I’m at the UserR! 2010 conference at NIST in Gaithersburg, Maryland. This is the main annual event for R users. there have been whole series of presentations on using R in education. The full program lists the Pedagogic talks (3 sessions, and 9 talks, on the first day).
What I’ve seen so far is great work on training people in data analysis, in statistics and (to some extent) in probability, The work is really good, and I have lots of new ideas. What’s been lacking, and what I need to think about more, is the other part. There are at least three other elements,
- asking intelligent questions, that is questions that are well enough specified to be answered, and well enough considered to actually matter.
- using research information, and clinical information, to support good clinical decisions
- data cleaning, data exploration,
Any thoughts?
Well, of course, I do not need convincing! I’ve been teaching distance learning courses based on R since 2005 – about 200-300 participants since….
But it is an uphill struggle getting health researchers to grapple with R when they have been inculcated with menu-driven packages such as SPSS (one of my pet hates). Many are resistant and feel that the learning curve for stats is hard enough without having to also learn R! But surprisingly many stick with it – thankfully. I’m not inclined to use the (as yet) primitive menu-like interfaces to R – this to my mind defeats the purpose – i.e. of encouraging statistical reasoning rather than pushing menu buttons…
Alan