Real-world experience

[Mapraputa Is]

This country's obsession with so-called "real-world" experience is simply insane. The only thing it's good for is to develop inferiority complex in newcomers. "I do not have any real-world experience, so what can I ask for? I should probably pay something, so they would allow me to work for them..." Ok, but let's look closer, what is this precious experience?
Most likely it means that you performed few dull mundane task over and over again. You did not have time to think about what you are doing or to learn new skills. I spend 11 years in real world and I wont give a dime for this [inappropriate language was removed by Mapraputa Is] "experience". Good education is what matters. I learnt much more from books, articles, Internet brilliant sites, JavaRanch, Michael Ernest, etc. etc. etc. than from all my [inappropriate language was removed by Mapraputa Is] experience.
In this country you cannot get a job as a "Key A presser" if you did not press the key "A" for last 5 years or you were unfortunately enough to have experience in pressing key "B".
Here is the apotheosis of best hiring practices:
"5 years or more desirable experience: XML, XSLT and other web-related communication design and coding; <...> 10 years programming Java and C++."
Happy hiring.

[Matthew Phillips]

I think that a lot of hiring managers don't realize what they really need. In some cases what is really needed is someone who has successfully helped another company with a similar problem. In other cases a person with well rounded knowledge would fit the bill better.
It seems that more and more, companies are trying to fit a round peg into a square hole. The experience demands far exceed the needs of the company and the pay scale they are willing to give. This was occuring even before the recession. Since the recession the problem has worsened. I think that eventually this will lead to greater project failure as key people jump to jobs they are more suited to.

[Mark Fletcher]

Isnt one of the reasons that these idiotic recruitment companies post such unrealistic requirements for experience (ie 10+ years experience in Java) is because theyre so desperate to secure a candidate to fill the position? If they can get the position filled, they get the commission on that position, rather than some other recruitment company.
I can imagine some recruitment consultants out there are so desperate to get the commission so they can put that down payment on that porsche theyve always wanted... its probably better to just sidestep these leeches and apply direct to companies.
Mark

[Rufus BugleWeed]

For sure, it is a scam. If you were a few months short of a degree with an MS on Comp Sci from a school with medium respect they would be hot on your case.
I posted two resumes on monster. Mostly my candidates have the same java skills. They have equivalent educational backgrounds. One has fifteen years of experience, the other has three.
The one with three gets hits the one with fifteen doesn't.
Those fifteen years of real world experience have proven to be very helpful in practice, but they count against me. The operators in Java are just so similar to the ones in C. The loops are so similar. The if statements seem so very similar.
Why don't they want to pay for their own experience and training instead of stealing it from somewhere else? ( GREED )
This demand for experience is how they get by with age discrimination.

[Shura Balaganov]

Originally posted by Mapraputa Is:
This country's obsession with so-called "real-world" experience is simply insane. ...
Ok, but let's look closer, what is this precious experience?
Most likely it means that you performed few dull mundane task over and over again. You did not have time to think about what you are doing or to learn new skills. I spend 11 years in real world and I wont give a dime for this [inappropriate language was removed by Mapraputa Is] "experience". Good education is what matters. I learnt much more from books, articles, Internet brilliant sites, JavaRanch, Michael Ernest, etc. etc. etc. than from all my [inappropriate language was removed by Mapraputa Is] experience.
In this country you cannot get a job as a "Key A presser" if you did not press the key "A" for last 5 years or you were unfortunately enough to have experience in pressing key "B".
Here is the apotheosis of best hiring practices:
"5 years or more desirable experience: XML, XSLT and other web-related communication design and coding; <...> 10 years programming Java and C++."
Happy hiring.

Right on the money. In fact, this position will be filled, I don't doubt it, but the person who will get hired better ask for at least 6 digits salary, or (s)he is a complete moron (knowck yourslef out by removing inapropriate language ).
There's also another tendency. Some companies have "slaves" working for them, i.e. people with H1B visas who are trying to get green cards. I've known some of these people, they are usually best-of-the-breed engineers who came from Eastern Europe or Asia, and they are getting paid VERY LITTLE. Som companies want to keep them slaves as long as possible (keep in mind, H1Bs can't simply change jobs, or they risk to be sent back). So companies have a process to get green cards. They have to prove that there are NO QUALIFIED CANDIDATES AVAILABLE in US to fill these positions. So they run ridiculous ads. After a while, their immigration lawyers collect "evidence" and send papers to INS.... This might take between 2 and 5 years. Imagine having Senior Java engineer and paying him $30-$40K - it's a party. 9/11 really spoiled this feast, so they are really pissed of now.
Here's the pandemonium:
http://www.zazona.com/ShameH1B/
Just another angle.
Last 3 interviews I had, my interviewers were scared sh1tless, because since their job requirements matched my skills pretty close, I was stronger then my potential "senior" partners. Bummer.
Shura

[Mapraputa Is]

Originally posted by Shura Balaganov:
In fact, this position will be filled, I don't doubt it, but the person who will get hired better ask for at least 6 digits salary, or (s)he is a complete xxxxx.

Hey, be careful! :roll: I am going to apply and claim seven... no, better eight years of heavy XML and XSLT experience. They said "5 years or more..." so I figured the more, the better? Shurik Balaganov will write me a confirmation letter that I did work as an XSLT developer since 86th day of Martoben 1985, and the fact that his country is presently non-existent will only add appropriate surreal tones to my brilliant resume.

[Shura Balaganov]

Originally posted by Mapraputa Is:
This country's obsession with so-called "real-world" experience is simply insane. The only thing it's good for is to develop inferiority complex in newcomers. "I do not have any real-world experience, so what can I ask for? I should probably pay something, so they would allow me to work for them..." ...
Ok, but let's look closer, what is this precious experience?
Most likely it means that you performed few dull mundane task over and over again. You did not have time to think about what you are doing or to learn new skills. I spend 11 years in real world and I wont give a dime for this [inappropriate language was removed by Mapraputa Is] "experience". Good education is what matters. I learnt much more from books, articles, Internet brilliant sites, JavaRanch, Michael Ernest, etc. etc. etc. than from all my [inappropriate language was removed by Mapraputa Is] experience.
In this country you cannot get a job as a "Key A presser" if you did not press the key "A" for last 5 years or you were unfortunately enough to have experience in pressing key "B".
Happy hiring.

AMEN
Here's my take on it. IT is a SERVICE INDUSTRY. Programmer generally is no different than waitress or store clerk. Although it employs people who make their living using brains, COMPANIES GENERALLY DON'T WANT THAT. They are interested in loading you up to wazoo with work (pressing "A" key) so that you don't use your brain much. THEY ARE NOT INTERESTED IN YOU LEARNING NEW STUFF, because once you know it, you might go somewhere else, and they'd have to retrain another person to do your job.
THERE IS ENOUGH PEOPLE IN POWER IN THIS COUNTRY. THEY CALL ALL THE SHOTS, AND THEY ARE NOT INTERESTED IN SMART PEOPLE, JUST IN OBLIGING ONES. THEY HAVE ENOUGH SONS, DAUGHTERS AND OTHER RELATIVES WHO NEED WELL PAYING JOBS TOO.
If you are super-smart, or super-slick, you might be able to get ahead and eventually succeed. Otherwise, DO WHAT YOU'VE BEEN TOLD, YOU GEEK!!!
I've been unemployed for 8 months now, and it has been the most rewarding, and the most enlightning 8 months of my IT career. So, real-world experience means something? Yes, you have to get a feel of it for 2-3 years. After that, it makes no stinking difference, not in IT.
Shura
[ May 21, 2002: Message edited by: Shura Balaganov ]

[Mark Herschberg]

Originally posted by Mapraputa Is:
This country's obsession with so-called "real-world" experience is simply insane. The only thing it's good for is to develop inferiority complex in newcomers.

While it may be overly excessive at some companies, I feel it is, generally speaking justified.
I came out of MIT with a BS in EE/CS and a msters degree in EE/CS. I worked under the great Ron Rivest, even! And you know what, I hardly learned a thing in school. MIT helped improve my analytical thinking. In terms of harder skills, allt hat was learned on the job. Even the soft skills, such as communications and team work I really learned on the job. Yeah, yeah, I knew how to write a paper and give talks, etc; I also did plenty of group activities in school. But understanding the "Peopleware" aspect of software, seeing how human error and micommunications lead to issues...you get nothing even close to that in school.
I was one of a cockiest kids coming out of school (at least in my own mind). I'm now one of the
biggest appreciators of experience.
--Mark

[Shura Balaganov]

Originally posted by Mapraputa Is:

Hey, be careful! :roll: I am going to apply and claim seven... no, better eight years of heavy XML and XSLT experience. They said "5 years or more..." so I figured the more, the better? Shurik Balaganov will write me a confirmation letter that I did work as an XSLT developer since 86th day of Martoben 1985, and the fact that his country is presently non-existent will only add appropriate surreal tones to my brilliant resume.

Love it! Let's do it!
Как Саратов поживает?
Shura
[ May 21, 2002: Message edited by: Shura Balaganov ]

[Reid M. Pinchback]

I'm going to chime in, disagree with the vocal majority, and agree with Mark. Folks, I'm sorry you've been interviewed by idiots in past, but I'm afraid you are mixing issues that shouldn't be mixed. Companies with absurd hiring practices are simply places you don't want to work. Focus on the rest, and then ask "in what way does past experience matter for these companies?".
I've had to work on a team to put together H/R practices for a department, have been trained to interview, and have had to do a lot of interviewing. There is a reason why competent interviewers focus on your experience. Just as you've seen companies conduct their hiring inappropriately, those hiring see a load of bullsh*t artists with no ethics and no skills trying to snow their way into a job.
That doesn't mean that interviewers should be expecting "X" years of experience (particularly when the technology in question hasn't been around as long as "X"). It does mean that they should be examining your experience to see if it will tell the hiring manager how you will perform in your new job.
The one and only "predictor" in hiring is past on-the-job behaviour, *NOT* booklearning. Book learning is beneficial for picking up some skills, but it doesn't tell you how somebody will approach their job and interact with their teammates.
If you don't believe that, look at your current idiot manager. How do you think he was hired? By somebody who didn't know to focus on the specifics of how that person conducted his supervisory duties in the previous job. The exec was snowed, and you got stuck with the results.

[Michael Morett]

I thought I'd take a more constructive route and point out to the folks at Merant how their ads might impact perceptions about their company.
Below is the email I sent them.
"Hello,

Briefly, I have no desire for this position. The ad was pointed out to me as a joke. But I thought it might be appropriate to provide feedback and point out that the requirement you mentioned ("10 years Java programming") is both impossible and (in the kindest words possible) ignorant. Java was not invented 10 years ago. You may have meant "10 years overall programming experience in Java and/or C++", but that is not what was written and in no way could it be interpreted that way. I will not even go into the "5 years XSLT" comment or the fact that this person should *also* be an Oracle DBA. Again, in the kindest words possible, this is a display of a fundamental lack of knowledge.

You left a negative impression of your company on the message boards. While this may or may not be important at this time given the job market at the moment, the market will turn around at some point. The laughter/anger/ill will generated can only be a negative to be overcome at a later date when it comes time to attract good candidates for long term careers at Merant.

I have no knowledge of Merant or it's products (or services?). But I can be honest and say even though I am employed, I would not be interested in working at Merant in any capacity, technology related job or otherwise, based solely on reading the job requirements of the position and the inference of the mentality of the management at the company. I would also not recommend Merant to any of my collegues and associates.

If the ad was solely atrributed to a lack of knowledge by HR, than that is of even more concern.

Employees are told never to burn bridges. I would encourage companies to heed the same advice.

In all sincerity, good luck on luring quality candidates to your company.

Michael."
I hope they took it as constructive criticism. I didn't mean it as a flame since I was not particularly upset by their post. We'll see if I get a response...

[Mapraputa Is]

Reid and Michael, you are just two cool guys. You people are what makes this country great. Do not take Map's and Shurik's cynical exchange too serious, if we did not like local practices, why would we be here?
It's just easier to spot funny things when you are a foreigner and weren't brought up in this culture. Having said that, I must admit that American hiring practices is the stupidest thing I saw here. Once I had an hour-long interview about a program, a moderately intelligent person would learn in half an hour. Another interview was wholely composed of questions, carefully collected in this book. They did not ask a single question about my skills! I do not mind these questions, but I cannot understand what makes people to utter these canned questions instead of simply asking their own. Every time I am inteviewed, I feel that I am not talking to a fellow human being, I am talking to programmed automates. You are saying these are not places one wants to work and you are right, of course. I only worry about all these working hours that are lost in asking meaningless questions. People, think about it this way. There is a population of software developers, and we all will be hired in the final account, whether you like it or not. So why don't you simply try to hire a randomly choosen guy, without all these time-consuming formalities? I bet, the result will be the same!

Originally posted by Reid M. Pinchback:
The one and only "predictor" in hiring is past on-the-job behaviour, *NOT* booklearning. Book learning is beneficial for picking up some skills, but it doesn't tell you how somebody will approach their job and interact with their teammates.

Well, since my secret identity was so rudely disclosed, I should note that in the opposite country people have neither resumes nor formal interviews (it is changing lately, though). The hiring process happens at random, I would say, and you know what? I haven't seen too many bad hiring or misplaced people. Most of job aren't that much challenging and randomly chosen guys and girls are perfectly capable of doing them :roll:
Shura, how did you figure out Саратов city, if I carefully removed all signs of my origin from my profile?
[ May 21, 2002: Message edited by: Mapraputa Is ]

[Shura Balaganov]

Originally posted by Mapraputa Is:

Shura, how did you figure out Саратов city, if I carefully removed all signs of my origin from my profile?

Слухами земля полнится....
Маргарита, very clever for you to hide under what looks like Indian name, made me wonder if it's truly so What does "It" stand for?
Didn't mean to be rude. You calling me Shurik gave you away. One can find (if one desires) your profile here:
http://www.javaranch.com/contact.jsp#MapraputaIs
People have accused me of using unappropriate name too, apparently someone leaked info about SB to local folks....I think it wouldn't be right to change it now, since I've been such a pain in the arse for some

Originally posted by Reid M. Pinchback:

That doesn't mean that interviewers should be expecting "X" years of experience (particularly when the technology in question hasn't been around as long as "X"). It does mean that they should be examining your experience to see if it will tell the hiring manager how you will perform in your new job.
The one and only "predictor" in hiring is past on-the-job behaviour, *NOT* booklearning. Book learning is beneficial for picking up some skills, but it doesn't tell you how somebody will approach their job and interact with their teammates.

Very thin ice here. Since there's no predefined resume formats, it's very tough to say how a person would behave at a job. Your HR should be a psychoanalyst. Most HRs try to "profile" you, and since they are not very good at it, they do it wrong a lot. At least, it is in my case. Reid, I wonder what did they teach you in hiring training. Might be useful for everyone to know
Shura
[ May 21, 2002: Message edited by: Shura Balaganov ]

[Mapraputa Is]

Originally posted by Shura Balaganov:
Маргарита, very clever for you to hide under what looks like Indian name, made me wonder if it's truly so What does "It" stand for?

Is? My last name - Isayeva...
Didn't mean to be rude. You calling me Shurik gave you away. One can find (if one desires) your profile here:
Rats. I forgot about this.
People have accused me of using unappropriate name too, apparently someone leaked info about SB to local folks....
So who do you think заложил you to Mark? :roll:

[Reid M. Pinchback]

Originally posted by Shura Balaganov:

Very thin ice here. Since there's no predefined resume formats, it's very tough to say how a person would behave at a job.
...
Reid, I wonder what did they teach you in hiring training. Might be useful for everyone to know

You don't need to be a psychoanalyst. The method is referred to as performance-based interviewing. You (a) decide on the skills - particularly behavioural skills - that you want a candidate to have. You then (b) walk the candidate through the history of something on their resume, like a major project. As they go through the history, you listen for things that either show you evidence of the skill, or you ask open-ended questions that would help you to verify the reality of the experience.
For example, I had somebody in an interview that claimed experience as a teaching assistant in a computer science class. I used to be in an academic support organization helping TAs and faculty. When the candidate couldn't recall the kinds of events and problems I new were typical of the job she had, then I had to doubt her. Either she was snowing me, or she hadn't learned much from the experience. Either way didn't mattter, I simply wasn't able to verify through this and other questions that she would have the kinds of experience and skill necessary to deal with a difficult end-user. Since that skill was important for the job, it was worth continuing the search for a better candidate.
[ May 22, 2002: Message edited by: Reid M. Pinchback ]
[ May 22, 2002: Message edited by: Reid M. Pinchback ]

[Shura Balaganov]

Reid, this is very helpful information, thanks.
I've been kinda employing Mark's "cocky behavior" principle: you have to allow your superstars to behave out of the ordinary.
That is, of course, if someone besides yourself has the same opinion on your abilities. Unfortunately, I don't have best of the best holding my hand.... Maybe it's time to grow up and smell the real world

Originally posted by Mapraputa Is:

So who do you think заложил you to Mark? :roll:

Значит мы квиты
Вы в каком месте в США, если не секрет? Я в районе Филадельфии. And Shura is my real name
Shura
[ May 22, 2002: Message edited by: Shura Balaganov ]

[Mapraputa Is]

This "10 years of Java experience" part makes me think that they in fact ask for something else (if we assume that there *is* something they are looking for). 10 years of experience is a rare requirement, and for reason.
Gerald M. Weinberg in his fascinating "Understanding The Professional Programmer" book states that "after three years, additional years of experience don't seem to add significantly to a programmer's productivity. Managers who hold to this model naturally are unwilling to pay a premium for many years of experience. Usually they will seek programmers in the job market with one or two years of experience."
The graphic presented shows that after 6 years of experience there is no more gain in productivity. (Actually things are more complex, the author shows other, more accurate graphics as well, and makes a lot of wit comments that complicate the statistics).
So when people ask for 10 years of experience, I suspect they actually want something else. More aged = supposedly more mature and more reliable person? Or maybe just an attempt to reduce a crowd of candidates... :roll:
Private Message for Shurik: I am in Portland

[Thomas Paul]

6 years? I would say it depends on what you did for those 6 years?

[Shura Balaganov]

I am guessing, that 10 years might be there because people with that much experience are usually a) experts in technology; b) have done design and architecture; c) managed less senior people (and most likely a number of them). Plus, maturity with age makes them an easier fit.
Naturally, I'd never take a person coming out of big old insurance company (industry I took just for the sake of an argument) to do cutting edge development, although (s)he might be perfect fit to DB and network design, infrastructure, etc.
Looking at it with a clear head, I can see why a lot of companies wouldn't want to take my "wise guy" persona.
Lots of technologies in requirement might serve this purpose of a) eliminating immature people; b) eliminating people who just sat on their pants for 10 years and didn't dare to learn new things?... Just me guessing here
Shura

[Mark Herschberg]

Thanks for the link Mapraputa. I just bought the book.
--Mark

[Shura Balaganov]

Oh, rats. I just read a description of it at amazon.

Beginning the book asking the question "How long does it take to make a programmer?", Weinberg points out that management often isn't knowledgeable and doesn't have the tools to tell the difference between somebody "trained" for six weeks and a journeyman programmer with many years of experience. "The point is not merely that there are people out there passing as professional programmers who shame us all, but that few managers have any way of telling if they're talking to one of them or one of us.

One of them or one of us, huh? So, since I don't think I ever had 2 projects using same technologies, and each project made me learn a new one, am I "one of them"?
[ May 22, 2002: Message edited by: Shura Balaganov ]

[Michael Morett]

I was focusing not on 10 years of programming experience, but rather the 10 years of Java experience requirement.
I can see the interview now....
Their Lead Tech: Michael, I see you have 10 years of Java programming experience. Impressive. How is that possible?
Michael: You guys started this game. I thought I'd play along.

[Mapraputa Is]

This quote is misleading!
Here is my (hoppefully more accurate) attempt to quote (=take words out of context):
"We ought to acknoledge that fifteen years of experience, in and of itself, need not teach you anything about programming. <...>
Just in case the last example is too subtle, let me list a few items I found in one day, reading the programs of "experienced" programmers:
1. Someone who didn't know what a remainder was when using integer division
...
5. Someone who doesn't use subroutines at all because "they're too complicated."
The list could be extended indefinitely. The point is not merely that there are people out there passing as professional programmers who shame us all, but that few managers have any way of telling if they're talking to one of them or one of us."
--------------------
"There are no answers, only cross references."

[Michael Morett]

Subroutines? Are you using GW-BASIC?

[Mapraputa Is]

The book was written more than 20 years ago, I suspect But it keeps its value, like "The Mythical man-month"
Actually, it was written in 1971!
[ May 23, 2002: Message edited by: Mapraputa Is ]

[Rufus BugleWeed]

Gerald M. Weinberg in his fascinating "Understanding The Professional Programmer" book states that "after three years, additional years of experience don't seem to add significantly to a programmer's productivity.

Now, I can say with 100% certainty, Gerald M. Weinberg is a moron.
Sir Edward Deming claims you can't measure the productivity of any employee in any non trivial job.

[Reid M. Pinchback]

That is a suprising quote coming from that source. Deming was the TQM god, right? You measure a process, you know the resources you put into that process, so you know how productive those resources were, at least with respect to the work done in that process. Software development doesn't fit a process model without a lot of careful thought, but it is possible. Measurement isn't easy, but it isn't a completely unsolvable problem either. Hell, there is even a well-regarded certification for one approach to it.

[Mark Herschberg]

Originally posted by Rufus Bugleweed:

Now, I can say with 100% certainty, Gerald M. Weinberg is a moron.
Sir Edward Deming claims you can't measure the productivity of any employee in any non trivial job.

That's extremely silly and defeatist. Of course you can measure productivity. It's not easy. it may not be very accurate, depending on the job and the tools, but you can correctly measure it.
--Mark

[Thomas Paul]

...it may not be very accurate... but you can correctly measure it.
For some reason that statement doesn't parse.

[Mapraputa Is]

Originally posted by Rufus Bugleweed:
Now, I can say with 100% certainty, Gerald M. Weinberg is a moron.

This is partly the Quotation Problem. I had either type the whole chapter, or to stop somewhere. From where I stopped, sir Weinberg continues:
"One problem with this model is that it may fail to consider problem difficulty. In most shops, the more experienced programmers are given, on the average, more difficult program to write. If the measures of productivity fail to consider problem difficulty, the more experienced programmers will naturally have their productivity underestimated."
Etc., etc.
Anyway... To give you more targets to attack here is one more misquote:

Originally posted by Gerald M. Weinberg :
As one programmer expressed it to me, "Why isn't my manager satisfied with a 5 percent increase in my productivity each year? Her productivity isn't increased at all in ten years."
I explained to him that a manager's productivity is measured in a different way - by the productivity of the people she manages. He then asked a very penetrating question: "Why, then, isn't my productivity measured in the same way - by the productivity of the programs I write?" Why indeed?

[Mark Herschberg]

Originally posted by Thomas Paul:
...it may not be very accurate... but you can correctly measure it.
For some reason that statement doesn't parse.

It's parses quite correctly when you apply the true definition of the terms. In this case, correct means when productivity has generally increased, the measurement will reflect this. When it decreases, the measurement will show this. Because the accuracy is not great, it may not register small changes (in some cases, depending on the model some might even argue that a small positive change may result in a negative change). The point is it may not give a very accurate one.
More loosely put, accurate means "generally right direction," even if not accurate down to the degree.
Think back to this discussions in high school science class about precise versus accurate.
Finally, consider Gib's Law (page 59 of Peopleware):

?Anything you need to quantify can be measured in some way that is superior to not measuring at all.? It may not be cheap, or perfect, but it?s better then nothing.

--Mark

[Jim Yingst]

Now, I can say with 100% certainty, Gerald M. Weinberg is a moron.
Actually, Map's quotation troubles stem as much from where she began the quote as where she ended it. Fortunately I have a copy of this book lying around, sent to me by a good friend some time ago. A more complete quotation:

Much typological nonsense has been written on the subject of "programmer burnout." This supposed phenomenon is characterized by a relationship between productivity and experience indicated in Figure 4.
[Fig. 4 omitted]
According to this picture, after three years, additional years of experience don't seem to add significantly to a programmer's productivity. Managers who hold to this model naturally are unwilling to pay a premium for many years of experience. Usually they will seek programmers in the job market with one or two years of experience.

It should be clear the Weinberg does not, himself, believe that this theory is accurate, and in subsequent pages he explains why. Basically it boils down to (a) the theory ignores problem difficulty - more experienced programmers are probably working on more difficult problems, and their apparent productivity is diminished by this difficulty factor. And (b) the theory only describes average behavior of groups of programmers, and ignores that some programmers become vastly more effective over time, while many others do not.
Those are Weinberg's objections, at least. For my own part, I'm sceptical of the studies themselves which claim to show diminished productivity benefits for experience past three years. No details are given on how these studies were conducted, and I can't help thinking that if they were using some lame-ass metric like lines of code, they could well see a lot more apparent productivity among less experienced programmers. So who knows where this all came from anyway?

[Mapraputa Is]

Thank you so much for your corrections, Jim Yingst

Originally posted by Mapraputa Is:
Actually things are more complex, the author shows other, more accurate graphics as well, and makes a lot of wit comments that complicate the statistics).

:roll:
Actually, all I wanted was to tell Mark about this book. I wasn't sure if he already read it, so to avoid insult I decided not to send him E-mail "here is a great book for you" which was my initial intention, but to mention the book in his protectorate, so he could do whatever he wants with it
Fig. 4 nicely fits in my disgust for so-called "real-world" experience, so I decided it's Ok if I quote it I still believe that if we take two programmers who works on similar tasks, one has, say, 6 years of Java coding experience and another... um... 10 years then all the difference will be because of individual abilities, not experience. 10 years of any homogenous experience requirement doesn't make sense to me.

[Mapraputa Is]

Originally posted by Mark Herschberg:
In this case, correct means when productivity has generally increased, the measurement will reflect this. When it decreases, the measurement will show this. Because the accuracy is not great, it may not register small changes (in some cases, depending on the model some might even argue that a small positive change may result in a negative change). The point is it may not give a very accurate one.
More loosely put, accurate means "generally right direction," even if not accurate down to the degree.
Think back to this discussions in high school science class about precise versus accurate.

One of the most amazing discoveries in my life was that there are different kinds of scales, and different "levels of measurability". I miss my books for accurate quotes but this place is giving decent definitions:
"Measurement is the assignment of numbers to objects or events in a systematic fashion. Four levels of measurement scales are commonly distinguished: nominal, ordinal, interval, and ratio."
When we think about measurement, we usually assume ratio scale, where the question "how many times more" makes sense. However, most of thing we "measure" do not lend themselves for this question. Can you say you like one person twice as strong as another? :roll: Time (interval scale) is another good example. Is 4 am twice as much as 2 am?
When we think that programmer's productivity cannot be measured, we assume that it cannot be measured in ratio scale.

[Jim Yingst]

Fig. 4 nicely fits in my disgust for so-called "real-world" experience, so I decided it's Ok if I quote it
Well that's fine. You just shouldn't attribute it to poor Gerald Weinberg, who only described the theory to refute it. It sure is a good thing someone sent me that copy of the book, so I could keep you honest.

[Shura Balaganov]

Ladies and Gentlemen,
I admire your fascination with measurements and accuracy (hey, I have a science degree myself), but the more accurately one tries to measure my performance, the closer I feel to being worked into McDonalds mentality: more burgers - better employee.
The point is that it is not day-to-day tasks that need to be recorded, in my mind. Because everyone has good and bad days. My former boss once admitted (and he used to be hard-core techie himself) that there are days when he can't get done squad. So it is okay to theorise all this onto masses, but applied to a particular case (let's say, me) the best way to do it would be by FINAL PRODUCT.
Because if you come to me everyday with your project plan, I am going to send you out for a nice journey.
I guess that would not make me an ideal candidate for a Project Leader.
Speaking of which... There is this subject of eXtreme Programming... Well, I actually had a chance to use it before I've read about it in books. What we've done was work in pairs on one of the projects (it just came to us naturally ), and I'll tell you it worked like a charm. Productivity and especially Quality was excellent.
Shura
[ May 24, 2002: Message edited by: Shura Balaganov ]

[Reid M. Pinchback]

As odd as it may sound, you don't get that much benefit from measurement if you worry too much about accuracy, particularly at the level of the individual employee.
The primary benefits of measurement are:
1) trying to measure forces you to understand how you do the work you do, instead of just blindly doing it
2) measurement enforces transparency; people can't so easily hide bullshit if their work is measured, particularly if those measurements are published!
3) if your ability to measure is reasonably sound, and your measurements indicate things are going in a negative direction, you know that you need to take action.
4) the measurements sometimes contradict what people believe to be true; we can construct mental models of the work and people around us, that are shown to be unsound when you bother to collect a few facts.
There are some really stupid uses of measurement. At the individual employee level, you have to be very careful about the conclusions you draw from measurement. If an employee isn't producing, is it because they are unwilling, unmotived, unskilled, because their supervisor is forcing them to be unproductive, or because others do work in a way that constrains the productivity of that employee? I've seen many developers do their work in way that makes them look productive, but completely undermines the activities of others. Ask any release engineer or DBA if they've experienced it!
The value of the measuring at the employee level is only to raise a red flag, but not to assume that the employee is the cause. By analogy, if you put a thermometer outside and it reads 96 degrees F, do you blame the thermometer for it being hot? No, but you do use the information to conclude you need to go and turn on the air conditioner.

[Mapraputa Is]

Originally posted by Jim Yingst:
You just shouldn't attribute it to poor Gerald Weinberg, who only described the theory to refute it. It sure is a good thing someone sent me that copy of the book, so I could keep you honest.

If this person could know the future, she would never send you that book
And there is nothing wrong with my quote. I gave it "as is", I didn't change a letter. By the way, Jim, when will you write a book? I want to quote it
[ May 24, 2002: Message edited by: Mapraputa Is ]

[Jim Yingst]

You said

Gerald M. Weinberg in his fascinating "Understanding The Professional Programmer" book states that "after three years, additional years of experience don't seem to add significantly to a programmer's productivity.

That's like me saying that Mapraputa Is says "The one and only "predictor" in hiring is past on-the-job behaviour, *NOT* booklearning." There is nothing wrong with my quote. I gave it "as is", I didn't change a letter.
Back to the topic, whenever I see people trying to measure productivity I think of Wally "coding himself a minivan". I can see where it may be useful on some levels, but attempting to apply it at an individual level can be woefully misguided. Re: Gib's law, I'd argue that "anything is better than nothing" is not necessarily true if the measure is used by less competent managers, who think the measure means something it doesn't. This can lead to inane policies that do more harm than good. If managers understand that the numbers they're looking at are somewhat related to productivity, fine. If they think they are measures of productivity, that's a problem.

[Mark Herschberg]

Originally posted by Mapraputa Is:
Actually, all I wanted was to tell Mark about this book. I wasn't sure if he already read it, so to avoid insult I decided not to send him E-mail "here is a great book for you" which was my initial intention, but to mention the book in his protectorate, so he could do whatever he wants with it

I would never be insulted by someone saying "Mark, you should check out this book, you might find it inetresting." (Unless it's a dummies book :-)
--Mark