What are the ways to measure a programmer’s knowledge?

Asked

Viewed 3,487 times

3

Just as there are techniques to measure productivity, there are techniques or ways to measure a programmer’s knowledge (both in a specific technology and in general)?

If yes, which ones? In programming there are levels of knowledge (Ex: I am Full Programmer C#, I must know how to implement x in the language)? Is there a linearity in this or is it possible to skip steps?

It would be a way to evaluate a programmer applying a proof based on the position he will exercise and what tools and technologies he will work with? (with the purpose of knowing whether it will achieve the proposed objectives, without external assistance as Google, to know his real knowledge).

methodology
  • 4

    I understand it’s a question with intent technical, but, I believe, that it is very easy to direct itself towards a subjective path, since knowledge (by itself) is already something quite subjective. Anyway, I think the path will be portfolio + certificates + knowledge tests.

  • The answer is subjective. As far as I know, there is no criterion set for universal measurement, as in the English proficiency tests (TOEFL, IELTS). And even so, these proofs have subjective collection and different focuses (business, academic, etc.). I believe that portfolio indicates more experience than level of knowledge, so I’ve seen it happen several times. Knowledge tests + certificates is a way. But they should not be taken into account for comparison of programmers' knowledge.

  • 12

    Probably tests. They will throw stones at me for saying this, but certificates and diplomas are far from indicating true knowledge. Academy is great until the subject sits in the chair and face a real problem with real time and that involves a mixture of situations that is no longer ideal from the beginning. Good tests have a higher chance to reflect the profile of the programmer.

  • 1

    Publish 10 questions and 10 answers in Sozão and do not receive none downvote == reasonable grade. Upvotes will measure the OK++ grade:D

  • 1

    I believe it is worth a debate on the subject. I always liked the subject. I even worked in a company where positions were well defined in 7 levels for developer. You knew what was expected of each position and the only thing you could do to evaluate it were Trials. We used coaching to track growth and help acquire knowledge. If there is a debate on the subject, I would like to participate

  • 2

    @Caputo I agree to make the discussion, and posted a question here that might help start it. :)

  • 1

    Nothing like putting said programmer to program, then you see the quality :)

  • I opened the vote to reopen, I think it’s concise now.

  • 4

    This question is being discussed at meta.

  • 2

    IMO The only valid metric to measure programmer knowledge is the number of times per minute you exclaim "what the fuck" when reviewing your code.

Show 5 more comments

3 answers

4

In Brazil, and specifically in the academic sphere, there is a unified proof (that is, applied throughout the national territory - and in some neighboring countries, such as Peru) called National Examination for Admission to the Post-graduation Program in Computing (POSCOMP).

This examination is organized by Brazilian Society of Computing (SBC) and aims to evaluate candidates for graduate programs in the country. The test evaluates knowledge in three fields - mathematics, computing fundamentals (theory) and computer technology (systems and applications) - distributed into 70 multiple choice questions. The topics covered are part of the Computer Science curriculum of the main Brazilian universities and colleges, and therefore do not evaluate the author in a specific technology, but as a form of general knowledge (although it is common to fall into the field of technology questions about SQL, syntax of languages like C++ and Java, etc).

Candidates are not "approved" or "failed" on the test. They receive individually (there is no general classification table) their evaluation in terms of the score obtained (hits per field) compared to the average and standard deviation of the country. I believe it is worth for a year (until the next race), but I did not find this information on the site to confirm. The individual result is not mandatory in many of the universities, serving only as an additional criterion for the selection of candidates. I also have no information regarding its use in other contexts.

The exam is considered quite difficult, having as average little more than 30% accuracy in your questions (source: Wikipedia).

  • I regard this test as a measurement of general knowledge and not specific. In addition to what is even written "aims to evaluate candidates for graduate programs". And who is not interested and does not want to do post? Would that be a negative point, so to speak? In the academic world, in particular, as it is described early on, it would be interesting, more for the labor market would not count as something that measured the professional.

  • @Leofelipe Well, the test could be used by companies to evaluate professionals in a similar way to what the TOEFL does for the English level, if someone so desired. But I really don’t think that’s likely to happen one day. Anyway, it is a form of evaluation that exists and is used, albeit in a specific context. I just made that clear in the answer. In fact, the question also did not limit context, so...

  • Yes @Luiz Vieira, could, but I think it would be work in vain. Because the key point of the question is "What are the ways to measure the knowledge of a programmer?". This makes it something already specific to a company, since it will want to hire a programmer of some certain technologies that it will work, like S#+Sqlserver+javascript, or php+mysql+jquery, or ZF2+oracle. That’s usually how the search for the company goes. So it would be more profitable, in my view, to ask for the Github link and view the applicant’s code history than the POSCOMP note (without taking its merit, of course).

  • 1

    @Leofelipe His argument about industry specificity is fair (yet the "technology" field notes in the test could be useful in the industry - but it’s silly to insist on this argument). Anyway, programmer Not just the one who works in the industry. The academy often also looks for people with specific qualifications and its own response (on evaluation with Github) is something that is also done there frequently (in the selection of specific projects). So I think the interest here in SOPT is valid for both worlds. :)

  • 1

    Claro @Luiz Vieira. ^_^

1

In addition to certificates of completion of courses, certifications etc a test requesting code examples in an environment without Internet access :) ... But in addition to technical knowledge other qualities must be taken into account, such as ability to work as a team, ethics, discipline, etc. I’ve seen excellent technicians, but scoundrels, just talking like that....

  • 1

    I agree with the certificate part, but in 80% of cases interviewers fail to develop an "exercise" that can actually measure the candidate’s capacity

  • So, @Mota, this "exercise" has currently been done by analyzing the history of Github.

1

What has happened a lot nowadays is to know what and how the programmer program. That is, see the open source projects that it contains, Github is currently used a lot to analyze this part. This way you can get an idea of how it solves the logical problems of a problem and the level of its code and the paradigm used.

In addition, of course, certifications, courses, experience time with certain technology. And also the HR part that is very difficult to identify in interviews, because many candidates do not always tell the truth about their behavior in the work environment.

Browser other questions tagged methodology

You are not signed in. Login or sign up in order to post.