Subscribe to DSC Newsletter

7 Traits a Big Data Scientist Shouldn’t Have

A lot has been written (including by me) on what it takes to be a good data scientist; what skills and traits are essential for the job and even how to acquire them.

But what traits don’t make a good data scientist?

Remember, I’m using data scientist as a bit of a catch-all term to include data analysts, engineers, programmers, and more.  But taking into account that we want data scientists to have a balance of common sense, business skills, creativity, a geeky love of statistics, and data skills management skills, I was curious what sorts of traits would be the opposite? What would add up to make a bad data scientist?

1. Someone who doesn't like scientific enquiries.

Let’s face it: if you like things cut and dried, black and white, no grey areas and no questions, data science is probably not for you. You might think it’s all about facts and numbers, but the truth is, every set of data can be interpreted in multiple ways, and more often than not, an analysis of a set of data will lead to some answers and a lot more questions. So if you don’t enjoy the pursuit of one hypothesis after another, this is probably not the field for you. 

2. Someone who loves to make decisions on gut feeling.

There’s nothing wrong with gut feelings. Some of the best leaders and CEOs out there have built their empires on gut feelings. And while there is a place for intuition in data science, it has to be backed up by the facts. If you find that facts just slow you down, data sciences may not be for you.

3. Someone who only likes the detail.

If you’re only good at number crunching, for example, and don’t want to zoom out any further than that, data science may not be for you. There may be a place for you in the data world, but a data scientist has to have the business insights, bigger picture, and creativity to interpret the numbers.

4. Someone who doesn't like details.

Likewise, someone who is not interested in the details (only a big picture person), will have a hard time as a data scientist, because the details paint the big picture. If you don’t want to be bothered with the numbers, cleaning the data, understanding the statistics, then data scientist is not the job for you.

5. Someone who is closed minded.

We tend to think of IT guys and data junkies as very staid, very left-brained, and not creative. But if that describes you, you won’t be a very good data scientist. Data scientists must be creative, outside-the-box thinkers who can come at any given problem from many different angles. They must be able to flip a question upside down and ask it 15 different ways to find the best way. And they must be open to the idea that their hypotheses and preconceived notions may not be borne out by the data.

6. Someone who hates failure.

If you really can’t cope with failure, data science is not for you. Data scientists face failure every single day in big and small ways. A query might fail, you might discover bad data, a piece of hardware or software goes down, there’s a bug (and another, and another) and worst of all: you may discover when you get your answers that you asked the wrong question. Data scientists must be able to cope with these small failures and see them as just bumps in the road instead of dead ends.

7. Someone who has only one fraction of the skills.

You’re good at statistics. Great!  Are you also good at programming, engineering, visualisation and business insights?  A great data scientist cannot do his or her job without many different skills sets, so someone with only one or two of the required skills isn’t going to be as effective. This isn’t to say that these aren’t skills you can acquire, but without them, you’re sunk. 

These are off the top of my head, but I’d be interested what you would add to this list. What other traits would make a BAD data scientist? I look forward to your thoughts in the comments below.

DSC Resources

Additional Reading

Follow us on Twitter: @DataScienceCtrl | @AnalyticBridge

Views: 8384

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Michael Shakhomirov on November 20, 2015 at 2:23am

Good article! I have just crawled all the people who call themselves data scientists on LInkedIn and analyzed their skills. Machine learning is at the top of the skills! And I think the results are well aligned with traits described above.  Check it out - http://bit.ly/1SAiuQU 

Comment by Ketan Puri on November 10, 2015 at 7:58pm

Hi Bernard,

This is a good article. I think you have missed a very important point of what is Data Scientist should not be.

The trait related to "Causality" and "Correlations". A data scientist should not look for cause and effect but correlations. The data can uncover multiple aspects which may not be causal in nature.

Thanks and Regards,

Ketan

Comment by Kirk Borne on November 6, 2015 at 1:41pm

I agree with Mike's comment -- Data Science is now considered a team sport in most organizations. So, it is essential to include a diversity of persons, some of whom may be really good at only one thing. But I do agree that every member of the team must have at least a conversational knowledge of the other things that comprise the data science toolkit --- e.g., Advanced Database / Data Management & Data Structures; Smart Metadata for Indexing, Search, & Retrieval; Data Mining (Machine Learning) and Analytics (KDD = Knowledge Discovery from Data); Statistics and Statistical Programming; Data & Information Visualization; Network Analysis and Graph Mining (Everything is a graph!); Semantics (Natural Language Processing, Ontologies); Data-intensive Computing (e.g., Hadoop, Spark, Cloud, …); Modeling & Simulation (computational data science); and Domain-Specific Data Analysis Tools. 

Comment by Mike Kennedy on November 6, 2015 at 11:24am

Well said, Bernard. There are of course exceptions to every rule, but in general you've covered many of the traits someone ill suited to be a data scientist may have. I would add lack of curiosity or "happy to stop at the first answer" as one - data scientists need to enjoy digging for answers.

One that I would be careful of is #7. Data science teams are made up of data analysts, programmers, managers and generalists - which is the role you describe. But if you aren't an amazing programmer for example, it doesn't mean you can't be a successful data scientist on a team with someone to do the programming. This goes for visualization and presentation - that's a different role from the detailed data work. It's unrealistic to expect someone to be able to do the entire process.

Good stuff!

Videos

  • Add Videos
  • View All

© 2019   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service