Why Don't People Check Their Work?

[not daSkeptic]

It seems to me that many people fail to check their work, instead assuming it is correct and creating messes for others to clean up. I've seen this a lot in IT, but I've also heard of it happening elsewhere too, so it doesn't appear to be specific to that field. A couple of examples ...

A simple program was written to take values from a small, known set of inputs and map them to another small, known set of outputs. When run, the program emitted values outside of the valid output set. The programmer had not tested his software against the full input set, but had instead just assumed that because a few test cases worked the rest would too. The problem was caught by a third-party who, upon seeing odd output, did the test themselves.

A technician was to change the DNS records for a set of websites, giving them new IP addresses. They instead deleted the records, effectively taking the sites offline. Had they fired up a web browser and verified each site as the changes were made, they would have caught the problem after the first one. At it happened, however, someone outside the organization discovered the issue and phoned it in.

Why do these things happen? Why don't these people take a few extra minutes to verify their work? Are they ignorant of the effect their actions have on others, or do they just not care? Don't they want to do a good job, or do they just want to get paid and go home?

 

[quixotecoyote]

Or perhaps the technician had a really long day with a dozen projects to work on and in trying to juggle more work than he could be expected to do, a to-do item got skipped.

Not saying that's what happened here, but when I get overloaded, the polish comes off my performance.

 

[gumboot]

Not checking your work can be expensive...

We were shooting a scene for a movie once that involved strapping the picture vehicle to a low-loader, pushing it out off the edge of a cliff, and using a digger to violently shake the whole thing, with stuntman inside looking appropriately scared. The camera was out on a jib arm to get the desired angle.

The camera operator didn't check their work, that is, if they'd secured the camera base plate to the head properly. Needless to say, they hadn't.

Fortunately at the last moment the grip decided, on a whim, to throw a safety line around the camera.

Potential cost of not checking that work? $250,000 + 3 hours of footage (conservatively, $20,000).

 

[not daSkeptic]

We were shooting a scene for a movie once that involved strapping the picture vehicle to a low-loader, pushing it out off the edge of a cliff, and using a digger to violently shake the whole thing, with stuntman inside looking appropriately scared. The camera was out on a jib arm to get the desired angle.

The camera operator didn't check their work, that is, if they'd secured the camera base plate to the head properly. Needless to say, they hadn't.

Is this sort of oversight an exception or something that happens regularly? I suppose I should have indicated that I run into this sort of problem at least once-a-week if not more often. Some are just annoying, but others (as you have illustrated) pose a much greater risk. What's worse is that I've actually had people get upset with me for catching these things.

 

[Gord_in_Toronto]

Expecting people to use their brains at all times is a lost cause.

There should steps in the procedures for doing anything important that call for checking and sign-off.

I was impressed when I worked at an engineering firm while in university that every engineering calculation was done twice, by two different people, and then signed off as correct by the second party. If a calculation was not correct, I had to do it over (without knowing the correct result). I don't think very many mistakes were made.

Maybe this concept has been lost in this century.

 

[rwguinn]

Expecting people to use their brains at all times is a lost cause.

There should steps in the procedures for doing anything important that call for checking and sign-off.

I was impressed when I worked at an engineering firm while in university that every engineering calculation was done twice, by two different people, and then signed off as correct by the second party. If a calculation was not correct, I had to do it over (without knowing the correct result). I don't think very many mistakes were made.

Maybe this concept has been lost in this century.

Nope--we still do it that way.
Anything important MUST be checked, independently. Not to say we repeat the work--just that the results are checked.
Often, the individual is too close to the work, and/or has too much invested in it (time, effort, whatever), to notice the problem. That's why, for example, NASA has a checklist, a person performing the checklist, and a 3rd individual making sure that the checklist is followed--a "checklist checker" as it were...

 

[Gord_in_Toronto]

Nope--we still do it that way.
Anything important MUST be checked, independently. Not to say we repeat the work--just that the results are checked.
Often, the individual is too close to the work, and/or has too much invested in it (time, effort, whatever), to notice the problem. That's why, for example, NASA has a checklist, a person performing the checklist, and a 3rd individual making sure that the checklist is followed--a "checklist checker" as it were...

That is really, really gratifying to know. I can venture outside my house again tomorrow.

 

[shadron]

It seems to me that many people fail to check their work, instead assuming it is correct and creating messes for others to clean up. I've seen this a lot in IT, but I've also heard of it happening elsewhere too, so it doesn't appear to be specific to that field. A couple of examples ...

Ahhh, another one of these. "Why does the rest of the world (or whatever segment I'm pissed off at) always let me, in my perfection, down?". It gets to be repetitious. Find and analyze the thread about IT people you were in a week ago, and tell me how many of the submitters, yourself and myself included, admitted to having problems themselves, rather than just all those anonymous "other people". While skeptics may exercise more care in thought than thers, I have doubts about that carrying over into routine job actions.

A simple program was written to take values from a small, known set of inputs and map them to another small, known set of outputs. When run, the program emitted values outside of the valid output set. The programmer had not tested his software against the full input set, but had instead just assumed that because a few test cases worked the rest would too. The problem was caught by a third-party who, upon seeing odd output, did the test themselves.

A technician was to change the DNS records for a set of websites, giving them new IP addresses. They instead deleted the records, effectively taking the sites offline. Had they fired up a web browser and verified each site as the changes were made, they would have caught the problem after the first one. At it happened, however, someone outside the organization discovered the issue and phoned it in.

Not wanting to intrude too far ino your rant, but what was your position with respect to these two particular problems? Middle management? If you are the supervisor, then I would think that allowing a third party to find the problem would indict you as well as the tech. Is not management responsible for the work product as well?

Why do these things happen? Why don't these people take a few extra minutes to verify their work? Are they ignorant of the effect their actions have on others, or do they just not care? Don't they want to do a good job, or do they just want to get paid and go home?

They happen because people are careless and lazy by inclination - almost all of us, almost all the ime. They also happen, as Quixote implies, because people like to go home on time, especially on Friday afternoons. They happen because occasionally they are preoccupied by a sick wife or child. And it happens because they want to get back to reading the latest entries in the thread after their last submission. Some of these are culpable, and some less so. What have you done, in your official position, to award the successful compleion of tasks, the maintenance of focus and carefulness? Or do you just apply the stick?

Given any thought to doing some retraining, perhaps?

 

[shadron]

Nope--we still do it that way.
Anything important MUST be checked, independently. Not to say we repeat the work--just that the results are checked.
Often, the individual is too close to the work, and/or has too much invested in it (time, effort, whatever), to notice the problem. That's why, for example, NASA has a checklist, a person performing the checklist, and a 3rd individual making sure that the checklist is followed--a "checklist checker" as it were...

Often, also, is the time constraint that a common worker has that he has no control over. The manpowering on this project allows you to change no more than 10.5 jhours of your time to it this week. So, if the recheck will extend to 12 hours, whatcha gonna do? Steal from the next project? Work unpaid overtime? Some do that, and I think it's mostly the young, unjaded, least experienced workers who do best at that. That's why man-rated hammers cost NASA $5K, but when the customer is the local hardware store, who obviously isn't going to pay $5, then where do you cut corners?

Real life adventure in economics: you go to your local hardware store a demand that your hammer be inspected as you speak about above. The hardware guy says, sure, he can do that. If you get everyone in the United States to do the same, they can be had for a mere $25 each. But if you're the only one, it'll be $5K. Plus tax. What're the chances?

It's true that important jobs require more care - all you have to do is identify them, and fund them. Going to the lowest bidder (ie, good old primary capitalism) runs counter.

 

[gumboot]

Is this sort of oversight an exception or something that happens regularly?

Usually not on that level of repercussion, but these sorts of things are relatively common. The film industry is very fast paced and high pressure, and most of the crew are specialists so everyone is left to do their own task - there's no oversight from someone else to check your work.

People in difficult situations, under pressure, make mistakes.

War seems to be another place where such mistakes are common, for the same reasons.

 

[Kaylee]

I’ve worked in two different fields (accounting and computer programming) and for several different companies.

I agree with rwguinn. I also think that being able to recheck one’s own work isn’t a skill that everyone has. Add lack of skill to the understaffing prevalent at many companies and you have a situation where if someone made a mistake the first time, they are more likely to make the same mistake the second time when they rush through their review.

Every group develops its own “culture” and expectations. I’ve found that most accountants, especially those that work in financial reporting, usually don’t make significant mistakes. Most places have no tolerance for those types of mistakes, and you either learn quickly not to make significant mistakes or you get out of the field. Auditors? (CPAs) It depends on the company they work for. Some have a no tolerance for errors culture. Others have a culture of “you must beat the prior year’s team time at all cost so we can make a bigger profit this year”--- that often leads to errors.

The manager often sets the tone also. Competition is good, but sometimes it can be overdone. You know it’s overdone when people don’t tell their co-workers about mistakes they happened to notice so that they can be fixed. They just sit back and hope that their co-workers get into trouble instead.

A weird sense of pride can be responsible for needless errors also. As a computer programmer I had worked for a small company owned by 3 partners. One partner specifically told all the programmers to never update or install anything on a production machine unless they were with another co-worker to serve as another pair of eyes. Well, people did that for the first week or so. But then that stopped. A small group of people wanted to present themselves as infallible and no one called them on it. Then that group grew because others starting to feel that they were presenting themselves in a bad light if they followed the partner’s directive and asked for a second pair of eyes. Crazy but true.

So … there’s a lot of reason why mistakes don’t get caught. Give me more time and I can probably remember other reasons.

 

[gumboot]

I know when I'm writing I can't check my own work. If I read a spelling mistake or grammatical error by brain automatically "corrects" it in my head as I read, so I don't detect it. Sometimes people just need someone else to check their work for them.

 

[patchbunny]

I find that after working on the same thing for a while, you simply stop seeing them. That's why I appreciate the outside review processes we have to go through before submitting projects for bid.

Doesn't make getting a drawing back covered in red pen any less embarrassing, though.

 

[tesscaline]

I know when I'm writing I can't check my own work. If I read a spelling mistake or grammatical error by brain automatically "corrects" it in my head as I read, so I don't detect it. Sometimes people just need someone else to check their work for them.

The technique I use to "proofread" my own writing is that once I've finished the draft, I set it aside for a minimum of 5 minutes (I prefer setting it aside for 10-15, but sometimes a deadline might interfere), and look at something completely different during those 5 minutes (say, a very short work task or going to get a fresh cup of coffee, or even read a forum thread if there isn't anything else to do at that moment). Only after that do I read over what I wrote. It tends to "reset" the thought processes and I catch the majority of my errors that way.

Of course, I come from the financial background that Kaylee referred to as having a "zero tolerance for errors" culture (and rightly so -- in my line of work, even small mistakes can get you fired), so the idea of having someone else find an error in my work is mortifying to me. It's almost an obsession for me that I have to have my work as error free as humanly possible before anyone else lays eyes on it. Unfortunately it can be both a blessing and a curse to be that way -- the blessing being that your bosses and co-workers adore you for not making more work for them, the curse being that ultimately you get to suffer the stress that you're sparing them

 

[tyr_13]

I tend to use the, 'read it backwards' method. You read each sentence, starting with the last sentence and going back.

Obviously, I don't do that for forum posts, as my terrible grammar and spelling shows. But this isn't work.

 

[Wangler]

Where I work, a GLP research laboratory, things have to be right.

That is why our software undergoes a rigorous validation and verification before use...even Excel calculation spreadsheets.

Everything requires one-over-one quality control, from data entry to manual calculations.

That is not to say mistakes are never made, but at least they are fully documented!

Also, our V&V process would likely have caught the two problems mentioned in the OP.

 

[Seanette]

I know when I'm writing I can't check my own work. If I read a spelling mistake or grammatical error by brain automatically "corrects" it in my head as I read, so I don't detect it. Sometimes people just need someone else to check their work for them.

I've caught myself doing that a few times, and it's pretty routine for Bob (which is why he has me proof-read important items like applications. In fact, just this "morning", he'd typo'ed an important word on an application he was about to submit. Good thing he did have me check it).

 

[DarthFishy]

It's amazing how easily mistakes can slip into work and how difficult it can sometimes be to catch it. I work as a lecturer at a University so part of my job is supervising 4th year (or Honours as they are called here) students in writing up their final dissertation.

I was working through one students work, studiously making corrections and comments using the track changes function in word. This student had a lot of mistakes in terms of spelling, grammar and formatting but eventual I though we found most of them and gotten his paper into a good shape.

After he had submitted I was looking for a specific source for a conference paper I wanted to write and I remembered a reference this student had quoted in his dissertation. I eventually found the quotation but imagine my mortification when that author was not been referenced in his bibliography. It had been quoted but not referenced or mentioned at all.

Worst part was I had totally missed this during all the hours I had spent correcting his work. I suppose I assumed he would know by now (especially after doing a course especially focusing on HOW to do research and reference etc.) how to do this. Obviously I was wrong

 

[rjh01]

My job for the last two years has been to check other people's work. The work involved is comparing a report written in Cognos (a report generator) with expected results.

Unless the report is only very small, I can find a major error in the work. Often I find several. This is a job that cannot be taught.

One of the problems is that the work may be done correctly as the person understands it, however they do not understand what is required.

 

[not daSkeptic]

Also, our V&V process would likely have caught the two problems mentioned in the OP.

My company's processes did catch the problems. That's how they were discovered. Unfortunately, a different company did the actual work, so we had no real say in their quality control. They were a collaborator with whom we shared a mutual client, so the most we could do is complain to that common party.