What do you actually learn on a data science course?

Stephen Dawson
3 min readMar 23, 2019


Tell a story with your data — Image by Becca Dawson — https://www.instagram.com/becca_dawson94/

Through a mix of FOMO and curiosity, I decided to enrol on a junior data science course.

It goes without saying, the course was hard work. It wanted me to learn to code from scratch and I really wanted to keep up with the assignments.

As the course progressed, it was clear that there was more to it than knowing how to write a While Loop or run a predictive analysis with Pandas.

Your approach, unravelling perceptions, and being able to talk about data properly were the most valuable strengths to be taken forward.

I’l outline what discoveries, strengths, and skill-sets you can take from a data science course.

Unravel the perceptions

Let’s get a few things straight.

Big Data: is just something that’s used to handle very large data sets.

Machine Learning: is an improved predictive analysis.

Artificial Intelligence: we’re not even close to this.

Data Analysis: is what you should be focusing on now. Taking a data set, cleaning it up, and using it to strengthen your conclusions. Simple.

Yes, Machine Learning will be the most interesting concept you’ll learn. It will sound like it solves all of life’s problems. Despite this, you’ll learn it’s to be treated with caution. It might play well in concept but it is hard to apply it in real life.

Persistence works

Yes, coding is one of the pillars of Data Science that can prove to be the most challenging to master — but it can also be the most rewarding.

Coming from a very basic coding background (some HTML 5 and CSS), I was taken back when I was given my first assignment; learn Bash.

Keep calm; dropping the ball is the worst thing you can do. Don’t assume that once you’ve been able to crack the code on the day you will be able to do it on every asking.

Approach is key

How you approach a problem will ultimately define your success.

Separate out your tasks into logical steps. Test your theories with smaller parts of the overall problem. When you’ve validated your assumptions press ahead with the grand finale.

I have this theory, can I produce an MVP to test and refine?

Understand what one function will achieve and how it will affect other variables. Often, there is more than one path to a solution and being creative about it will help.

Problem-solving should be an everyday part of your role. Especially if you’re looking to improve the way you work.

For me, this was the most beneficial skill to learn.

It’s not all coding

Coding will help you succeed in setting up your data for analysis. The next step, turning it into a strong conclusion, can be tricky.

Firstly, go back to the basics — know the difference between mean, median, and mode. Understanding statistics will allow you to determine the most suitable root for reaching your conclusions.

Secondly, consider how to take your findings and translate them into plain English. Your audience will be one step behind you — think of how they will interpret you.

This is a highly desirable skill to have. Get lots of practice in.

Finding the right people

If you’re enthusiastic you will want to know who to listen to. Do a little digging beyond the top result on Google — there will be someone or something that appeals to you in your own way.

It’s safe to say networking will help a huge amount too, especially if you want to take the leap from digital marketing to a full-time data science job.

Wrapping up

Data Science isn’t magic — it’s common sense, requires sensible thought and respect. Keep at it, learn how to approach it, and be prepared to translate it to your audience.