When I was ready to start doing more with data, I went looking for courses that would help me achieve three things:
- Test and refresh my understanding of basic statistical concepts when analysing or interpreting data
- Introduce me to the skills and tools I would need to work with big data, including the role for coding
- Experience learning through MOOCs (massive open online courses).
At the time I was looking Future Learn, which is run by The Open University and links to University courses from all over the world, was a good fit. They had relevant courses about to start, I didn’t need a background in statistics or calculus to take them and their courses are free, unless you choose to purchase a certificate of completion. An added bonus is that many of them can continue to be accessed after the formal completion date, which takes the pressure off if you are juggling work and family commitments while trying to learn.
There are lots of other courses and resources available out there. This list of free courses [http://www.openculture.com/freeonlinecourses], includes data science and statistics and there are websites which help you search for MOOCs. Coursera is one of the best known providers, including some of the top Universities. For example, three programmes (programmes being a series of linked courses) recently available through Coursera and all suitable for beginners are an introduction to social sciences research and analysis, using R; the practice of building and leading a data science team; and getting started on the concepts and tools to become a data science practitioner. These are all slightly more time intensive than the courses I selected from Future Learn, and are not free. They look interesting for more in depth skills development, when you clear on the area you want to focus on, rather than the broader overview I was looking to start with.
The first course I completed with Future Learn was Data to Insight, an introduction to statistical data analysis and to the free analysis software iNZight, which is based on the R programming language, but doesn’t require the user to be able to code. The course is suitable for beginners, although I would not describe it as basic, and structured so that those with more experience can jump around to the most relevant content.
This course is very useful if you are responsible for commissioning research or evaluation, or looking at outputs from data analysis. Through understanding some basic statistical concepts you will be better able to critique the work you are presented with, particularly around bias and sampling, relationships between variables and how conclusions are reached. Also for interpreting different outputs, including graphs, charts and statistical terms such as p-value and confidence intervals.
It is also a good place to start if you are a qualitative researcher looking to build your quantitative understanding and skills. Using iNZight to undertake practical analysis at every stage of the course really helped bring the statistical concepts to life and embed the learning. Plus, I was left with a working knowledge of software suitable for analysing relatively large datasets, of the type discussed in my post on administrative data. And the experience to compare, at some point, using a software programme based on R with using the R programming language directly.
The upcoming posts cover my experience of other courses, including social media analytics to analyse Twitter, an introduction to coding for data analysis and a big data analytics programme.