Cracking Statistics Interviews for Data Scientists

A statistics interview is a technical interview that evaluates necessary skills for the daily work of a data scientist. You have to be able to ace this interview if you want to land a data science job.

However, statistics is a broad subject. Thankfully, you do not have to know all of it to be a data scientist. This blog will break down what you do need to know. After all, knowing what to expect is half the battle, so let’s dive in.

What Do You Need to Know?

For data scientists, a statistics interview typically focuses on 3 areas of knowledge:

• Probability

  Probability questions cover 3 major categories: probability basics (expectation, variance, permutation, combinations, etc.), conditional probability (Bayes’ rule), and probability distribution (commonly used discrete and continuous distributions such as binomial, normal, and long-tailed distribution). Probability questions will likely appear the most in interviews.

• Hypothesis Testing

  For these questions, you need to know both terminology (power, p-value, confidence interval, etc.) and testing methods including parametric (Ex: z-test and t-test) and non-parametric (Ex: chi-squared) tests.

• Regression

  You need to be familiar with linear and multiple regression. Regression will likely appear the least in interviews, but you still want to be prepared.

Types of Questions

Those are the areas of knowledge, but how will you be asked about those areas? There are 3 types of questions you can expect:

Conceptual Questions

These questions are not interested in your math skills but rather your ability to explain concepts. You will be asked to explain concepts to both a technical and non-technical audience.

Example Questions:

• Probability: What is the distribution of average time spent per user?
• Hypothesis Testing: Explain the p-value and confidence interval to a non-technical audience.
• Regression: What are the assumptions of linear regression?

How you approach these questions depends on the target audience. There are different approaches for someone with a technical background versus someone with a non-technical background.

For a Technical Audience:

A technical audience may be familiar with the terminology, but it’s important to stay organized so that your explanation is clear. Following these steps can help you stay organized.

1. Start with some context.
2. Define the concept. Try to keep it simple because this shows a higher level of understanding.
3. Explain what changes in values mean (for concepts that can be represented by numbers).
4. Talk about how this concept is applied in practice (optional).

For a Non-Technical Audience:

You can use the same steps to organize your explanation that you used for the technical audience, but there are some additional things you should keep in mind.

• Use examples and analogies.
• Avoid using technical terms.

For all audiences just remember to keep your answers clear and structured. You always want your answer to be as easy as possible to understand because that shows that you truly understand the concepts.

Questions Involving Calculations

Explaining the concepts is not enough to get the job. You also have to be able to do the math, so expect questions that require you to do calculations.

These questions could either ask you simply how to solve a problem or have you go further and provide the exact answer.

Example Questions:

• What’s the probability of getting two heads among 10 tosses of a fair coin?
• Given two groups of users, compare the click-through rates and draw conclusions as to whether the two click-through rates are the same.
• Can you lay out the testing steps and draw conclusions?

These questions are all about whether you have the skills to go along with the knowledge. The final type of question also evaluates your skills.

Coding Questions

These questions are all about your implementation skills. You need to show that you not only understand what to do but that you can do it. These questions give you a chance to show that you can get results.

The best way to prepare for this is to practice coding. The 10 days of statistics by HackerRank is a great place to start.

Final Thoughts

Remember that if you want to ace interviews you have to be prepared. Now that you know what to expect in statistics interviews, you can study and get ready to crack your next interview.

If you enjoyed this post and want to see a longer version, check it out here.