
Nationwide Data Scientist Interview Questions
& Process
Real candidates share what happened, how many rounds they had,
and how the experience turned out.
Based on 20 interview experiences · FREE TO READ
Candidate interview experiences
First-hand accounts from people who interviewed at Nationwide.
Data Scientist
It took them forever, like 2 months, to decide. First, there was a modeling challenge due by the end of the week. You had to build a model and write a detailed report. Then, there was a panel interview (3-5 people) to go over the model, kind of like an in-person meeting with a manager. You'd think they'd read your report and understand your thinking before passing you, but I felt like they just glanced at the results. Since my results were good, they passed me initially. I even made a nice presentation explaining why my model was correct, but they asked stuff that was already in the report. It was obvious they hadn't read it. So, for a model review, just go with the submission because they probably won't check it. The data they gave was simulated, so I found a great algorithm that worked really well, which might have ruffled some feathers. They seemed to expect someone to struggle with the data, do tons of feature engineering, and use complex algorithms. When I solved their simulated data with a simple, mathematically sound algorithm, they thought I got lucky. I think I just analyzed the data, did an exploratory analysis, and found the best algorithm for the model (roc 0.96 classification with low FP and low FN on a skewed dataset). If they wanted me to act like it was real-time data, they should have provided real data. After that, they stopped talking to me and sent a rejection email almost a month later via automation.
- Standard machine learning and data science questions will be interspersed as you discuss the model.
- Questions about multicollinearity, random forest, and logistic regression were some that I recall.
Data Scientist
So, there were two parts to the interview. First, I had a Zoom call where they asked me some basic personal questions. After that, they sent me a small data analysis project to work on, which took me about 4 days. Then, a few days after I submitted the project, we had another Zoom interview. They asked me about how I tackled certain issues in the project, like dealing with unbalanced data and what data leakage means. A few days after that, they inquired about my PhD completion timeline. Ultimately, I didn't receive any offers.
- Regarding the data analysis project, how did you approach the issue of unbalanced data?
- Could you explain the concept of data leakage in the context of data analysis?
- Can you provide an estimated completion date for your PhD?
Data Scientist
The interview was pretty straightforward. It mainly consisted of behavioral questions and questions about the technical skills mentioned on my resume, focusing on how I've learned and used them in previous roles.
- How did you get ready for this interview?
- Tell me about your journey in learning and applying the technical skills listed on your resume.
- Can you share an example of how you've utilized [technical skill from resume] in your work history?
Nationwide Data Scientist Interview Questions
Quoted word for word from Nationwide interview reports.
“How can we tell if a group doesn't have enough data to be included in our analysis?”
Read reports →“What are some basic Git terms?”
Read reports →“Could you explain the concept of variance in a way that a child would understand?”
Read reports →“Can you provide an estimated completion date for your PhD?”
Read reports →“Which problems are inherently serial in nature?”
Read report →“Can you explain logistic regressions?”
Read report →“Could you explain the concept of data leakage in the context of data analysis?”
Read report →“How would you describe variation in elementary school students?”
Read report →“How would you go about processing a data table with 50 million rows?”
Read report →Formats, difficulty and experience
Across all 20 Nationwide interview reports.