Read on Twitter
I have been conducting a lot of phone screenings for Data Scientist roles lately, and I can only speak to my own experience and opinions, but I can give a few tips. (Note, these are not "universal" tips that will work for every interviewer, but things I noted)
I'm coming up with this on-the-fly and it isn't a prepared list, so don't expect these thoughts to be super-organized!
When you're in a phone screening, you usually have a short amount of time to clarify things on your resume and demonstrate your work (and how you communicate your process and challenges and results).
The way I do a phone screening, I'm asking you to explain in more depth something I see on your resume and I'm curious about - either I want to better understand what you meant, or see how you actually talk about what you've done.
Since it's over the phone, you have to explain these things verbally and succinctly, so it's good to practice explaining your resume and work to different types of people who might be calling you:
For example:
-someone very technical who wants to assess your skill level
-someone in management who wants to get a feel for how you explain your work to someone who isn't a data scientist
-a potential teammate who wants to get a sense of what you might be like to work with
-someone very technical who wants to assess your skill level
-someone in management who wants to get a feel for how you explain your work to someone who isn't a data scientist
-a potential teammate who wants to get a sense of what you might be like to work with
If I ask you to tell me about the tools you used for a particular project I see listed under a past role on your resume, "They are listed on my resume" is not an ideal response.
I want to hear you talk about them, how&why you used them, get a sense of your comfort level w/them.
I want to hear you talk about them, how&why you used them, get a sense of your comfort level w/them.
Academic projects are OK to talk about, but make sure you can talk about more than the assignment and the outcome. I want to know that you've come across problems you'll experience on the job, and overcame them. I want to know what role you played if it was a group project.
I want to know how you decided what question to answer, and what approach to take. I want to know if you tried different tools or techniques, or had to learn something new in order to accomplish what you were attempting to do.
Knowing that someone gave you an assignment, you used tools taught in class, & you presented one result, isn't that impressive, even if the result was good.
If that's the only kind of project you've done, consider re-doing it later with different tools and techniques, &comparing.
If that's the only kind of project you've done, consider re-doing it later with different tools and techniques, &comparing.
Or, using the same tools & techniques on a raw dataset you found online, that wasn't given to you by an instructor because they knew it was "clean", and what the outcome should be.
If you're describing a project you've done at work, be prepared to talk about it from different stakeholders' perspectives. Some might want to know your technical approach and model evaluation metrics.
Some might want to know *why* you were working on it, and what business value you provided.
Make sure you can put your work in context, and not just talk about it like a calculator.
Make sure you can put your work in context, and not just talk about it like a calculator.
For example, if you're talking about having built a dashboard in Tableau, talk about what was the purpose of the dashboard, who the target user was, how you figured out what data to use, why you made the data viz design choices you did, how it was used, how you improved it.
This shows that you did your work thoughtfully, and added value beyond "completing a task you were assigned".
It also shows you were engaged - showing some enthusiasm for the work you did and the work you're applying to do is important.
It also shows you were engaged - showing some enthusiasm for the work you did and the work you're applying to do is important.
I'm going to pause for now, but let me know what other questions you have about doing phone screenings/interviews for data science roles!
Note: this takes place after your resume has "made the cut" and the employer wants to find out more about you.
Note: this takes place after your resume has "made the cut" and the employer wants to find out more about you.
Another thought: I'm saying some things in this thread that might sound like they require long explanations. Sometimes the interviewer does want depth, but also be aware of the limited time, especially for a screening call.
It's ok to pause and ask, "Do you want me to continue and explain more?"
That gives them a chance to either cue you to continue, and possibly direct your response to a particular drill-down topic, or move on to another question they want answered.
That gives them a chance to either cue you to continue, and possibly direct your response to a particular drill-down topic, or move on to another question they want answered.
There are certain data science topics you should be prepared to give answers to that show that you've come across a topic.
One example: How did you know your model was good? (with follow-up questions digging into possible model evaluation pitfalls)
One example: How did you know your model was good? (with follow-up questions digging into possible model evaluation pitfalls)
I don't have a list of these, and many lists I see online cover a too-broad range of topics, where I think in a screening call (at least the way I conduct them), the questions asked should be directly related to your resume, or to items in the job description, so narrows it down.
So, maybe look at lists of common data science questions, narrow it down to the ones relevant to what's on your resume or what's in the job posting, and be prepared to answer those.
I feel like this might not be the best advice, though, not having data about what other interviewers ask in a screening. So take this as advice, but don't be surprised if you experience something different.
Read up at least a little bit about the company. If they ask "why would you want to work [for our company or in our industry]" and your answer is that you don't know anything about the company, but a recruiter invited you to apply, that's not ideal.
At least browse their website before the call and get a sense of what they do. And come up with a personal/professional reason why you might want to be involved in that industry/company.
What else haven't I answered that you want to know?
I think a lot of the advice here boils down to being able to put yourself in the shoes of the interviewer. Depending on the company, you could be being screened by someone in HR, someone in management, or a team member.
There can sometimes be a LOT of resumes/calls to get through, and the person might be doing it on top of their normal job. It's unlikely the only thing they have to do that day is sit down with a stack of resumes and call people.
And with the nature of the field of data science (and sometimes-bad job postings), it can be hard to sift through and find someone who is an ideal match for the job.
Your job in the phone screening is to:
-ensure them you can do what's in the job posting
-ensure them you can do what's on your resume
-provide enough "signal" for them to get a sense of your experience during a short call, among many other calls
-give a sense of your potential
-ensure them you can do what's in the job posting
-ensure them you can do what's on your resume
-provide enough "signal" for them to get a sense of your experience during a short call, among many other calls
-give a sense of your potential
I think I gained experience in this area during a past job in sales (yep, I sold Cutco knives!).
In sales, you learn how to do things like:
-address possible objections before they come up
-demonstrate how what you're selling matches their need
-get to the point while also being interesting enough to keep their attention (which involves "reading the room")
-address possible objections before they come up
-demonstrate how what you're selling matches their need
-get to the point while also being interesting enough to keep their attention (which involves "reading the room")
If you have more questions along these lines, let me know!
