Subscribe to DSC Newsletter

5 Signs That You Are NOT a Data Scientist

Data Scientist is the rock star job title of the moment, and why not? There is a huge demand and a very small pool of qualified candidates. And those who get hired are making big six-figure salaries — who wouldn’t want in on that?

 

But just because you aspire to be a data scientist doesn’t mean you already are qualified to be one.  Here are a few signs you’re not qualified for that unicorn job yet:

 

  1. You don’t have the advanced skills. One study of data scientist salaries found that 88% of data scientists have at least a Master’s degree, and 46% have a Ph.D. The fields vary from math to statistics, computer science to engineering, or even economics and operations research. But the fact is that without the training of an advanced degree, it’s incredibly rare for someone to have the technical skills necessary to be a data scientist.
  2. You come from strictly a research or academic background. All that said about advanced degrees, people that only have experience in the academic world must work on developing their business acumen. The best data scientists will be able to relate the pure data to the real-world business applications.
  3. Excel is your primary analysis tool. If Excel is your workhorse, you might be working with data, but you are not a data scientist. But you probably already knew that. On the other hand, knowing how to use Hadoop, Python, and AWS doesn’t guarantee that you’re right for the job either, unless you can back that up with examples of experience with unstructured data.
  4. You don’t add anything to the data. For me, the most important quality of a data scientist is the ability to add value to the data through analysis and interpretation. Anyone can present the facts that data provides.  A good data scientist will be able to present those facts along with interpretation and visualization that will assist the executives and non data scientists in the organization make sense of it and make important decisions.
  5. You aren’t creative.  It might be just the stereotype of the scientist or statistician, but people don’t tend to think of a data scientist as being creative.  However, creativity is a key trait for a good data scientist because in the end, you are a storyteller. Data is useless without context, and it is the data scientist’s job to provide context and show how the data can help solve complex problems.

 

Of course, if any of these applies to you, it doesn’t mean that you will never be a data scientist, just that you’re not there yet. There’s honestly no such thing as an entry-level data scientist, but there are any number of jobs that could lead to a career in data science if you concentrate on building the right skills.

 

What would you add to my list of signs that someone is NOT a data scientist?  I’d love to hear your thoughts in the comments below.

 

DSC Resources

Additional Reading

Follow us on Twitter: @DataScienceCtrl | @AnalyticBridge

Views: 17382

Comment

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

Join Data Science Central

Comment by John L. Ries on May 18, 2016 at 11:04am

Sorry but science is about knowledge and understanding; not about how businesses can apply it (that is the province of management consultants).  So yes, you can be a data scientist with a purely academic background, just as you can be a chemist or physicist with such a background; you may not be employable outside of academia, but you can still be a scientist.

Comment by Sione Palu on August 27, 2015 at 4:07pm

Quote :  "There is a huge demand and a very small pool of qualified candidates."

Sorry Bernard, but that's completely false. There are tons of them out there who can do analytics.  Data science is basically analytics, end of story. Everything else like hadoop & big data, can be learnt on the job, there's no big deal there. Someone who's got grasp of deep data analysis is qualified,  from non recognized domains as Electrical Engineering, Physics, Computational Economics/Finance to Biomedical Engineering.

Its the people who think they know data analysis are those that think there are shortages out there. Remember, Data Science is a recently invented term, but other disciplines have been doing large scale data analysis for decades, but their job titles were'n't called data scientist.

One of the big Wall St  hedge fund is Renaissance Technology which was founded by Prof James Simons of MIT in the early 1980s.  His quantitative R&D researchers are physicists, mathematicians, astronomers and computer scientists?  He hardly hire economists or those with background in finance. You have to ask why? Well, because the type of backgrounds that he looks for do come from a deep data analytics domain, but not because they have some job titles called data science. 

"Renaissance hedge fund: Only scientists need apply"

http://www.reuters.com/article/2007/05/22/simons-hedge-idUSN2135575...

We have bypassed some aeronautics engineer (control system design), mathematician (optimizations), biomedical engineer (medical signal processing), simply because they miss out making the interview short lists. We knew they were all qualified because their backgrounds are modeling & data analytics, but we only have limited number of positions.

There are lots of data analysts (or perhaps we can call them data scientists as of today's terms) out there. I can find them easily because they're abundant out there. Some may ask why I think I can find them easily. Well, because we don't advertise for someone with all the big data hype terms. We emphasized first & foremost that quantitative background is what we look for. Then computer language comes 2nd.  Hadoop & other big data technologies come 3rd (as it can be learnt on the job), and so forth. Math skills & Computer language knowledge can be said to be one sets of skills because its hard to find a mathematician or data analyst these days who don't do their analysis using computer packages.

Comment by Dr Vincent Micali on August 27, 2015 at 8:25am

Yes indeed I concur in the comments made here below. Your 5 signs might not be the only ones but they certainly are the most important ones and you seem to also have elegantly ordered them. Well done and thank you. your article will provide a robust goal to those that are aspiring to be Data Scientists and a benchmark to those that think they are.

Comment by Vincent Granville on August 26, 2015 at 2:21pm

The gap between academia and business depends a lot on the university. Some colleges partner with big companies, allowing students to work on real big data projects and do internships. However, many still teach mostly old R, sometimes Python, and old techniques (see the textbooks that they use, the content has barely changed since 1960), with limited practical projects offered to students, making graduates of these schools even less competitive than someone who self-learned the tools, or someone who learned in a business environment, on the fly. But I believe that there is also a place for true data science research. Maybe in academia, or as in my case, independently in my own research lab and intellectual property foundry.

Comment by Benjamin Bertincourt on August 26, 2015 at 1:55pm

Thank you Bernard, this is a neat and well summed up compilation of the most fundamental traits that we are looking for in data scientists.

I feel that point 3. is not exactly at the level of the others though. Maybe you meant here that data science implies the use of tools that are quite a bit more involved and in general a data scientist has a certain expertise with advanced algorithmic and development paradigms (machine learning, mapReduce, version control, etc ...) that you don't necessarily expect in Business analysts (although that title requirements did evolve too).

I need a bit of help on point 2. though and I must confess that I exactly fall under that category of a researcher moving from academia to the private sector so I am certainly biased (hence I'll try to be as nuanced as I can manage :). Where do you see the largest difference between business and research lies, and what constitute the business acumen that researcher tend to not develop in academia ? The way I experienced it, researchers have to stick to deadlines (experiments are time constrained and fundings do not last forever), they have to justify of their time usage and results (you have to reach tenure to suddenly get relieved from this and that does not happen before your 40s), they have to recruit people and advance projects on lean budgets that they often manage themselves, finally they have to raise more funds through grants and proposals. This last item is not anecdotal, this is used, and very effectively I might add, as a way to ensure that the science that is being done follows a strategy toward objectives decided at a much higher level. Overall, the way researchers work has evolved a lot in the past 40-50 years with the centralization of administrative services in Universities and institute leaving very few personnel per department.

I was not attempting to be vindicative here, I am really trying to understand whether there really is a big gap between academia and businesses nowadays or if that gap is perceived. I also want and need to find out whether I am missing what business acumen relates to and how I can work on it without already being on the job.

Finally, I feel that domain knowledge is most definitely important for the job, but again I would expect it to be something you learn while doing the job in most cases. Overall, I feel that junior data scientist as a position where you bring technical skills and learn to implement them in service of a business, might as well be a reality since, in practice, most employers have a hard time figuring out their needs in term of data analytics and the technical part is probably the easiest to envision.

Follow Us

Videos

  • Add Videos
  • View All

Resources

© 2017   Data Science Central   Powered by

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