What do you do when an interviewer asks a question you do not know?

[RoGeorge]

The "how many cows" type of question is OK.

For an engineering job, the expected answer will be to show how you estimate something when very little information is known.
The result can be a number and an error range, the error range can be a few orders of magnitude big.

It's not about numbers, but how you get to those numbers.

[brucehoult]

When I was in my interview for my current company, the manager engineer asked me: "How many cows are there in Canada?"

So... how would all of you answer that?

It's an excellent question. Estimation.

For example:

How many people are in Canada? More than Australia I think .. 30 million?
How much milk does each person use? Hmm .. maybe a litre a week. 50 litres a year.
How much milk does a cow produce? Let's say 10000 litres a year.

3e7 * 5e1 / 1e4 = 35e4

350,000 cows to produce milk. Neglecting imports and exports.

You also need some more to produce butter, cheese etc. And meat. Let's triple it and say one million even.

You're not going to get an exact answer this way, but I bet it's the right order of magnitude.

[glarsson]

My estimation was 5M based on comparison with Sweden. Defining cows as beef cows and milk cows (ignoring oxen, calfs, etc) the official numbet seems to be just below 5M for Canada.

[brucehoult]

When I was in my interview for my current company, the manager engineer asked me: "How many cows are there in Canada?"

So... how would all of you answer that?

I'd need more information around the context of the question to give a more effective answer.

Furthermore, smart arse trick questions like these which put the interviewee under unrealistic, unnecessary and undue pressure simply indicate to me they're not a particularly nice place to work in, and have too much time on their hands not doing real work. Can you imagine how intrusive and overbearing their HR dept is likely to be in your daily working life? No thanks!

I totally disagree!!

Nothing trick here. The ability to quickly do an order or magnitude estimation is vital to any good engineer or programmer to know if something is even worth trying, to notice someone is bullshitting you etc.

You're not expected to get the right answer to 3 decimal places! Getting the correct exponent is probably enough, and within a factor of two is perfection.

My favourite to ask a candidate who claims knowledge of high school physics, but I've never seen anywhere as a question: What is the mass of the Earth's atmosphere?

You only need to know two physical constants and one math formula (maybe two related ones) to get this right to within 10% or so. I'll even prompt you with their values if you at least figure out which ones you need but can't remember them.

[Mr. Scram]

It's an excellent question. Estimation.

For example:

How many people are in Canada? More than Australia I think .. 30 million?
How much milk does each person use? Hmm .. maybe a litre a week. 50 litres a year.
How much milk does a cow produce? Let's say 10000 litres a year.

3e7 * 5e1 / 1e4 = 35e4

350,000 cows to produce milk. Neglecting imports and exports.

You also need some more to produce butter, cheese etc. And meat. Let's triple it and say one million even.

You're not going to get an exact answer this way, but I bet it's the right order of magnitude.

Import and export are important, as some countries vastly overproduce and others do the opposite.

It's not whether your answer is correct though, it's how you approach the question. Even a totally wrong result can be right.

[tggzzz]

When I was in my interview for my current company, the manager engineer asked me: "How many cows are there in Canada?"

So... how would all of you answer that?

It is a reasonable question for the reasons others have given. But it is the type of question an HR-droid or technically non-competent manager might ask.

For that reason I would prefer a more engineering specific question such as "A manufacturer of toy cars for children approaches us and asks us to design traffic lights controllers, so that the model roads are more realistic. How would you proceed?".

I would want to see a the client ask questions to elicit a better definition of the acceptable solutions, plus several possible technical implementations (most of which will be quickly rejected), plus an appreciation of how to structure a project plan.

[NivagSwerdna]

According to google there are around 950k Milk cows in Canada so 350k is well within an order of magnitude.  There is roughly 4x the number of beef cows... the grand total including bulls, heifers, steers etc... just under 12M

But of course that is irrelevant... the question is asking you how you would go about reasoning the answer not the answer itself.

The OP's question is a simple trap... when asked a question that is out of range some people just b*llsh*t; nobody wants to work with one of them... so they get rejected.  The way to tackle it is say what you know, stop there and then offer reasoning to go further but to be upfront that this part of the answer comes with a degree of uncertainty. IMHO saying you would use google isn't a bad answer if you admit something is beyond your knowledge.

Slightly off topic... once I was interviewing and the candidate went out of their way to explain how they worked night and day, including weekends to solve a problem... I rejected them... if they can't get the work done in the time allocated then... harsh but fair. 

[Howardlong]

When I was in my interview for my current company, the manager engineer asked me: "How many cows are there in Canada?"

So... how would all of you answer that?

I'd need more information around the context of the question to give a more effective answer.

Furthermore, smart arse trick questions like these which put the interviewee under unrealistic, unnecessary and undue pressure simply indicate to me they're not a particularly nice place to work in, and have too much time on their hands not doing real work. Can you imagine how intrusive and overbearing their HR dept is likely to be in your daily working life? No thanks!

I totally disagree!!

Nothing trick here. The ability to quickly do an order or magnitude estimation is vital to any good engineer or programmer to know if something is even worth trying, to notice someone is bullshitting you etc.

You're not expected to get the right answer to 3 decimal places! Getting the correct exponent is probably enough, and within a factor of two is perfection.

My favourite to ask a candidate who claims knowledge of high school physics, but I've never seen anywhere as a question: What is the mass of the Earth's atmosphere?

You only need to know two physical constants and one math formula (maybe two related ones) to get this right to within 10% or so. I'll even prompt you with their values if you at least figure out which ones you need but can't remember them.

The trick is that they are deliberately trying to make the interviewee uncomfortable with an open ended question with apparently little relevance to the post they're interviewing for. It shows a lack of respect, and suggests (to me at least!) an underlying culture of bullying, cliquism, siloing and trying to catch people out, rather than an atmosphere of fairness, openness, co-operation and transparency. It just doesn't sound like a nice place to be IMHO.

On the other hand, asking a technical question as you suggest leaves it far more clear to the interviewee what you're after, without any unnecessary patronising and divisive smart arsery.

[RoGeorge]

The trick is that they are deliberately trying to make the interviewee uncomfortable ... It shows a lack of respect

Yes, the interviewee is deliberately put in an uncomfortable and stressful situation, but the interviewer is NOT doing that because of a lack of respect. It is doing that because the interviewer wants a glimpse of how the interviewee deals with unknowns, or with stressful situation.

Even the most smart and decent person can turn into a stupid jerk during a crisis situation, and that could be very bad for business. Nobody likes an employee that might be triggered by just an unpleasant question.

[VEGETA]

Their purpose of that question was to measure how can I handle odd questions in front of experienced engineers and managers, and how strong is my personality. Also, how can I deliver good enough answer whether I know the true answer or not...

I answered as follows: "Canada is not a country famous for cows or milk product, so there is no available specific information about it. However, a country like Switzerland is famous for cows and milk products, thus I know they have 600000 cows and all of them registered and maintained.

So for Canada, if you want an answer then I would need to make a specific research to be able to give you results according to available resources."

Then they followed this with tough technical and personality questions and I answered them perfectly. I got the job because of my personality and talking skills since I did not get the 1st place for exams (I got 2nd).

[RoGeorge]

Of course, there is always a chance to pass the theory, but step into troubles at the practice test:



 

[Howardlong]

The trick is that they are deliberately trying to make the interviewee uncomfortable ... It shows a lack of respect

Yes, the interviewee is deliberately put in an uncomfortable and stressful situation, but the interviewer is NOT doing that because of a lack of respect. It is doing that because the interviewer wants a glimpse of how the interviewee deals with unknowns, or with stressful situation.

Even the most smart and decent person can turn into a stupid jerk during a crisis situation, and that could be very bad for business. Nobody likes an employee that might be triggered by just an unpleasant question.

I understand exactly what they think they are doing, but there are other far more representative ways of doing it. While of course I'd give it a reasonable shot, I'd walk out of there and think "what a jerk", and think twice about what I'd be letting myself in for for the reasons I already cited.

I wonder what they'd think of someone who stated "11,750,000, as at January 1 2017, a 27 year low"?

[RoGeorge]

One more thing to add, about what to expect after getting the job. No matter what, sooner or later something like this will happen to any of us:

[tggzzz]

The trick is that they are deliberately trying to make the interviewee uncomfortable ... It shows a lack of respect

Yes, the interviewee is deliberately put in an uncomfortable and stressful situation, but the interviewer is NOT doing that because of a lack of respect. It is doing that because the interviewer wants a glimpse of how the interviewee deals with unknowns, or with stressful situation.

Even the most smart and decent person can turn into a stupid jerk during a crisis situation, and that could be very bad for business. Nobody likes an employee that might be triggered by just an unpleasant question.

I understand exactly what they think they are doing, but there are other far more representative ways of doing it. While of course I'd give it a reasonable shot, I'd walk out of there and think "what a jerk", and think twice about what I'd be letting myself in for for the reasons I already cited.

I can see where you are coming from, but I wouldn't go so far as that without corroborating evidence.

Once, during the milkround I was asked how, if I was an ancient Egyptian overseer, I would be able to tell that the minions had finished their pyramid-building shift. I decided to have fun, and dreamed up half a dozen different ways. Eventually the interviewer asked me if I knew the answer he was looking for (a large eggtimer, yawn); only at that point did I come to a negative conclusion.

Of course, depending on the job, knowing how someone might behave in a difficult situation might be important information directly related to the job. It all depends on the job.

I wonder what they'd think of someone who stated "11,750,000, as at January 1 2017, a 27 year low"?

I would ask where that information came from, and how accurate they thought it was

[NivagSwerdna]

Once, during the milkround I was asked how, if I was an ancient Egyptian overseer, I would be able to tell that the minions had finished their pyramid-building shift. I decided to have fun, and dreamed up half a dozen different ways. Eventually the interviewer asked me if I knew the answer he was looking for (a large eggtimer, yawn); only at that point did I come to a negative conclusion.

I think the Hourglass is a relatively new invention and that might not be the correct answer.  I think there is evidence of water based timing in Egypt... not sure about Hour glasses... think of the difficulty of construction (and also you can't use regular sand... ).

[brucehoult]

Import and export are important, as some countries vastly overproduce and others do the opposite.

Certainly, and that would be a very important consideration if one wanted to estimate for New Zealand where the vast majority of dairy and meat production is for export!

In the case of Canada I have just enough domain knowledge (coming from a dairy farming family in New Zealand, though I escaped to university and the Big Smoke) to know that Canada isn't a big exporter or importer of dairy and meat, producing pretty much exactly what it needs itself. In fact Canada is infamous for its "supply management" system that stands in the way of anyone (such as NZ and Australia) that wants to make proper free trade agreements (the kind that fit on both sides of a sheet of A4) including Canada.

[ejeffrey]

When I was in my interview for my current company, the manager engineer asked me: "How many cows are there in Canada?"

So... how would all of you answer that?

IMO that is not a very good interview question for most jobs, but the way to answer is to state that you would need to make some guesses or estimates, and state what they are.

Something like:

Population of canada is X
Milk/dairy consumption per capita is Y
Dairy production per dairy cow is Z

Beef consumption per capita is A
Average yield of a slaughtered cow: B
Average age of slaughtered cow: C

Multiply/divide those out to get an estimate, then state what factors are not included: breeding stock, imports/exports, etc.

Its not a hard question if you are used to Fermi problems, and the point would be to see if you can identify the key parameters that you would need to research to get an accurate answer.  I just think they are not usually great interview questions because they are sort of designed to "shock" someone already in a stressful situation.

[T3sl4co1l]

When I was in my interview for my current company, the manager engineer asked me: "How many cows are there in Canada?"

So... how would all of you answer that?

I'd need more information around the context of the question to give a more effective answer.

Furthermore, smart arse trick questions like these which put the interviewee under unrealistic, unnecessary and undue pressure simply indicate to me they're not a particularly nice place to work in, and have too much time on their hands not doing real work. Can you imagine how intrusive and overbearing their HR dept is likely to be in your daily working life? No thanks!

Either you Just Get It(R), or you understand what one of these questions is getting at; or... you don't get it at all and you get frustrated with the whole idea.

What's the idea?  Orders of magnitude.  IGNORE REAL NUMBERS, COMPLETELY.  Scary?  Why?  How is it any more scary than rounding a number to any other arbitrary precision?  See, the idea generalizes, too!  That is why it is useful.

1. There's about 10M people in Canada.  (Why?  Well, Canada is smaller than the US, and the US is about 100M, so it must be around 10M.)
2. A person eats about a pound per week of steak, and there are a hundred weeks in a year.  (The first is an underestimate, and the second an overestimate.  Justifying rounding like this is okay, just don't sit there thinking about how all the rounding stacks up, you'll take forever.)
3. A cow lives about 10 years before slaughter.
4. A cow is made of about 1000 pounds of steak.

Therefore, there's about 10B pounds of steak consumed in 10 years, and 10M cows to supply it (in equilibrium).

Or in short, about as many cows as people, which might be an easy shortcut.  (On that note, pigs are about 1/10th the size of cows, so you'd expect 10 times more.  Chickens live much less than a year, so should number about the same as pigs.)

Multiplying orders of magnitude is adding exponents, so you can do it in your head without jotting down a single thing on paper.  Anyway, if this sort of question pops up out in the field, you won't have access to the internet anyway, and maybe not the pad of paper either!

Such arguments are great for ballparking whether something is even feasible at all.

The technique I think was first popularized as Fermi's estimation method.  Of course, a man of that intellect would be more than comfortable processing the numerical value as well as the exponent, but the average joe can still get within a factor of 10 or 100 without any real practice!

Tim

[sokoloff]

Such arguments are great for ballparking whether something is even feasible at all.

The real world application of this is quite valuable at times. I can't tell you how many times I've gotten a proposal at work where it's clear that some fair amount of effort went into the background research and preparation and yet I can find a glaring hole in the proposal in 15 minutes of reading and checking for plausibility of numbers. (This proposal would require that our conversion rate in e-commerce would increase from 6% to 125%, or this proposal ballparks our current average shipping prices for 1# air freight packages domestically in the US at $80/parcel.)

[AndyC_772]

Honesty is the best policy here.

Completely agree; the alternative really isn't an option.

If I interview you for a job, the thing I'm most interested in is how you approach unfamiliar technical problems. After all, that's how engineers spend the majority of their time.

That means I need to first find out the limits of what you do know, so I can start to explore what lies just beyond that limit. I want to find out whether you're honest with yourself (as well as with me, of course) about where your expertise lies, and how you approach technical problems that you can't already tackle with absolute confidence.

However smart and experienced you may be, I can virtually guarantee there will be some technical topic which I know better than you do - bearing in mind, of course, that I'm the one who gets to choose the topic.

All I'm doing is trying to make you think, to extrapolate from things you already know, to help solve problems you've not encountered before. I'm not trying to be mean; I just don't learn anything particularly useful if I ask a question you already know the answer to.

My favourite interview question is a bit more immediately relevant to a junior EE than counting cows.

"Imagine you're designing a new board, and you have a choice of two different package types for a major component. One is a BGA, the other a QFP. Tell me about all the things you might consider when making a decision between one and the other."

It's open-ended, has no 'right' answer, and allows a candidate to demonstrate knowledge across a breadth of areas including design, manufacturing, supply chain and repair. I can choose to either mention any that are missing, or delve into more detail around areas that are obviously of interest.

[WorldPowerLabs]

Well, first let's model the cows as spheres...

...no, wait; that would be a physics problem.


In all seriousness, honesty has always worked best for me.  In an interview setting, I'll answer the question as far as I can with what I do know and I'll see if there's any way to estimate or derive the ultimate answer the interviewer is asking for.  If I get completely stuck, I'll say that the question is a bit outside of what I've historically worked on, but that I can find out.  At work, on occasion, when a supervisor has asked me to do something WAY outside of my realm (like redesign a complex cast metal housing), I've said that I'm not an expert in what he's asking for and that while I'm willing to work on it, it may be best to ask one of the mechanical engineers.  Perhaps not the best attitude, but I'm a bit beyond the "young and eager to please everyone" stage of my life... now I'm a semi-cranky, semi-jaded, chronically tired and overworked EE.

Funny story:  I went on an interview once and the interviewer asked me how I would design (on a block-diagram level) a filter to address a certain problem they were having.  Well, I thought about it for a minute and explained what I would do.  The guy's eyes widened a bit and he exclaimed, "I think that would work!  I hadn't thought of that!"...

On another interview, the interviewer was a bit of a cocky and confrontational fellow, and I didn't much care for his attitude.  He asked me how I would design a circuit to perform a certain function based on some digital inputs, without using a microcontroller.  I worked out a design using flip-flops and some glue logic, but in retrospect, I really wish I'd used relays.

When I was in my interview for my current company, the manager engineer asked me: "How many cows are there in Canada?"

So... how would all of you answer that?

[tggzzz]

My favourite interview question is a bit more immediately relevant to a junior EE than counting cows.

"Imagine you're designing a new board, and you have a choice of two different package types for a major component. One is a BGA, the other a QFP. Tell me about all the things you might consider when making a decision between one and the other."

It's open-ended, has no 'right' answer, and allows a candidate to demonstrate knowledge across a breadth of areas including design, manufacturing, supply chain and repair. I can choose to either mention any that are missing, or delve into more detail around areas that are obviously of interest.

Exactly.

Such questions can be very revealing about the candidate's previous experience, their ability to extrapolate and "pick up" unfamiliar topics in the future, and their general personality. Counting cows is for HR-droids, managers, and engineers that are out of their depth[1]

[1] unless they demonstrate otherwise by asking relevant questions

[KJDS]

I've interviewed many people over the years, and it's very easy to spot bullshit.

I also ended up having a rather frank discussion when being interviewed when being questioned about Smith charts. If you're doing the interviewing and don't know the subject well, at least make sure that the answer you have written down is correct.

[cdev]

Whenever I have gotten a question like that in an interview I am always honest and then I just go through the thought process with the interviewer with the goal of us finding out the answer. That's always the thing to do.

In a job too, being able to learn new skills quickly is more important than anything else.

[BBBbbb]

Meanwhile in Australia:


Scam attempt or a fake, but still a good laugh.