.

And they emerge with a science degree and a piece of paper which says that they can "do" statistics. Sigh.