Do we really need puzzles for interviews?
Despite the fact that the blog is called “Learn to work,” I will not teach anyone here, but simply share my thoughts about interviews and puzzles.
Recently, I conducted a survey "How do you feel about using puzzles at interviews?" . Taking into account the fact that polls are not welcome at Habré (although from my point of view this is not correct, as Habr is a good resource for conducting any kind of research and polls are not always created according to the principle “I am interested”), but however, he did not go into deep minus, which indicates a relative interest in the topic. Thanks for voting and supporting the poll.
We will quickly go through voting options (we will go from a negative attitude to a positive one).
')
It would be foolish to expect that everyone likes to solve puzzles, and it would also be naive to believe that everyone loves this particular style of interviewing. About 30% of voters believe that the use of puzzles does not correlate with the ability to solve real problems.
This question is very strongly correlated with other questions like “whether you need a higher education,” “whether you need knowledge of mathematics,” etc. Despite the fact that very many people before the loss of their pulse prove that they don’t need to learn and that they don’t really need to know anything, the majority still agree that having these qualities is a definite plus . So here we can assume that the ability to solve puzzles is a plus, but not a necessary skill.
Let's answer the question: Does the ability to solve puzzles correlate with mental abilities? I think the answer is yes. But here I am risking to be immediately abandoned rotten tomatoes shouting "and who are you to decide for everyone?" And they will be right. But let's look at it on the other hand, more global. The approach to mental testing was born not today, but a hundred years ago. And the first to adopt it were large innovative companies and the defense industry. This is how large companies operate - from Microsoft to Google. If we assume that there is no connection, it turns out that Google and Microsoft are full of profane people who are unable to distinguish a normal candidate from a bad one without psychological distortions. Accordingly, such an approach would have to bring in the future large financial losses (or bankruptcy of these companies), which we do not observe. Thus, despite the fact that it is theoretically very difficult (even practically impossible at this stage of development) to prove unambiguous connections, in practice we get quite adequate results. This I led to the fact that further we will assume that some connection is still present.
This is a good position, which is held by 25% of respondents. In fact, a neutral approach is often the most effective, since a person abstracts from tasks and sees before himself only one goal - to successfully pass an interview and get a good offer from an employer. But this option is not very relevant to this article, so we will not dwell here for a long time.
So consider Microsoft, Google, big banks and another 18% of Habr's readers. This may indicate that those who voted for this option are able to solve puzzles or they are impressed by this style of interviewing . With the first option is clear - it turns out to solve puzzles, why treat positively? But the second option is associated with personal preferences, so the percentage was a little less. Probably, sad interviews are already fed up, so I want something new and interesting.
I admit that I voted for this option and the purpose of the survey was to check how many people share my opinion. As expected - a lot, namely a little less than 29%. Despite the fact that I really like to solve all kinds of problems and spent a lot of time on this issue, I began to notice some problems in this kind of knowledge testing approach during real interviews. We will discuss this further.
There are actually more disadvantages than advantages. Consider them in more detail:
In addition, the person conducting the interview must answer a few questions for himself:
Another ethical issue is the fact that many (mostly strong) candidates may regard the puzzle questions as an insult. This can ruin the interview at the very beginning.
Puzzles come in different types - logical tasks, tasks for weighing, tasks without an answer, etc. In fact, it is not the ability to solve puzzles that are often tested, but it is checked whether the particular puzzle belongs to which class of problems . Otherwise: what they want to hear from you? As soon as you understand what it is, you have solved the problem by 85%.
I will give a simple example in which I will try to explain why this does not always work.
Once at school competitions they gave tasks of this type: prove that the sum of cubes of three consecutive positive integers is divisible by 9. Many people in this place thought it was elementary, others thought hard how to calculate this, especially without computing devices. And in fact, the solution is very simple, if you know that this is a mathematical induction problem, the solution of which is to prove that the expression is valid for n = 1, n = k and n = k + 1, i.e. the problem ultimately boils down (mostly) to simple arithmetic operations. Thus, a person who knows the type of problem will solve it quickly (and almost any task), while another person who does not know the approach will hardly solve it in a reasonable time (I doubt very much that a candidate is in a place or two in a minute or two) will offer just such an answer). There is a problem when testing is not what was intended.
Why doesn't it work for real-world tasks? Probably, because, when solving real-world problems, as a rule, there is already at least some kind of input data - scope, ready-made solutions, code samples, Google, etc. At a time when you have no source data for solving problems.
Therefore, if a company still wants to use puzzles, then it needs (IMHO):
Although, I believe that if a company is determined to use puzzles, then the best solution is to select a candidate for 2-3 hours, when he can be alone and think about solving problems. This will help him gain strength, collect his thoughts and show his best.
I didn’t like the interviews with the puzzles that I went through, despite the fact that I usually dealt with them. They were really some kind of non-professional and detached from life. It seems that the Western colleagues shared their methods, and forgot to learn how to use it. Which leads to funny and sometimes bad results for candidates and companies.
Finally, I will recommend the book "How to move Mount Fuji," because of which all these thoughts were born. As for me, the book is definitely worth reading, but you should not take everything as a truism.
Thanks for attention!
Recently, I conducted a survey "How do you feel about using puzzles at interviews?" . Taking into account the fact that polls are not welcome at Habré (although from my point of view this is not correct, as Habr is a good resource for conducting any kind of research and polls are not always created according to the principle “I am interested”), but however, he did not go into deep minus, which indicates a relative interest in the topic. Thanks for voting and supporting the poll.
We will quickly go through voting options (we will go from a negative attitude to a positive one).
')
I treat the use of puzzles in interviews negatively, because I believe that the ability to solve puzzles does not correlate with the ability to solve real problems
It would be foolish to expect that everyone likes to solve puzzles, and it would also be naive to believe that everyone loves this particular style of interviewing. About 30% of voters believe that the use of puzzles does not correlate with the ability to solve real problems.
This question is very strongly correlated with other questions like “whether you need a higher education,” “whether you need knowledge of mathematics,” etc. Despite the fact that very many people before the loss of their pulse prove that they don’t need to learn and that they don’t really need to know anything, the majority still agree that having these qualities is a definite plus . So here we can assume that the ability to solve puzzles is a plus, but not a necessary skill.
Let's answer the question: Does the ability to solve puzzles correlate with mental abilities? I think the answer is yes. But here I am risking to be immediately abandoned rotten tomatoes shouting "and who are you to decide for everyone?" And they will be right. But let's look at it on the other hand, more global. The approach to mental testing was born not today, but a hundred years ago. And the first to adopt it were large innovative companies and the defense industry. This is how large companies operate - from Microsoft to Google. If we assume that there is no connection, it turns out that Google and Microsoft are full of profane people who are unable to distinguish a normal candidate from a bad one without psychological distortions. Accordingly, such an approach would have to bring in the future large financial losses (or bankruptcy of these companies), which we do not observe. Thus, despite the fact that it is theoretically very difficult (even practically impossible at this stage of development) to prove unambiguous connections, in practice we get quite adequate results. This I led to the fact that further we will assume that some connection is still present.
I treat the use of puzzles at interviews as neutral, because I believe that the interviewer has the right to ask any questions at the interviews
This is a good position, which is held by 25% of respondents. In fact, a neutral approach is often the most effective, since a person abstracts from tasks and sees before himself only one goal - to successfully pass an interview and get a good offer from an employer. But this option is not very relevant to this article, so we will not dwell here for a long time.
I am positive about using puzzles for interviews, because I think that the ability to solve puzzles speaks of a high level of intelligence
So consider Microsoft, Google, big banks and another 18% of Habr's readers. This may indicate that those who voted for this option are able to solve puzzles or they are impressed by this style of interviewing . With the first option is clear - it turns out to solve puzzles, why treat positively? But the second option is associated with personal preferences, so the percentage was a little less. Probably, sad interviews are already fed up, so I want something new and interesting.
I love to solve puzzles, but I think this is a bad way to test candidates.
I admit that I voted for this option and the purpose of the survey was to check how many people share my opinion. As expected - a lot, namely a little less than 29%. Despite the fact that I really like to solve all kinds of problems and spent a lot of time on this issue, I began to notice some problems in this kind of knowledge testing approach during real interviews. We will discuss this further.
The complexity of this approach
There are actually more disadvantages than advantages. Consider them in more detail:
- you need to have in stock a large number of tasks and puzzles;
- it is necessary to understand (at least remotely) which types of tasks need to be given to a specific candidate;
- criteria for evaluating responses should be clear to the candidate;
- Candidate should not know questions and answers (this is a problem, since it is possible to “prepare” for basic questions)
- the interviewer himself may not know the answers to the questions and their meaning ;
- The candidate's answer may differ from the expected one, which again raises the question of the evaluation criteria.
In addition, the person conducting the interview must answer a few questions for himself:
- Will I take a person to work if he answers n questions correctly?
- Will I take a person if he answers this question and does not answer this one ?
- How much time will I give a person thinking?
- Will I consider anything other than the ability to solve puzzles?
Another ethical issue is the fact that many (mostly strong) candidates may regard the puzzle questions as an insult. This can ruin the interview at the very beginning.
Why are puzzles a bad job interview?
Puzzles come in different types - logical tasks, tasks for weighing, tasks without an answer, etc. In fact, it is not the ability to solve puzzles that are often tested, but it is checked whether the particular puzzle belongs to which class of problems . Otherwise: what they want to hear from you? As soon as you understand what it is, you have solved the problem by 85%.
I will give a simple example in which I will try to explain why this does not always work.
Once at school competitions they gave tasks of this type: prove that the sum of cubes of three consecutive positive integers is divisible by 9. Many people in this place thought it was elementary, others thought hard how to calculate this, especially without computing devices. And in fact, the solution is very simple, if you know that this is a mathematical induction problem, the solution of which is to prove that the expression is valid for n = 1, n = k and n = k + 1, i.e. the problem ultimately boils down (mostly) to simple arithmetic operations. Thus, a person who knows the type of problem will solve it quickly (and almost any task), while another person who does not know the approach will hardly solve it in a reasonable time (I doubt very much that a candidate is in a place or two in a minute or two) will offer just such an answer). There is a problem when testing is not what was intended.
Why doesn't it work for real-world tasks? Probably, because, when solving real-world problems, as a rule, there is already at least some kind of input data - scope, ready-made solutions, code samples, Google, etc. At a time when you have no source data for solving problems.
Therefore, if a company still wants to use puzzles, then it needs (IMHO):
- warn the candidate that puzzles will be used
- explain the evaluation criteria and the time that the candidate will have to think about
- say whether only the right answer is important, whether the reasoning is important and in what proportions it will all be evaluated
Although, I believe that if a company is determined to use puzzles, then the best solution is to select a candidate for 2-3 hours, when he can be alone and think about solving problems. This will help him gain strength, collect his thoughts and show his best.
I didn’t like the interviews with the puzzles that I went through, despite the fact that I usually dealt with them. They were really some kind of non-professional and detached from life. It seems that the Western colleagues shared their methods, and forgot to learn how to use it. Which leads to funny and sometimes bad results for candidates and companies.
Finally, I will recommend the book "How to move Mount Fuji," because of which all these thoughts were born. As for me, the book is definitely worth reading, but you should not take everything as a truism.
Thanks for attention!
Source: https://habr.com/ru/post/97476/
All Articles