Saturday, 8 June 2013

Student choices between SAS and R in Individual Coursework

This is the third post is a small series of posts reflecting on my teaching in a new class (all teaching materials can be found here) introducing students to SAS and R on our MSc course. The two previous posts have so far just been a reflection on student's attitudes towards each piece of software.
In the first of those posts I said how I was slightly surprised at how students had chosen to use SAS for 1 particular question where they had the option of the language. In my opinion it was a problem much easier to tackle with R (all the course works, class tests etc can be found on this site). I also mentioned in that post how I asked students which language they preferred. Almost all students answered that it depends on the task (which is a great answer) but after pushing them for a particular decision a strong majority seemed to prefer R. I posted on G+ recently about a particular interaction I've had subsequently with a student which seemed to confirm this attitude of needing to find the correct tool for the correct job.

In the second post I described how students in their group presentations (asking them to teach me something I had not taught them) mostly evaluated SAS v R for certain tasks. It was great to see them identify strengths and weaknesses for each language.

This post is about their choices in their individual coursework.

Similar to their class test (which I discuss in the first post of this series) there was a question which allowed students to choose a language in this individual coursework component of the class (which can be found here).

In my opinion this question made use of quite a big data set (generated by some research I'm currently doing) and I thought it would probably be simpler to approach in SAS. About 52% of the class agreed with me whilst 48% seemed to still prefer R. First of all, I could be wrong and R could indeed be better suited for this question, secondly it might also be a reflection of the personal preferences that the students seemed to indicate when I asked: most students seemed to prefer R. If the latter is the case then I suppose it's nice to see that students not just realise that there's a better tool for a given job but also a better tool for a particular person doing a given job. I'll be keeping a track of this over the years and see how (if) it changes.

In my next post in this series I'll start to reflect on some of the teaching methodologies I used.