A Mock Interview Experience from a Data Lover with Imposer Syndrome

I recently finished my data science program at Flatiron and was provided with an opportunity to do a Mock Interview at Skilled for Data Science Technical practice.

I will give a little bit background of myself before analyzing my mock interview experiences further. My undergrad and grad school were not in a data science related field but I had taken one intensive data science course, wanting to build skills in the industry I am into, particularly working on analytics and optimization works. I suffer from imposter syndromes and tend to be extra nervous for exams, interviews, and any forms of assessments (the stress can start adding up a week before any of events). I even get memory-slips of anything or getting very disorganized on answering any questions.

My interviewer had many years of experiences as a data scientist with academic background in machine learning. Learning all those background, certainly, made me felt more stressful about this practice as my inner voice told me “No matter what you say, the interviewer might think you are very stupid.”. It is very bad but I tend to beat myself up already before the interview.

The practice was in part — Conversational (experiences & stats) and Live Coding.

Part 1 Conversational part:

For the first big question, I was first asked “tell me about yourself”. I talked about my actual story and how I found data was relatively under used in the industry which became my motivation to learn more. Some great feedback I want to share include:

1. The story is good but a bit too long
2. This is the chance to make you stand out, therefore, integrating more of your thoughts & passion for data science would be highly valued

Then, the second big question came down to describing a DS project that had challenging. I asked for a minute to think and drew down some points. However, the second I heard the word ‘challenge’, I started to direct my thoughts about what are difficult things that I had to deal with; in which, I started my story just one sentence introduction then jumped right into the details. Some great feedback I want to share include:

1. The project story needs to start with an overall big story (i.e what was the question you wanted to answer)
2. Do not start with getting into the details super fast without giving a good picture of the project first

3. A good structure for project description would be:

The overall picture → Data acquisition and processing → Model selection and decision → Process of training and improvement

The third big question consisted two statistical related questions and which were standard probability and statistical test design.

For probability question, I did calculations and gave an answer without walking through my thoughts and of course, the calculation wasn’t correct which deemed no credits. Then, I went through my process with the interviewer and suggested approaching a probability question — walk through the thought process before you reach an answer; in this way, you can definitely get partial credits or you might be able to correct your answers.

For statistical test design, some good preparation work include:

1. Checking the website of the company
2. Learn about basics of their products

The statistical test design can be heavily focused on their products, and therefore, the basic understanding would be very beneficial.

Part 2 Live Coding:

I was tested on basic Python functions writing. Getting used to the new coding window was a little stressful and of course, in the middle, as I forgot the exact spell of a function, I was asked if I could check the documentation of a particular library. During the time I was coding, the interviewer noticed some problems and gave me quick hints so I could be on the track to 100%. However, I was very focused on completing the function writing and could not fully digest the hints. After this, the interview offered insights about what I can also expect in a DS coding interview:

1. SQL queries
2. Model building and training
3. Basic coding questions to play with the data

He brought to my attention a small but very critical point. Many times, the DS world is very collaborative and therefore, proper communication is important. When he offered the hint, it was a great help and I should have paused and thought more about it before proceeding further (just like how we listen and react to comments from colleagues, supervisors, and clients).

It seems that some small actions of you during a coding assessment can reflect your other parts. Even as a person who value teamwork very much, I, actually, edged out my strengths during this assessment.

In summary, technical interviews are extremely stressful. With experiences in other forms of technical interviews, this is still new and nerve-wrecking for me. I often see it as an exam or assessment where people can possibly find flaws or silly side of me. Afterwards, it can be hard for me to recover. However, why not treat it just as a new experiences to get your try on a problem and share your experiences with experts. It is never possible to predict how the other side reacts to your thoughts but keep it positive to yourself!