Official science: People do less, make more mistakes on Friday afternoons Boffins have spent two years monitoring the computers of office staff at a large Texas energy concern and found that workers did less and made more mistakes in the afternoon – particularly on Fridays. The study, published in the journal PLOS ONE, comes from authors Taehyun Roh, Chukwuemeka Esomonu, Joseph Hendricks, Anisha …
"found that workers did less and made more mistakes in the afternoon – particularly on Fridays"
Ah HAA ! Yes ! I see it : false advertising !
The title leads the reader to conclude that it is on Friday afternoons that more mistakes are made. That is not the conclusion of the boffins.
Clickbait ! Clickbait !
At our place, where I never go anyway, I've noticed upgrades, deployments and other big changes are often booked in for Friday evenings, the logic apparently being that if it does go pear shaped it can be rolled back with the minimum of disruption to the work week, ie either it works or it's rolled back before Monday morning. I assume the relevant people doing the job are pre-booked into Saturday and maybe even Sunday on a provisional basis, ie they are there and expecting to work and the "bonus" is going home as soon as the job is done, hopefully without issues and still the weekend, or most of it, off.
For those (younger? persons) who don't know or never worked out why Friday was called poets' day:
_Piss _Off _Early _Tomorrow's _Saturday.
This fine piece of research confirms that there are productivity gains in having your workforce taking an early mark and absconding to the pub on Friday afternoons.
Besides you can never get a pew in a CBD watering hole much after 4.30pm. Even standing room is at premium. Certainly I am been forced to enjoy a few overpriced ales on the Pitt Street pavement served out of a street window. Summer afternoons in Sydney is excellent drinking weather :)
As the study was based in Texas I doubt the Friday afternoon result was due to lunchtime pub sessions ( regular almost compulsory event in 2970s London).
I think the afternoon flagging is due to the excessive hours gullible mercans get conned into.
You can only expect about 4 hours of high quality work from a techie per day long term. Conning them into 10 hour days only makes more work for the bug reporting software.
@James Anderson
"4 hours of high quality work" from a techie?
You serious? So much? Citation needed.
I sacked* all my "techies" years ago when I realised it was cheaper to change to an O/S that didn't need tech support as it seems the o/s I chose just does what it is supposed to do. With (so far) no problems.
*When I say "sacked" I really mean paid them off, big time.And we all still go out a couple of times a week for a booze up
My US colleagues tend to follow a pattern of having whichever poor soul is on-call being the one to do CMs/patching/etc after work (~5pm their time) on Friday evening.
Given that they know our culture (hic!) and possibly more so that we can't mess with some servers during their working day, we tend to do our round of the same thing on Saturday mornings. People have commented on if it's wise doing such things with a hangover, but on the main occasion I recall making a boo-boo I hadn't been out the night before.
I was gonna say, I'm very confused by all of this.
5 on a Friday is THE deployment time for anything big. Yeah, it means my weekend is fucked. But it means the next week I'm taking Tuesday, Wednesday, and maybe Thursday off if things went well over the weekend, and it means I've got until Sunday evening to roll everything back if it didn't work.
And of course I take a nap Friday afternoon before doing anything big. If something interrupts my nap, maybe updates get delayed a week.
The one where your phone rings Friday lunchtime (when you're in the pub, of course) and you're told the week's invoicing has fallen over. The second week it happened, after sorting out the mess, the rest of the afternoon was spend going through the source code - 3rd party application but not quite enough source to compile provided - to find out just what they'd done to make the database engine blow its memory allocation.
At my main work, in Spain the Friday afternoon is officially off.
Now we just have to expand it to UK, IE, FR, NL, BE to have everybody happy.
(if it did work for having a day off for your birthday - it started in SP and was expanded Europe wide - no reason it couldn't be done for this too)
The authors themselves state "other activities not observed in this study could significantly impact overall productivity. For example, the reduction in computer use might be caused by engaging in different types of activities, such as participating in meetings or planning on Friday".
So the results are only relevant to physical computer use, and the study is also weak because "[t]he study included participants from various computer-oriented, office-based positions, such as admins, geologist, accountants, and engineers, within the company. Although demographic and other individual factors were not collected, these characteristics were assumed to be distributed evenly, and the effects of outliers would be minimal, due to the large sample size" -- a somewhat suspect assumption. A distinction should have been made between performance of routine processes on computers and the use of computers in support of intellectual tasks. Indeed the words 'thought' and 'think' are absent from the paper.
Hey, at least they stated their assumptions.
I recently read a paper on how exercise modes affect blood pressure that was reported in the the press akin to "These two exercises could reduce your dangerously high blood pressure" where they made the assumption that the control groups in the papers they were doing a meta analysis of were sufficiently similar to allow all the studies to be linked. Of course they (1) didn't actually state the implicit assumption or (2) provide any evidence that the assumption held thus rendering the paper little more than prettily worded bunkum and the conclusions worthless.
Unlike this paper, which at most is going to be used to justify a bit of Friday afternoon bunking off, that paper is going to possibly genuinely and seriously affect the health of people who get an exercise prescription from a medic who (a) didn't read more that the conclusions or a report of them, (b) isn't smart enough to spot the terrible methodological mistake.
Almost the entire literature on blood pressure and health is contaminated. Study after study is either poorly designed, or, worse yet, cleverly designed by a pharmaceutical company to correlate a measurable change in a physical sign with "good outcomes." The medication does push the sign in the right direction, to be sure: that can happen only in patients with less underlying morbidity. Then, the "positive" study results come about from unmasking patients with less severe underlying causes, who where going to have better outcomes anyway. The medications themselves have hardly any benefit, and often quite the contrary.
If a lot of these medications actually worked, the "number needed to treat" would be much lower.
Years ago, the Lancet bravely pointed out that really good treatments only needed small studies--that massive studies were powered to show clinically insubstantial benefits--and that they were, ipso facto, unethical. It was brilliant, and of course, ignored.
...or made at any time on any day of the week.
With a few exceptions - some Jags and Range Rovers - BL cars were crap. The only good thing to say about BL's "cars" is they were slightly less water soluble than those from Fiat/Lancia/Alfa-Romeo - or a spoonful of instant coffee.
Was spent rewriting a template for a change about 4 times, every time I thought it was done I found another bit to add to the list. To be fair it is a template for upgrading a site and we want to have one template that can be adapted for the 40 building we need to do 4 times a year. So it was worth it.
As the finished document with 1300 words with the rewrites and the other documents I was working on must have typed about 8-10000 words yesterday…..
No I don’t know how many typos I made only that at the end the documents made sense.
I sometimes think I am a writer with a sideline as a senior network security engineer and not the other way round….
I was employed to work helping to build a hospital, and almost all of the people I worked with were Irish, everyone took lunch-time off at the local pub and our lunches were normally only about 6 to 10 pints. Once we all got back and started climbing up the outside of the building (there were only ladders, no lifts) we were a little slow but always very careful after lunch, Monday to Friday every week. No problems at all.
I guess you now know how I started hearing Brendan Behan ... "I saw a notice which said, 'Drink Canada Dry' and I've just started." ... LOL, a lovely memory, drinking lunch with all those Irish guys.
If we changed to all working Monday to Thursday, would this result in all the errors being moved to Thursdays?
Or if 20% of us got any one day off, would that spread the errors more evenly across the week?
Even before I had (semi) retired, I had announced that I will not be working Mondays or Fridays. Now, part-time, I am less worried about my errors!
The last time I was working for somebody else, Friday afternoon was often used to catch up on documentation, what I managed to get done that week with a prioritized list of things that need to get done the next week. I didn't want to get stuck into working on a design that I wouldn't get done and have to pick up days later. Monday morning was an all-hands meeting and conference call and I didn't want to try and do all of my notes that morning. I could spend Monday morning before the gathering to go through my email and get out some calls before people on the East coast of the US were off to lunch.
When I interviewed at a company that I would have loved to work with, they had what was called "5 o'clock math". Any serious calculations done at the end of the day had to be reexamined the next working morning for errors. The management preferred that the last 1/2 hour or so was used for tidying up and making sure to-do lists for the next day were ready to go. It was another company on the West coast of the US near Seattle and it could be important to get calls done first thing.
The only stuff that gets worked on for IT on Fridays are helpdesk type requests or emergency situations that need to be fixed ASAP. We don't change the network, do any work on server upgrades or patches, new server deployments, etc.
Basically, it is a do not do any work on a Friday that you might be required to fix or undo if it goes sidesways over the weekend policy.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
