Practice the Case Study Method to Nail Your Data Science Interview Practice the Case Study Method to Nail Your Data Science Interview
Landing a job in Data Science is more than just technical knowledge. You’ll need to ace your data science interview and... Practice the Case Study Method to Nail Your Data Science Interview

Landing a job in Data Science is more than just technical knowledge. You’ll need to ace your data science interview and stand out from the crowd. You may feel ready for any technical test or set of questions, but preparing for an interview is another matter.

More than ever, employers are looking for data science experts with the soft skills necessary to communicate complex concepts, problem solve and innovate, and work with a team. Many questions at a data science interview are asking for technical knowledge on the surface and something else entirely underneath.

Get ready for your data science interview with a reliable method for answering questions to showcase your expertise and your soft skills — The Case Study Method.

Why practice the Case Study Method

Data scientists must be great storytellers. The massive size of data makes gleaning insights and then presenting them in an understandable way a package deal. You cannot be a successful data scientist if you can’t do both.

It’s more than just drawing up a cool visualization. Understanding the real questions involved in searching for answers within data takes a level of social skill beyond simple analytics or mastery of Tableau.

You’ll also need excellent communication skills because data science is bleeding into other departments. You’ll likely be working with sales, marketing, finance, and legal, among others, to put data insights into motion. Hiding at your desk will not be an option when businesses know that data is now gold, and everyone in the organization is seeking data literacy.

Data scientists must also be great problem solvers. Data rarely moves in business like it does in the squeaky clean world of academia. In business, you’re massaging data, getting creative, and understanding what to keep and what to leave behind in the fast world of business.

The most prominent example of this happened at Google a few years ago. Google’s founders set hiring algorithms to filter based on technical expertise, owing to a belief that only technologists could understand technology.

A few years later, a real look at the data revealed that Google’s most successful hires had a variety of soft skills, including communication and empathy. STEM skill was dead last.

Employers wanting to replicate Google’s expanded hiring practices based on that study are using data science interview methods designed to uncover these soft skills (or “power skills,” as Forbes suggests).

How the Case Study Method works

When an employer uses the case study method, they present you with a scenario. It’s your responsibility to provide an answer to this scenario by using your experience and then explain your answer using soft skills.

The case study could happen in a short time frame, or it may be something you complete over the course of a few days. In both, candidates who’ve proven their technical skills have a chance to showcase how their performance fits the team ideals the company has in mind.

The case study will be a real problem you’ll solve using your wits and your expertise. It will revolve around the messy world of data. There’s no exact or right solution to this problem. Instead, the interviewer is looking for insight into how you think and work.

Case study example

The interviewer gives you a scenario:

Last month, there was a spike in sales for a specific product in the company offering. The company would like to uncover the possible reasons for that unexplained spike to see if conditions could be replicated to drive revenue through quarter-end. What could have caused that spike, and how do you confirm?

To do well on this task, you’ll need to produce:

  • Good analysis — The results must be free of errors or omissions, and it must be meaningful for the business and its target question.
  • Good communication — The employer wants to know that not only can you secure good analysis but that you’ll also be able to explain what you’ve done to stakeholders, technical or not. This involves a compelling narrative and answering questions satisfactorily on the spot.
  • Professionalism — Candidates may have to answer the same question more than once as nontechnical people grapple with the process. Difficult discussions could happen that lead to going a different direction with the data than what the data scientist would prefer. You must handle these professionally.

What most candidates do wrong with a case study

There are a few common problems with a candidate’s presentation during a case study:

No narrative — Just presenting findings isn’t what employers want. Instead, they’re looking for the ability to tell a story. 

Solution: Present an insight you found and what that could mean specifically for the business

Too technical: The presentation is far too technical for stakeholders to understand, including unnecessary information about the process itself.

Solution: Stick to the business side of the presentation and leave the technical discussion with the data science team itself. Get to the meat of the insight.

Not professional: Candidates that make it to the case study are technically proficient, but not all can handle the pressures of presenting findings. 

Solution: Don’t take questions about the methodology personally, nor inquiries that repeat. This is the time for patience for the nontechnical side of the business.

Practicing the case study data science interview method

The best way to get used to this type of data science interview is to practice. Research scenarios or fine cases on your own and run your numbers. Take it to the next level by visualizing the data and enlisting the help of a nontechnical peer (or family member) willing to sit down and let you explain.

Remember to prioritize the story above all. You should be comfortable talking about metrics, experimentation, and modeling, as well as the overarching insight as it relates to business value. Once you can blend these two things, you’ll be on your way to acing the case study interview.

Editor’s note: Ready to learn more in-demand skills, including data science interview prep, machine learning frameworks, and more? Subscribe to Ai+ Training and get access to hundreds of talks on-demand, live training sessions, and discounts to ODSC events!

Elizabeth Wallace, ODSC

Elizabeth is a Nashville-based freelance writer with a soft spot for startups. She spent 13 years teaching language in higher ed and now helps startups and other organizations explain - clearly - what it is they do. Connect with her on LinkedIn here: https://www.linkedin.com/in/elizabethawallace/