Analytical interview questions

[captZEEbo]

What would be an acceptable "thought process" for any of the original questions?

For some of the questions are they looking for estimates, or are they looking for answers with variables?

[bisonbison]

BPS,

Let's take "How many ping-pong balls fit in a 747?" - a question I've never asked or gotten in an interview.

Good answer:
Are we filling just the cabin? Are the seats and overhead compartments in place? Are the balls in boxes or loose? Yes, no, loose? Ok, I think a 747 has about 9 seats per row + 2 aisles. Let's assume the seats and aisles are each 2 feet wide. That's 22 feet. If there are 40 rows, each 3 feet deep, that gives us 120 feet long. I'm six feet tall, and there's about a foot of clearance for me, so the cabin is 7 feet tall, it's a rounded cylinder but it's not a semicircle, so I'm just going to treat it as a rectangle for now. 22 * 120 = 2640, 2640 * 7 = 18480 cubic feet. Let's say that the curvature of the plane cuts out 1/4 of that space. A quarter of 18480 is 4000 + a quarter of 2480, so that's 4620, meaning our space is 13860 cubic feet.

I think a ping pong ball is about 1 cubic inch, and there are 12x12x12 cubic inches in a cubic foot, so we get 12x12x12x13860 = 23 million ping pong balls.


Bad answer: I think it's about 120 feet long and like 20 feet wide. I guess with the height and curved walls it'd be average 5 foot height. A ping pong ball is about an inch and so we'd get 2400x5x12x12x12 in there, so 20 million ping pong balls.

Conclusion: Basically the same answer, but the good answer walks the interviewer through all the work and assumptions, lays out the thinking and allows you to go back and fiddle with certain things. "Oh, the seats are in the plane?" Well, that's x inches deep by y inches wide by z inches tall, and there are m seats, so we'd lose n balls."

[bisonbison]

Why are manhole covers circles instead of rectangles?

They don't have to be aligned to fit, easier to move, less material for same volume in the covered cylinder.

[Vyse]

These are horribly stupid questions that tell you nothing you can't figure out in a normal interview.

[econophile]

bison,

i would think the main reason some people give "good answers" is that they have read some case interview book that tells them they are supposed to answer the questions this way.

[Big Poppa Smurf]

[ QUOTE ]
BPS,

Let's take "How many ping-pong balls fit in a 747?" - a question I've never asked or gotten in an interview.

Good answer:
Are we filling just the cabin? Are the seats and overhead compartments in place? Are the balls in boxes or loose? Yes, no, loose? Ok, I think a 747 has about 9 seats per row + 2 aisles. Let's assume the seats and aisles are each 2 feet wide. That's 22 feet. If there are 40 rows, each 3 feet deep, that gives us 120 feet long. I'm six feet tall, and there's about a foot of clearance for me, so the cabin is 7 feet tall, it's a rounded cylinder but it's not a semicircle, so I'm just going to treat it as a rectangle for now. 22 * 120 = 2640, 2640 * 7 = 18480 cubic feet. Let's say that the curvature of the plane cuts out 1/4 of that space. A quarter of 18480 is 4000 + a quarter of 2480, so that's 4620, meaning our space is 13860 cubic feet.

I think a ping pong ball is about 1 cubic inch, and there are 12x12x12 cubic inches in a cubic foot, so we get 12x12x12x13860 = 23 million ping pong balls.


Bad answer: I think it's about 120 feet long and like 20 feet wide. I guess with the height and curved walls it'd be average 5 foot height. A ping pong ball is about an inch and so we'd get 2400x5x12x12x12 in there, so 20 million ping pong balls.

Conclusion: Basically the same answer, but the good answer walks the interviewer through all the work and assumptions, lays out the thinking and allows you to go back and fiddle with certain things. "Oh, the seats are in the plane?" Well, that's x inches deep by y inches wide by z inches tall, and there are m seats, so we'd lose n balls."

[/ QUOTE ]

So basically the difference is in the level of detail in the assumptions made?

edit: Sorry if these are dumb questions, I feel retarded that I don't get the point of these, but it's really making my head hurt to figure out how these are truly useful interview questions.

[Blarg]

[ QUOTE ]

How would you locate a book in a big library if there is no cataloging system and no librarian to help you?

[/ QUOTE ]

I'm not sure I'd want to work at a place that asked me this question. It's sooooo formless and requires considerable assumptions to do better than exhaustive, sequential searching.

[/ QUOTE ]

I'd ask someone. Several people, if I had to. Why reinvent the wheel if somebody already knows? Waste of time.

[Blarg]

You purposely left out the main one I guess.

[Blarg]

[ QUOTE ]
These are horribly stupid questions that tell you nothing you can't figure out in a normal interview.

[/ QUOTE ]

I agree and think these are more for the interviewers to get their jollies while feeling smart than anything.

Laying out one assumption after another, which is what most of these questions require, is merely mechanical. A reasonably attentive monkey can do that.

Often IQ tests at least require some creative thinking somewhere, not just plugging numbers into a formula.

[RunDownHouse]

[ QUOTE ]
Why are manhole covers circles instead of rectangles?

They don't have to be aligned to fit, easier to move, less material for same volume in the covered cylinder.

[/ QUOTE ]
And so they don't fall through the hole they're covering.