Slashdot Mirror
PCs Use More Sick Days Than People
lunarscape writes "ZDNet is running an article about the 'absentee' rate of PCs in various UK workplaces. According to the article, while the average employee was out sick seven days a year, the average PC was inoperable due to a virus nine days a year. The article also discusses junk e-mail's impact on productivity, with one business reporting that 99.84 percent of all incoming mail is spam."
The moderators really need a -1: Didn't Get The Cultural Reference option. I believe the parent was refering to a Dilbert cartoon (Dilbert used to "40% fact" to alarm the Pointy-Haired Boss.)
Anonymous Kev
Proudly posting as AC since 1997
(Finally got a dang account in 2004)
Vbug, a Microsoft developer support company based in the UK with just six employees, received around 720,000 e-mails messages in a month, 99.84 percent of which were spam.
No its the figure for one company for one month.
Do not try to read the dupe, thats impossible. Instead, only try to realize the truth
What truth?
There is no dupe
There is nothing in the article that says the survey was for PCs in workplaces.
It just says "A survey of 2,500 UK e-mail users found that 70 percent of users had been infected by a virus in the past year." It then relates that to average UK worker sick days. Nothing says the PC's were in the workplace.
Which of course makes MUCH more sense. If the average PC atany workplace I know of was down for 9 days a year heads would roll. That's insane. Average PCs at my company are down maybe a fraction of a percent due to viruses because there are professionals making sure it stays that way.
So this article is basically "70% of random HOME users were infected in a year."
Businesses seem to have been asked only about spam.
Doesn't seem like news at all.
Operator, give me the number for 911!
Not really. 2500 is plenty people; what's more important is that the sample is representative (of whatever group you want to talk about). As long as your sample is representative then 250,000 people will not give you significantly better results than 2500.
Igor Presnyakov stole my hat
That's a bogus way to conduct a poll. By definition, you are only getting data from people who go to that site.
It's called a "self-selecting sample" and in statistics it's a no-no.
2,500 randomly selected sample points will give very accurate results, and in fact a lot of poll-takers would be envious of such a large sample.
Apparently even the poster didn't RTFA - the article states:
Our corporate workstations were affected significantly enough by virii last year to be down a total of less than a single day each. Still more downtime than we'd like, but nothing like nine days. Now spam - that's another kettle of fish altogether...
"Never attribute to malice that which is adequately explained by stupidity." - Hanlon's Razor
Contrary to popular belief, the sample sizze required for a given level of statistical precision is NOT some big percentage of the population if sampling is random. Think of having a medical test on a blood sample. Since blood is well mixed, small amounts drawn from anywhere in your body are representative of the whole amount. There's no need to take 20 or 30% of your blood, nor to spike you all over your anatomy, they just need enough to work with. Thank goodness!
Not only did you not get the joke, the mods didn't either.
Real statistics show that people are ~ 1.5 times more likely to call in sick on Monday and ~ 2 times more likely to call in sick on Friday, as opposed to Tuesday-Thursday.
That figures out to roughly:
Monday: 23.5%
Tuesday: 15.5%
Wednesday: 15.5%
Thursday: 15.5%
Friday: 30%
Casual Games/Downloads
Actually it's exactly 40% (for businesses operating on the standard M-F business week) and it is a proven fact. You just have to be careful about how the fact is stated so that it's clear you're taking the percentage of M,F from M,Tu,W,Th,F.
"Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
And for anyone who's curious, this 7 days v. 9 days statistic seems to be significant by chi-squared analysis (yes, I did waste the time to check)
I'll introduce you to the 405
Been there done that. However, Seattle has a 405 too, and in my experience, just as bad or worse.
You forget that sick days and vacation days are separate.
----- Question authority, but not ours. Hate the man, but we're not him.
That sounds great, but as a desktop support drone at a major company (40,000 people) I can tell you it doesn't always work out that way. Here's what really happens:
The user has a 4 year old CPx laptop the company won't replace because it doesn't have the budget (unless you're a director or higher). The OS gets fried from spyware, adware, viruses, etc. All the spares are ancient systems too.
Backups and restores take longer because no one seems to keep files on any network servers. The 350 MB limit on storage space doesn't help. Those 4 year old hard drives sometimes fail, and we don't have a budget for data recovery, so it's up to desktop support to make a best effort.
Systems are often out of warranty, so we have to scavange parts from reclaimed systems. If we can't fix a system, then we have to order another reclaim in from the warehouse, which takes at least a day.
We have a couple loaners, but they won't have the user's files or custom programs. If it's not too busy we can transfer data from the old hard drive pretty quickly. But if there are a lot of tickets, take a number and expect to wait a few hours.
All that adds up to a lot of downtime over a year. I had one poor guy who went through 3 laptops in 2 weeks because of this. By then I just gave him a loaner to keep since we where getting DOA reclaims.
Oh, and as for viruses, our team here is pretty good, but we did have one virus where we had to go desk-to-desk with a patch CD. Some people where down for the entire day.
Absence per year:
At work:
~250 Linux systems: 1-2 hrs/yr
~20 Solaris systems: 1-2 hrs/yr
~25 Windows systems: 2 day/yr?
~10 Macs: 2-3 hrs/yr
Then again, we have serious firewalls, and bought a Barracuda spam/virus filter. The Linux downtime is almost all hardware-related (old, dying PS, cheap memory - yes, we're getting away from these). Mac downtime is mostly hardware, and one flaky OS9 app.
At home:
2 Linux systems: 1 day/yr
2 Win systems: 1 day/yr
Good firewalls, only the Linux systems have internet access. Linux systems are always on, Windows are on mostly when used, so guesstimate is for lost time. Down time for Linux systems is mostly trying something weird or adding hardware. About half the Windows downtime is for that.
I also have a production Linux server at a colo. It's been up 499 days, and was down for maybe 2 hours the previous year. So 1 hr/yr.
I have a good firewall for this system, too.
Lessons? Even Windows systems can show up *if* you have a secure environment and educated, trustworthy users. We have, just today, though, implemented a "no IE" policy. And without Windows, life is even easier.
(For the record, TCO/system at our site, and my house, is *much* lower for the non-Windows systems. 8^)
They're like Slashdot mod points. Use 'em or lose 'em.
Think about it. His sig is self-referential.
Want to improve your Karma? Instead of "Post Anonymously", try the "Post Humously" option.