R be dragons

August 18, 2010
By

(This article was first published on Timothée Poisot » R, and kindly contributed to R-bloggers)

Hic sunt dracones used to be placed on maps, as a way to denote a dangerous or otherwise unexplored territory. We might as well write it all over R-related material used in introductory classes, because students seems to be really afraid of what they will find.

Last year, I had the chance to teach a bunch of master students some fundamentals of R. The course was called « Analysis of biological data » or something along these lines. It was a joint course in biostatistics and an introduction to R, and I was in charge of two groups of students for this last part. I will be doing it again this winter. I was really surprised by the fact that these students were not really looking forward to the first class.

After asking them what they expected from the class and why they were anxious about it, they told me they feared that the combination of statistics
(statistical thinking seems to be something we lose, because it is obviously active as soon as 18 months!) and computing will result in too much new concepts to handle.

Yes, R has a steep learning curve, a fact that some used to call it an epic fail [what is an epic fail?]. So does LaTeX, so does C, and so does almost any other language. If we wanted to use something with a smooth learning curve, we would use Statistica or JMP. That would allow the students to go into a button-clicking frenzy, they would get test statistics and p-values by the dozen, and… well, they will likely do the same thing that I did when I was a student using Statistica and JMP : perform a bunch of tests that are not necessarily valid given their data, have no global understanding of how things work, and learn roughly no statistics at all (or, it can be that I was a disastrous student…).

Because no matter how difficult it might seem to be proficient in R, you end up learning a lot of things in statistic in your daily use. Just the help pages for most tests are rich in information. So, the goal is not to make the learning of R easy (actually, I guess it is, but it will never be natural to some people). It is to design the introduction to R so that is will maximize the interaction with statistics, and help learn things jointly in the two fields. To this regard, I think that a document like R for beginners (PDF) is not optimal (while it is great and I used it many times, I always had the feeling that it expected you to know R in order to learn R), while simpleR (PDF) or IPSUR do a great job.

Does any of you have experiences in teaching simultaneously R and statistics? I’ll be curious to know how you approached the situation…

To leave a comment for the author, please follow the link and comment on his blog: Timothée Poisot » R.

R-bloggers.com offers daily e-mail updates about R news and tutorials on topics such as: visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...



If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Tags: , , ,

Comments are closed.