This seems to be a way for them to screen for the 1000s of entry level applicants, who don't have large supporting work histories. They need to screen them somehow. One way is "Only from this list of schools." Another is "Based on what some person thinks of your resume writing ability as a new engineering grad." Another is "Based on what our algorithm thinks of your ability to place keywords in your resume." I don't think this is that much worse an alternative for those who may not have other means to get into a prestigious company.

When I was last in school I struggled with "networking your way into a company" and wound up landing my internship (and ultimately full employment) on the one company who offered a test before the "Fly you out to meet us" step.

They will probably fly you out after this, for another stage of the interview. But at the "FizzBuzz stage", they do really need to make sure that the person answering the question is the same person they're later flying out. There are a surprisingly large number of college students who've gotten used to "paying for grades", who attempt things like that.

Maybe they should just conduct fizzbuzz then?

I think the point is that even if employers do conduct fizzbuzz, some candidates use someone else's answer.

Some do and you will see this at on site. So how big is the harm really?

Wasting thousands in travel and interview costs?

those are the people you should hire. why would you pay to reinvent the wheel when you really need someone to invent better IAM controls?

Because the point of FizzBuzz is not to get a working FizzBuzz class that can be integrated into an application, it's to see if the candidate can actually produce a working version themselves.

Imagine you're hiring someone to do your taxes. Sure, you'd expect them to use a calculator to do all the hard addition/subtraction, or more likely spreadsheets. But if you asked them to add 5 and 3, and they said "I would use a calculator, this is a stupid question, I can't just tell you off the top of my head" then you're probably not going to hire them.

It's a super basic test of 'Can this person code themselves out a paper bag."

Thank you! For a few hours there I couldn't believe I was the only one in the thread that understood this.

which can be determined entirely without the aforementioned methodology. if you read the guys article, you'd see that he'd already completed the basic tests demonstrating basic competency.

being able to solve problems is more important than how you solve problems. (generally speaking, yes i know there are exceptions)

Because people who are both incompetent and deceitful are an utter waste of everybody's time, and also a liability?

we both know those people make it past fizbuzz anyways, and they are left in the dust fairly quickly anyways.

And that's why Amazon's phone-interview coding exercise is a rather complicated coding question delivered under proctoring, rather than simply fizzbuzz itself: to actually filter those people out early and avoid the cost of flying them to Seattle, rather than just "going through the motions" of doing so.

We both know these are the people you were advocating hiring in your previous post.

I interviewed with them a decade or so back. First interview was a phone screen. Ok makes sense. Second interview was another phone screen. Never mind the fact I lived about ten miles away. On there other hand two years ago I interviewed with a company on the other side of the country. After a half dozen phone interviews they decided to fly me out... For more interviews

I agree with your reasoning, i think the alternative perspective is that traveling somewhere to interview is a significant time commitment on my part, and I might welcome alternative testing approaches where I don't have to fly somewhere until I'm reasonably confident I'm going to get an offer.

the internship IS the interview.