Sampling Data

With a high median base salary of more than $116K and accolades including “2016’s Best Job in America” and “the Sexiest Job of the 21st Century,” the data scientist is becoming an attractive professional role. As more people apply for this highly regarded job, it is important to be able to differentiate a true, skilled data scientist from a fake when conducting a data scientist interview.

What to Look For in an Interview

To spot a fake data scientist, one must first understand what a data scientist is and what they do. Data scientists are innately curious professionals who are tasked with making sense of data. They work with both structured and unstructured data—organizing it, extracting insights and presenting their findings to business leaders, team members or clients. Data scientists often work with multiple data sources at once, looking at the data from multiple angles to fully understand what it means.

1. Know the difference between a data analyst and a data scientist

While both roles involve curious individuals who study data to gain insight and create a story, they differ in many of their responsibilities. Data scientists tend to use more advanced statistics and programming than data analysts do. Another difference is that data scientists don’t just analyze the data—they might acquire it, move it and manipulate it using machine learning, engineering or programming skills. When conducting a data scientist interview, you may want to ask them which specific tools they use to manipulate data and why.

2. Make sure candidates are expert problem solvers

Regardless of what a candidate’s resume might look like, a true data scientist will have strong problem solving and analytical abilities. Though a data analyst might be able to arrive at a correct solution, a real data scientist will go about problem solving in a specific way. A good question to ask a candidate during a data scientist interview is to solve a problem, then pay attention to the way in which that person thinks and acts.

3. Look closely at the data scientist’s resume

It is likely that a data scientist will have a graduate degree, possibly even in mathematics, statistics or computer science. Even without a graduate degree, data scientists should have technical experience in Python, Hadoop or SQL, if not more than one. They should also have deep knowledge of SAS or R.

4. Be sure the data scientist can show results

Beyond possessing knowledge of specific programming, a true data scientist will know how to effectively use it to solve problems. Instead of just asking whether a candidate knows how to use Hadoop or Python, ask for clear examples of how they have worked with each of them in the past to solve specific problems and what the processes and outcomes were.

5. See if the data scientist asks the right questions

A good data scientist might want to ask you questions about the data your company works with so they have a better understanding of what they are getting themselves into and whether they can effectively use the data at hand. Is the data centralized or organized? Which kinds of technology and platforms do you have in a place? What kind of budget is there to purchase adequate tools and software? An unskilled data scientist might not think to ask these types of questions.

Following these steps will help guide you towards identifying a true data scientist. But remember that there are some good impostors out there. Before conducing data scientist interviews, either try to further educate yourself on the industry or have a skilled data scientist present in the interviews so they can ask the appropriate questions.

If you are interested in becoming a true data scientist, consider pursuing an online degree in data science.