The Slowness of Using an Unfamilliar Package

Since I came to this job twenty years ago I have used SPSS as my main package. This means I am highly fluent at doing things in SPSS. For instance I generated a set of data to demonstrate ANOVA on in a couple of weeks as finding a suitable one was not going to be easy. We had a couple of suggestions but they did not actually provide the data to demonstrate on. This took me a couple of hours using quite an esoteric part of SPSS. The major argument was creating the data structure beneath the surface.

However I have got to teach the course in R. Now I have used R and even its predecessor S Plus and I have got it to do the things that I wanted in the past but it is not a tool that I use every day. Therefore I downloaded the latest version and started to use it.

It has taken me two to three hours just to write a script so as to get the data into R into a form that I can go on and analyse. The thing is I will think something like "I really need to make a comment in the script here so people know what I am doing". In SPSS I know that I just need to put a "*" at the start and a "." at the end and that is it. With R I have to find what the correct term is, I think it is still "comment" then look that up in the books index and then hope there is an example which I can copy. It turns out that I just need to put a "#" at the start of the line to turn it into a comment.

Then I hit the real snag. I have a number of variables like sex where "1" is for "male" and "2" stands for "female" and I would like to have these appear as "male" or "female" in the data. This which is basic stuff I teach with SPSS (rather than technical knowhow that I use) is not well covered in the R books I have. Instead they go straight into rearranging the data and then instead of how to get simple descriptive statistics they go straight into graphs. Alright if I am teaching a package with unfamiliar data graphs are a good way to build users competence with the data and familiarity with the data set. So two hours later and I have just got the data in!

Meanwhile I am secretly wondering if I can't download the gui interface called R Commander. I might yet particularly if it generates scripting.