Opinion
Like most Data Scientists these days, I learnt (and am still learning) the techniques on the job. Back in 2014, I graduated with a scientific degree which taught me the foundations of data analysis, statistics and visualisation. After that, I fell into Data Science and landed my first job as a Data Scientist. Probably a familiar story for others out there too?
Earning the job title of ‘Data Scientist’ without having a directly related qualification was odd to me. I knew a few of the concepts, had worked with relational databases and dabbled in a small amount of code, but that was pretty much it. I also wasn’t as dedicated to learning back then as I am now, so most topics I learnt were dictated by the projects I was working on at the time.
I realised I was getting left behind…
As a few years passed, Data Science grew massively in popularity as a career path. As more online resources became available and data professionals were getting churned out of dedicated Data Science programmes, I realised I was getting left behind. My coding skills were underdeveloped, my statistics knowledge was rusty and my analytical approach to solving problems was quite frankly, primitive.
Now, you might be thinking that I’m being pretty harsh on that version of myself from a few years ago. Well, considering that a lack of effort in developing your technical skills gets you nowhere fast, I decided to change everything. I needed to be better. I needed to be more skilled. I needed to make a real effort so I could keep up with the Data Science market.
This wasn’t a straight forward journey. Learning how to learn again as an adult quite frankly, sucks. Pretty quickly, I ran into the questions: Which online course is the right one for me? Should I study a Masters degree? Should I just give up and cuddle cats for a living?*
A year or so into my learning journey, I had found an online learning platform that I was comfortable with and I was starting to learn and apply my new skills to my day job and reap the benefits. I was feeling better about myself. For a short while.
I would come across online quizzes or forums where I still had no idea what they were talking about, mentioning python programming concepts or statistical practices that seemed basic yet I didn’t know what they were or how to use them. I would be tasked with new projects at work and I wouldn’t even know where to start. How is this possible? All this dedicated time to learning and I still sucked?!
The challenge I was facing was caused by that great little human emotion called ego.
Fortunately, at around the same time as this realisation, the online learning platform I was using updated the programme to which I was enrolled and some new modules popped up that I hadn’t seen before. They had the titles like ‘Introduction to…’ and ‘Learn the basics of…’ and so on. At first I thought, ‘what a drag! I don’t need to go over the basics of Pandas or have an introduction to statistics, I’ve already learnt all this!’. Yet I had to now complete these new courses to finish the whole programme.
And that’s exactly what I needed.
The real trouble I had wasn’t my lack of technical knowledge or my ability to learn. Anyone is capable of learning. The challenge I was facing was caused by that great little human emotion called ego. My problem was thinking I didn’t need to go over the basics because I thought I already knew it. I assumed that, owing to my few years of working as a Data Scientist with my new found skill set and having learnt some of the ‘fancy stuff’ in the form of machine learning techniques, by proxy I must already understand all the basics.
WRONG.
This is by far one of the WORST mistakes you can make as a Data Scientist.
Assuming you have a solid understanding of a topic just because of your job title or the number of years you’ve held it can 1) make the learning experience of other topics harder because you won’t have all the prior pieces to the puzzle, and 2) lead you towards making grave errors when applying and interpreting models simply because you don’t understand the nuances and core data concepts.
Studying those courses that are aimed at complete beginners was the best thing I’ve ever done for my career. I had so many ‘ah-ha’ moments and was able to gain a more comprehensive understanding of the other complex topics I had previously tried learning yet couldn’t quite grasp.
I became a lot more confident as a result. I become bolder at work. I would challenge stakeholders to really dive into the data and work with them to ask questions and shape the analysis. We would aim to derive insights that they didn’t think would be possible to obtain from the data. I wouldn’t be afraid to learn new analytical techniques as I had a stronger foundation beneath me and so the learning process was quicker.
With my more comprehensive understanding of statistics, I was able to explain to stakeholders why the result was or wasn’t significant using real world anecdotes to aid my explanation, resulting in more engagement and trust in my work. Coupled with taking a course on data storytelling and how to write effective presentations (which I don’t think I had ever spent any time learning prior to this), I could share my results and recommendations in an effective way that would keep my audience engaged. Simple stuff, big difference.
My takeaway; be humble and make sure you know the basics.
If you don’t, read about the theory or take a course or two until you’re confident enough to tell everyone about the workings behind a neural network or whatever you just learnt. If you can have a conversation about the foundations of a Data Science topic with someone who is not a data person, yet they can follow along with your descriptions and explanations, then you’re on to a winner (providing you haven’t just made it up just to look super cool).
Oh, and in case you’re wondering what the asterisk is after the cat cuddling question?
*This option is still up for debate 🙂