The Receptionist Test (The Interview before the Interview)
thedailywtf.com
thedailywtf.com
Is it that there are lots of people who lie and say they're computer people when they aren't (compared to the rates in other occupations)? Is it because you can't fake computers like you can fake other jobs? Is it because computer people are used to being mistreated and undervalued in a way that is unacceptable for other people? I'm really curious if anyone has thoughts on this.
However, there are lots of other fields that demand a certain level of competence for you to pursue them. Interviews at investment banks last all day, and they ask for pretty serious mathematics exams among other things. Lawyers must pass the BAR, which is fairly brutal. Actuaries must pass the actuarial exam. Musicians and dancers must audition. Aspiring professors must give a series of 'job talks', opening up their research and program to the criticisms of an entire faculty, after which they continue to meet one on one for hours at a time.
By contrast, the hours spent in Google interviews seem comparatively simple.
The simple truth is, some fields are performance-oriented. Either he is the best Tuba player auditioning for the chair or he goes home with nothing for his efforts.
They don't. If you can reverse a string in place in C, or reverse the words in a string, you can get hired at 90% of programming jobs.
It boggles my mind when developers seriously believe lawyers, graphic designers, chefs, etc. don't require a high "standard of competence" in their own skill sets, just because they aren't e.g. messing with pointers or refactoring.
I have interviewed over 2500 programmer applicants, all having passed a "resume screening". I terminated 50% of the interviews before reaching the coding test because I knew by then that it would be a waste of time. Of the other 50%, 90% couldn't pass the simple coding test. I ended up hiring or referring less than 100.
Anyone else have similar experience with these percentages?
It's pretty amazing how easy it is to get a degree in compsci or game development and have no programming competence whatsoever. What amazed me too was the astounding quantity of terrible CVs, even among people who supposedly have a substantial amount of experience. I used to laugh at articles telling you to spell check your application and give other, seemingly extremely basic advice, but I swear about 80-90% of applications have glaring spelling, grammar and semantic mistakes and no discernible logical structure. Come on, how hard is it to assemble a chronological list of things you've done?
"All your base are belong to us!"
Perhaps if you had a different way of pre-screening candidates, something besides resumes.
Perhaps even if the pre-screening itself took much more time then scanning a resume - overall it might end up saving time?
Judging by myself: FSCKING YES!!!
The only thing I hate more than resume writing is Java,
This freaking "resume screening" selects best resume writers, not best programmers.
Its basically just a bunch of buzz words on the skills you have, and a paragraph about your last few positions/roles.
As to why computer jobs require competence, it is because any engineering position (computer, or otherwise) requires competence. Would you live in a building without knowing that it won't collapse tomorrow because the construction company didn't know how to calculate the required strength of the steel beams needed to support the structure?
A company won't go under because a secretary does not know how to make perfectly formatted word files from the get-go. Where as it will fail if a software company's key product is managed/coded by incompetence.
Also, I'm glad they didn't call this the "administrative assistant test".
My cofounder now says that he wish there were an earthquake for every interview...
The biggest test is that of go-getter and not be finicky about the kind of work you do.
Seriously though, Ford's test sounds silly. I'm sure he had some kind of justification for it, but I doubt it really did a good job of screening out who would and wouldn't be a good employee.
Do you want to hire someone that does things 'because that's what they always do', or because they collected data and acted appropriately?
I'm not saying salt is a good indicator - there are a million other ways you could test it - but I think that's the gist of it.
It's a weed-out class.
What's wrong with presenting the problem in a straightforward manner and see if the applicant can debug the steps, telling the interviewer what he's trying and why at each step? For starters you'd get way more out of his think flow. For "authenticity" you can add that this was a real problem that Cindy faced the other day and you may ask her questions.
People tend to try what they know first (ie: pounding ctrl-p or ubuntu), and if it happened to be the issue, they were merely lucky and may not be competent at all either.
How hard is it to ask them "What else haven't you tried? What would you have tried next if this didn't fix it? Why did you try those steps in that order?"