An upsurge in paid work meant professional development took a back seat for a while. In this post I talk about finding a realistic way forward for developing my data skills, within a limited amount of time.
In my last post, ten months ago, I described lessons learned from dipping into different software and programming languages in a fairly random way (i.e. based on what was free, at introductory level and potentially useful to the kind of work I do). A bit of trial and error, and a bit of useful reading, gave me a clearer sense of direction – to pursue using R for data analysis. But to get motivated, and maintain momentum, I had to overcome unhelpful self- talk along the lines of:
‘what’s the point, it moves so quickly I can’t keep up, it will all be different by the time I am back working enough hours to use it on a project, ugh I will probably have to reinstall R or at least update it because I hadn’t used it for so long, I only have 4 hours a week (assuming no holidays, sick kids etc.), I’m not good enough at statistics to make it worth doing, I don’t have time for the immersive approach everyone recommends, ooo look what’s on Netflix…’
Like Goldilocks, I had been losing too much time on things that weren’t quite right – courses that were too hard, or not quite enough . This meant I was stopping and starting, always rebuilding momentum from scratch. I needed a path that could be broken into small steps for the pockets of time I have available across any given week. I also needed to go back and cover the very basics of working with R and RStudio. And I am interested in links to computational social science / data analysis for social science.
First, I tried this course from Udemy. It was on sale for $10, had no set date for completion and positive reviews. You listen to a series of short lectures, very bite sized chunks, with practice exercises to work through at the end of each section. While I often forgot lots of the detail by the time I started the practice exercises, this was no bad thing. It mimics real life – I had to find the data sets, upload them to my RStudio and work out how to do the things I couldn’t recall. This is different to the embedded and self-contained practice environment of other online courses I tried.
But this course is not an introduction to statistics or R, and assumes a working knowledge of both. I had to use online resources to work out how to set up working directories, upload files and retrieve previous work. While a problem-solving mindset for programming is helpful, it is not the most efficient way to work out the basic functions when you are just starting out.
For back to basics with R and RStudio I tried two of Data Camp’s free courses. Introduction to R explained the basic terms and walked me through the main functions. The first chapter of the Introduction to R Studio IDE course is also free. It is a good intro into how RStudio is organized, around four main panes (console, environment, files, source) each with subtabs. I sometimes flipped between the DataCamp taster exercises on their website and trying things in my own version of RStudio, to check I was doing ok in the ‘real’ world.
The experience with DataCamp was sufficiently positive that I signed up for a year’s membership (I paid £229, they often have deals that come up with discounts of up to 50%). DataCamp has a data analyst with R track (plus access to everything else). It met my ‘pick up and put down in small pockets of time’ criteria, with learning broken down into bite sized chunks. And if all else fails I at least try to do 10 minutes per day via their daily practice challenge. The price compared favorably with other things I had been looking at, and that I might come back to. I can also combine this path to make more efficient use of the Udemy courses, as I will have a better grasp of the basics, and to design my own practice analysis. I will let you know how it goes.