September 11, 2013 at 8:29 am
simon.crick (9/11/2013)
I agree 100% with the idea of a probationary period.However... you still have to decide which candidate to hire for the probationary period, and it is still 6 months wasted if you get the decision wrong.
Therefore, you still need some way to choose the best candidate, and I still believe academic qualifications are the most reliable indicator of long-term potential.
Sitting someone in front of SSMS will tell you how useful they will be on day 1, but not how useful they will be in 6 months or a year.
Simon
I like the idea of probation periods, but I agree that it doesn't reduce much of the pressure. You are still out a person and time if things don't work out.
I'm not sure I think academics are a long term indicator. They do show someone is willing to work through a long term project (a degree), and can learn, but it doesn't speak to long term motivation, to being motivated to be a DBA, etc.
Follow me on Twitter: http://www.twitter.com/way0utwest
Forum Etiquette: How to post data/code on a forum to get the best help
My Blog: www.voiceofthedba.com
September 11, 2013 at 8:35 am
In Steve's first article in this series he pointed out that most of the certifications test one's ability to memorize. This hits the nail right on the head. It's a skill I personally stink at. I hold no certifications and yet I've never been unemployed for more than a few weeks during my IT/development career. Knowing the technical aspects of the technology is only one part of a good fit for the candidate. And even with that I've always questioned the value of certification to test the technical aspect.
I've worked with developers and IT people that have had several certifications and they can't think, reason or exercise good judgement or debug their way out of a paper bag. The certifications don't test that. They test memorization. When I interview, I want to know if a person can DO development. So I sit them in front of VS or SSMS with a live internet connection and make them write something. I want to see how they go about solving the problem. Are they reasonably facile with the tools or clicking around aimlessly? I don't care that they go to the internet to look something up. Most of the troublesome problems I face are with some third party's piece of code that doesn't do what it says it should do. Can the candidate research problems and find solutions? What's their temperament like when they struggle with something? Certifications don't test that.
Some of the most intelligent productive people I've worked with didn't have degrees or certifications. They had a passion for creating. That's what I look for in a candidate.
September 11, 2013 at 8:45 am
Steve Jones - SSC Editor (9/11/2013)
simon.crick (9/11/2013)
Why not base your decision on the candidate's academic qualifications?College and university exams are far more rigorous and comprehensive than any interview tests can ever be.
Choose a candidate with good exam grades in a relevant subject area from a reputable college/university and you can't go far wrong.
Base your decision on a small number of random interview tests and you will get random results.
Simon
Not sure I like this. College/University isnt' a training ground. There are plenty of people, for all I know the majority, that work in technology, but didn't major in CS/IS.
The college exams often don't translate to the real world. Even in closely aligned programs, like medicine, law, accounting, there is a good amount of training done for the candidates later. I do wonder, however, how often companies in those fields regret their hires?
I am one of them. I got my first job in computing before I went to college/university. After 5 years I realised I could never compete with graduates who had a deep understanding of the underlying concepts in computer science, so I went to college, got my A-levels, then went to university and got a degree in Mathematics. I then went back into computing and now work as a senior software engineer responsible for some very large and complex databases.
When I left university I would have certainly failed the SSMS test because I had never had any exposure to Microsoft technology. Therefore, I know from personal experience that rejecting someone based on an SSMS test would be wrong, and I'm pretty sure that positively choosing someone based on a limited SSMS test would also be wrong because it only proves a very narrow range of skills and doesn't really prove the deep understanding of the underlying concepts that is necessary for problem solving and innovation.
I'm not saying that academic qualifications are a 100% reliable indicator of ability, but based on my personal experience from both sides of the fence, I strongly believe they are the best available indicator, certainly for entry-level jobs.
Simon
September 11, 2013 at 9:10 am
Steve Jones - SSC Editor (9/10/2013)
Comments posted to this topic are about the item <A HREF="/articles/Editorial/102072/">The Future of Knowledge Measurement</A>We'll never be able to completely and accurately measure a person's skills in technology. At least not in any cost- and time-effective way.
First, a lot of you will recall I am not a fan of certs. I think they are useless. However that does not mean I agree that you can't measure skill. I think you can. Unfortunately, companies don't want to pay for proven skills. I was going to include some examples, but those aren't quite on topic so I won't. Suffice to say that certs aren't going to be fairly valued on either side they way they are currently done.
OK, so how can we measure skills, assuming we can find companies that are willing to compensate employees for being productive? First, it is easy to hand out questions like you suggest. This is part of testing ability - if you know the answers you PROBABLY have some level of skill in the field. Next, interview them. Ask them how they would handle very specific issues. Some of these should be soft skill questions, some should be trouble shooting questions. These should be "think through the problem" type of questions, not specific "what would you type at a command line" questions. You should get a feeling for how a person thinks, how they analyze, how they develop a theory of what might be wrong. You can even produce scenarios where the person being interviewed is led to a false conclusion, to get an idea of how they pigeon hole themselves. There is no reason this could not be automated, with written answers.
Slightly off topic, but truly related is that part of the process of measuring someone's technical skills should be to include communication. I have worked with lots of foreign born individuals that could communicate in English as well or better than those born and raised in English speaking areas. These individuals have typically been superior in what they did. Those that couldn't write a complete sentence, or couldn't convey a thought or question, typically can't program to save their life! Why? Well learning a spoken/written language is not much different than learning a programming language. There are technical people that never need to talk to non-technical users. Those that do, should be evaluated for the ability to SPEAK CLEARLY about their area of technical expertise.
Dave
September 11, 2013 at 9:19 am
simon.crick (9/11/2013)
Why not base your decision on the candidate's academic qualifications?College and university exams are far more rigorous and comprehensive than any interview tests can ever be.
Eeek! Depends on where they went to school. I know of one university that values income more than education, and refuses to fail ANYBODY no matter what. I doubt they are unique.
Then you have the professors that teach opinion and not fact, and grade based on that. How valuable is if that someone learned an opinion?
A college degree says that an individual is able to learn, nothing more. It does NOT mean they know anything in particular, nor does it make them an expert. Everyone I know that has taken networking course work in college is of the opinion that it was nothing more than theory and basically useless in the real world. I know my database classes were at least as bad. I would go farther and say that about half the database courses I have taken at professional training schools have been just as useless.
Dave
September 11, 2013 at 9:34 am
djackson 22568 (9/11/2013)
simon.crick (9/11/2013)
Why not base your decision on the candidate's academic qualifications?College and university exams are far more rigorous and comprehensive than any interview tests can ever be.
Eeek! Depends on where they went to school. I know of one university that values income more than education, and refuses to fail ANYBODY no matter what. I doubt they are unique.
What university would that be?
September 11, 2013 at 9:43 am
patrickmcginnis59 10839 (9/11/2013)
djackson 22568 (9/11/2013)
simon.crick (9/11/2013)
Why not base your decision on the candidate's academic qualifications?College and university exams are far more rigorous and comprehensive than any interview tests can ever be.
Eeek! Depends on where they went to school. I know of one university that values income more than education, and refuses to fail ANYBODY no matter what. I doubt they are unique.
What university would that be?
A university more interested in income does not strike me as one that would hesitate to sue if their name was mentioned, so I don't believe I will say anything more. As I said, I doubt there is only one like that.
Dave
September 11, 2013 at 9:50 am
I agree with Dave62 on this, a probation period of 6 months. That's what my company has and it has worked for us, some mistakes but that can be expected. Nothing you do will guarantee a 100% success. I would say looking at someone's work history would help weed out some bad people. If you haven't stayed at a job for more than a year or two would be a giant red flag to me. Either you couldn't do the job or you are only interested in 'moving up'. And what guarantee will I have that you will stay here for very long.
With the easy of, coming to this site:-), and other online sources to find answers to help you solve any problem, to dismiss someone just because they can't memorize some commands for your test seems pointless.
I'm no DBA, but I'm sure I could find the answers I need if I had to do it. Doesn't mean it would be the best or fastest way to do it and it may take me more time. But it could be done. The willingness to learn is key. IMHO
-------------------------------------------------------------
we travel not to escape life but for life not to escape us
Don't fear failure, fear regret.
September 11, 2013 at 9:55 am
Recently, we were looking for a SQL developer that could become part of our team, could develop, and could convert a user story into a solution. We had run into people who look good on paper, but many who had embellished their SQL skills. To combat this I devised a problem for them to solve.
The problem provides a script to generate 2 tables with data, and then describes what the data and the tables represent. Then we create a hypothetic business problem/objective that can be solved with the information in the tables. The "Correct" solution requires multiple levels of knowledge, requires careful reading and checking of the requirements contained within the business problem. A "basic" solution requires about 30-40 minutes of analysis of the data in the tables and analysis of the requirements. I expect the actual solution to take about 10 minutes to write and test. My own solution is a Select statement of about 8 lines involving an inner JOIN, and either a UNION or OR condition in the where clause, an aggregate function, a GROUP BY, and a HAVING.
After initial screening of resumes, we send out the problem to the candidate and ask them to send their solution in 1-2 days before their interview date and be prepared to discuss their solution in-depth during the interview. I make myself available to the candidate by cell phone if they have questions about a nuance contained in the problem at any time before the interview. I review their solution and may test it if it involves an approach that I wasn't anticipating. The candidate is expected to be able to speak about their solution during the interview
This is what I am trying to achieve: 1) Can the candidate figure out how to run the script to load the data. 2) Can the Candidate read and correctly interpret English (my clients are all English speakers). 3) Does the candidate pay attention to details and nuances contained in the business problem. 4) Does the candidate reach out to someone to clarify a requirement or just "assume" they know what a vague requirement really means. 5) Ascertain whether the candidate is a brute force or elegant solution developer. 6) Whether the candidate understands the basic constructs of SQL. 7) Whether the candidate understands more advanced concepts of set theory (how they combine the tables group where clauses). 8) give a common framework for having the candidate discuss a technical solution.
In reality, I'm not at all interested in whether the candidate produces a solution with the 23 rows that a correct solution produces. I'm concerned with all of the thinking, interpreting, and interacting skills. I can teach the language and the business, but I can't teach how to think, interpret, interact, research and pay attention to details.
If you'd like a copy of this exercise, I'd be happy to forward it to you privately.
Jim
September 11, 2013 at 10:23 am
jmoney 69422 (9/11/2013)
...I devised a problem for them to solve.
I like this approach. If the candidate was inclined to "get help" from somebody else, they would be caught out when discussing the solution with you in the interview. I think I'd like to take a look at your question.
Hakim Ali
www.sqlzen.com
September 11, 2013 at 10:45 am
jmoney 69422 (9/11/2013)
...I devised a problem for them to solve.
Great idea, if I was interviewing I would like this approach.:-) Got any openings? 🙂 Just kidding, or am I?;-) Guess it would depend on location, benifits and pay. 😀
-------------------------------------------------------------
we travel not to escape life but for life not to escape us
Don't fear failure, fear regret.
September 11, 2013 at 10:47 am
djackson 22568 (9/11/2013)
patrickmcginnis59 10839 (9/11/2013)
djackson 22568 (9/11/2013)
simon.crick (9/11/2013)
Why not base your decision on the candidate's academic qualifications?College and university exams are far more rigorous and comprehensive than any interview tests can ever be.
Eeek! Depends on where they went to school. I know of one university that values income more than education, and refuses to fail ANYBODY no matter what. I doubt they are unique.
What university would that be?
A university more interested in income does not strike me as one that would hesitate to sue if their name was mentioned, so I don't believe I will say anything more. As I said, I doubt there is only one like that.
While I know theres some examples of institutions failing accreditation, this doesn't mean that we should discard the benefits of higher ed, especially in the context of this discussion.
September 11, 2013 at 11:03 am
Hi,
If you consider the rise of Massively Open Online Courses (MOOC), an open source testing ability with 'reasonable' testing fees and security would be a great thing. Professionals need a way to expand their knowledge and be able to demonstrate that knowledge in an independently verifiable way.
Also, if the testing facilities were separate from the MOOCs, you could ask the students to list the preparation courses they took, and list their efficacy as well. (Test driven teaching anyone?)
People will always game any system. But, we want to have a way for anyone to learn, improve, and prove that knowledge. We will still have to interview, test, and have probation periods, but we can bring more interested and qualified people into the workforce.
Thanks,
Peter
September 11, 2013 at 1:24 pm
simon.crick (9/11/2013)
I agree 100% with the idea of a probationary period.However... you still have to decide which candidate to hire for the probationary period, and it is still 6 months wasted if you get the decision wrong.
Therefore, you still need some way to choose the best candidate, and I still believe academic qualifications are the most reliable indicator of long-term potential.
Sitting someone in front of SSMS will tell you how useful they will be on day 1, but not how useful they will be in 6 months or a year.
Simon
Still no guarantee that people will change after that 6 months to a year period and they become the employee from hell. I have seen them slyly work the system and the people they work for until they are in like a tick on a dog's behind. In the government sector it is much harder to get rid of them then. It's not impossible, just much harder.:-D
"Technology is a weird thing. It brings you great gifts with one hand, and it stabs you in the back with the other. ...:-D"
September 11, 2013 at 8:39 pm
jmoney 69422 (9/11/2013)
The problem provides a script to generate 2 tables ... a GROUP BY, and a HAVING.
Actually I would be interested in seeing this test as well.
I would probably go through #temp tables to get to it. But I have been doing SPs for enough years that I take a different attitude on some stuff.
----------------
Jim P.
A little bit of this and a little byte of that can cause bloatware.
Viewing 15 posts - 16 through 30 (of 45 total)
You must be logged in to reply to this topic. Login to reply