Also to add- psychometric tests are such a bad tell of a persons ability to critically think. I understand it’s a proxy measure and a way to filter out candidates but I would prefer a case study over those annoying tests !
Re: take home assignments, I just hired two new Product Analysts using a “live” 45 minute case study. Ability to test basic technical proficiency, communication, and critical thinking all at once
One thing I would challenge is the "Say no to take-home assignments" point. Data science portfolios are easy to copy-paste from other sources. Profiles from others are easily duplicated, especially in the age of LLMs. I've recently begun hiring and there is A LOT of spam in the hiring process today that wasn't present even just a few years ago.
I've found properly tuned take-home assessments and the subsequent discussion of them in an interview are a reliable way to assess technical, critical thinking, and communication skills. Yes, it places an additional burden on the candidate's time but I wouldn't call that disrespectful. Some people shine when they have the time to think and analyze in a focused environment.
What I would do is share interview questions with candidates ahead of time, such as a root cause analysis or a deep dive into a case study - like why a particular metric declines or accelerates. During the interview, I would offer them input details. This way, you can evaluate their framework, logic, and examples based on their response, and you also have the ability to dive into particular aspects, such as tooling, methodology, or context, as needed.
Agree about significantly higher volume of spam / irrelevant / not qualified leads disrupting now the interview process (which indeed was not the case before 2022). And, yes, portfolios are easy to copy-paste, but good-targeted portfolios also help to spot technically qualified candidates for the initial screen / next step of interview.
Additionally, receiving a rejection hurts less when it's because your portfolio doesn't fit the role, rather than after you've spent a week with multiple late nights on a particular assignment that you'll never see again.
Also to add- psychometric tests are such a bad tell of a persons ability to critically think. I understand it’s a proxy measure and a way to filter out candidates but I would prefer a case study over those annoying tests !
Ohh, psychometric tests is another painful story, haha. Absolutely.
Re: take home assignments, I just hired two new Product Analysts using a “live” 45 minute case study. Ability to test basic technical proficiency, communication, and critical thinking all at once
Great list of resources, thanks!
One thing I would challenge is the "Say no to take-home assignments" point. Data science portfolios are easy to copy-paste from other sources. Profiles from others are easily duplicated, especially in the age of LLMs. I've recently begun hiring and there is A LOT of spam in the hiring process today that wasn't present even just a few years ago.
I've found properly tuned take-home assessments and the subsequent discussion of them in an interview are a reliable way to assess technical, critical thinking, and communication skills. Yes, it places an additional burden on the candidate's time but I wouldn't call that disrespectful. Some people shine when they have the time to think and analyze in a focused environment.
What I would do is share interview questions with candidates ahead of time, such as a root cause analysis or a deep dive into a case study - like why a particular metric declines or accelerates. During the interview, I would offer them input details. This way, you can evaluate their framework, logic, and examples based on their response, and you also have the ability to dive into particular aspects, such as tooling, methodology, or context, as needed.
Agree about significantly higher volume of spam / irrelevant / not qualified leads disrupting now the interview process (which indeed was not the case before 2022). And, yes, portfolios are easy to copy-paste, but good-targeted portfolios also help to spot technically qualified candidates for the initial screen / next step of interview.
Additionally, receiving a rejection hurts less when it's because your portfolio doesn't fit the role, rather than after you've spent a week with multiple late nights on a particular assignment that you'll never see again.
Great guide for anyone preparing for ds/ analytics interviews. Thanks!