Stories
Slash Boxes
Comments

SoylentNews is people

posted by martyb on Friday November 02 2018, @12:36AM   Printer-friendly
from the pi(x)=substr("3.141592653589793238462643383279502884197169399375105820974944592307816406286208",1,x) dept.

Recently there was an article discussing how poorly today's Silicon Valley approaches the question of testing their information technology candidates' intelligence.

If you were a hiring manager, how would YOU test YOUR candidate's intelligence?

I was mulling this over recently, when, for unrelated reasons, I found myself researching algorithms to be used in calculating 'pi' (en.wikipedia.org/wiki/Category:Pi_algorithms).

As it so happens, there are currently 15 known algorithms for pi. At least one of the algorithms can be used to generate arbitrary digits of pi (en.wikipedia.org/wiki/Bailey–Borwein–Plouffe_formula), and that might be relevant. (I'm not a mathematician, or a programmer, as such - I'm a sysadmin - but even those who are will likely be surprised to learn that such a thing is possible. For more information, please see en.wikipedia.org/wiki/Spigot_algorithm.)

And so I would ask my candidate to pretend that he was responsible for designing a library of math functions, and to offer me an opinion on which algorithm should be implemented, in our hypothetical library of routines, to calculate 'pi'.

Some prior experience in programming is required - you have to have written your own functions. No programming languages are required. No coding. Not even pseudo-code! No right answers. No wrong answers. Just pure thought.

You don't need to be a programmer to take this test and succeed. You don't need to do anything on a whiteboard. You just need to be somewhat mathematically inclined ... somewhat literate ... and, a nerd.

A real nerd. Not one of these fake Silicon Valley nerds. You need to have books on your shelves. Not DVDs.

Points for asking what the library will be used for. The value 22/7 might work great for roughing out the roof of a gazebo. Not so good for calculating orbits!

Points for implementing multiple algorithms and letting the user decide for themselves.

No time limit, no pressure ... but I would want to hear back from my candidates, within a day or two, via email.

Compare my test to the puerile tests involving balls, and strings, and calculating 2^64 in your head, in real time, and ask yourself which of these methods REALLY exposes intelligence?

Now, you're in charge.

What would you do?


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 5, Insightful) by MrGuy on Friday November 02 2018, @02:43AM

    by MrGuy (1007) on Friday November 02 2018, @02:43AM (#756731)

    A former client of mine would have a conversation with someone. Not just an open conversation, but a focused conversation on a topic. They'd put two interviewers together with the candidate, and suggest 4-5 interesting topics to discuss (e.g. "Should we always make sure that we've found the very best possible idea before we implement anything?") The candidate should think about the problem, and express an opinion. Then there's a conversation about the topic - why do you feel that way? What principles or overarching ideas guide you to that opinion? Have you considered this problem or alternative? (The interviewers deliberately offer opposing or possibly contrary viewpoints at least some time). Here's a hypothetical situation - is this what you think should happen given your position? Hmm...maybe it needs to be more nuanced than that - maybe I should revise it in some cases, but in general I still believe X to be the superior position.

    The point of the exercise isn't really to probe the specific opinion expressed. It's to a.) see how someone approaches a topic where there's not an obvious right answer, b.) see how they go about logically presenting an opinion c.) see whether they can express an opinion as an outgrowth of larger principles and ideas in a coherent fashion, d.) listen respectfully to contrary positions, e.) think critically about their own position, and possibly be willing to revise it, and f.) just generally behave like someone who's intelligent, a good listener, reasonable, but willing to express and defend an opinion.

    If you can do all these things, you're probably going to make a good team member. If you can't, you're not going to speak up when you should, or you're going to be constantly expressing strong opinions and throwing a fit when they're questioned, or you're going to have a slapdash "It sounded good at the time and now it's too late to change my mind!" approach, or you're going to be getting into pointless arguments all the time, or you're going to be a jerk to your teammates.

    Starting Score:    1  point
    Moderation   +3  
       Insightful=2, Interesting=1, Total=3
    Extra 'Insightful' Modifier   0  
    Karma-Bonus Modifier   +1  

    Total Score:   5