The world’s leading publication for data science, AI, and ML professionals.

Debunking Data Science Myths

The Truth They Don't Tell You About Data Science

Photo by Manyu Varma on Unsplash
Photo by Manyu Varma on Unsplash

When I first set out to become a Data Scientist I’d often hear/read various myths being thrown around the Data Science community. Now, I believe many of these myths have matured and evolved into even more annoying myths. You’d often find that these myths usually stem from the misconceptions of people who’ve struggled to break into the field, people of high influence with the field expressing subjective views that are misunderstood by followers, or common confusion in general.

"Data Science is all About Model Building"

Let’s start by introducing Kaggle; Kaggle is a very popular Data Science platform that hosts a flurry of various Data Science competitions. The team that comes up with the best solution, meaning that they have the best score on the leaderboard, are deemed winners along with other people also built high scoring models.

This isn’t to discredit Kaggle or any of the other Data Science competition websites because I actually think they are great learning resources and one of the best ways to improve your modeling skills. However, many Data Scienetists are often tipped to go to these places to practices. If you’re new to Data Science and all you’ve been exposed to is various algorithms, as well as competitions, it is quite easy to conclude that this is what makes up Data Science.

There is actually much more than building models to Data Science, in fact, modeling is typically what takes up the least time in comparison to data collection, data cleaning, etc. It’s often easy to miss this because much of the data we are given is already collected and processed when we take part in competitions, and the majority of our focus is on optimizing to for a certain evaluation metric.

You Need A Technical Degree

I really thought this misconception has been cleared up by now, but still I recieve many messages from people asking whether it’s necessary to go to university to become a Data Scientist.

Now, the answer is not so straight forward because though you don’t need a degree to become a Data Scientist, there is still many benefits in going to university and earning a degree. For example, it is one of the most natural forms of networking available, and you will get a deep theoretical in whatever technical subject you decide to do which often translates well into Data Science.

Despite what anyone else says, I would always say that one should never completely write off going to university, but keep at the back of your mind that it is not essential that you have a degree to become a Data Scientist. Therefore, it is essential you establish and understand the motivations you have for going to university.

Note: Other factors come into play also such as considering your financial situation and more.

Python is The Best Language / R is the Best Language

Initially I thought this debate was banter so I laughed it off. It turns out there are people that are passionately engaged in serious debate about what programming language is better for Data Science out of Python and R. Additionally, there are also many newcomers that remain in limbo because they don’t know which programming to begin with.

Here is my opinion; Whatever programming language you decide to use, learn it and do it well because It truly doesn’t matter. The most important part of being a Data Scientist is to use your skills to find valuable insights in data which means the tool you use to do it doesn’t matter as long as you do the job.

To decide what programming language you should learn as a beginner is quite simple. Go on job boards and read the job descriptions of various jobs that you think seem interesting and learn the most common language from them all – You’d probably find that most job descriptions say you may know one or the other (it doesn’t matter).

Wrap Up

There are most definitely more Data Science myths but these were what came to mind at the time of writing. If you can think of some, leave a response in the comments and I will clap for it.

Let’s keep the conversation going…

Kurtis Pykes – Data Scientist – Upwork | LinkedIn

If you like this post you may also like:

How I Stay Up-To-Date With Data Science

How I Levelled Up My Data Science Skills In 8 Months

5 Hacks To Overcome Data Science Imposter Syndrome


Related Articles