February 5, 2004 at 5:35 am
Can't you buy special support when having Windows 2000 Datacenter Edition that guarantees 99,9%+ availability. Think I read this somewhere.
--
Frank Kalis
Microsoft SQL Server MVP
Webmaster: http://www.insidesql.org/blogs
My blog: http://www.insidesql.org/blogs/frankkalis/[/url]
February 5, 2004 at 5:35 am
Interesting comments. I must say we face the same issues in the medical profession. That's why you have to be alert and read the fine print on any of our journal articles. And frankly, what makes it so difficult to sort out the truth.
But may I suggest a balance: statistics are not completely manipulatable (is this in the dictionary?). But those who are proficient with them and understand how they were derived can easily "spin" them one way or another and someone who either doesn't understand them or doesn't take the time to carefully scrutinize exactly what is said or written can be easily misled. It comes down to diligence and discipline, IMHO.
In my observations, almost all medical reporting (CNN, networks, etc) is rarely if ever accurate because the reporters don't understand the data they are reporting. This can actually be dangerous and on more than one occasion detrimental to the public.
SMK
February 5, 2004 at 5:46 am
Statistics and military intelligence share a characteristic that they should be used as as a predictive tool or indicator to say what is likely to happen.
They shouldn't be used to justify something in retrospect.
February 5, 2004 at 6:06 am
But may I suggest a balance: statistics are not completely manipulatable (is this in the dictionary?). But those who are proficient with them and understand how they were derived can easily "spin" them one way or another and someone who either doesn't understand them or doesn't take the time to carefully scrutinize exactly what is said or written can be easily misled. It comes down to diligence and discipline, IMHO.
Of course, statistics are not that manipulatable. The processes and formulas are given and cannot be changed. That would be far too obvious and foolish. But the final result strongly depend on your assumptions and the raw data you've collected and your intention when analyzing and interpreting that data.
David hit the nail right on the head:
Lies, damn lies and statistics.
--
Frank Kalis
Microsoft SQL Server MVP
Webmaster: http://www.insidesql.org/blogs
My blog: http://www.insidesql.org/blogs/frankkalis/[/url]
February 5, 2004 at 6:42 am
Statistics don't lie, but liars use statistics --Dr. Harrington
-- J.T.
"I may not always know what I'm talking about, and you may not either."
February 5, 2004 at 9:39 am
A little off topic, but it was funny when I heard it. Glad you guys like it.
February 10, 2004 at 10:27 am
statistically speaking, i don't think your article had the intended affect. after reading it i felt like i should mistrust all statistics because they are all slanted in some way.
i am of the opinion one should be able to account for these unforseeable data hickups in the fact that most statistics come with a scientific degree of error. i guess most people don't realize that fact. statistics are and can be a scientifically calculated. it is the degree of error that will always be inquestion. that is where a structured environment comes into play. if you have an environment that documents down time to a server, then the degree of error becomes smaller.
the one thing i would leave readers with is the fact that one would definitely have to question statistics that seem drastic. to say that all system changes have a 50% success rate is a very drastic measure. that is failing in most school systems. i would be curious to see the degree of error for that number as well as the process used to derive that number. but if someone were to tell me 99.99%, well i would assume that's close enough to 100% for me and the degree of error wouldn't much mater.
Doug Guerena
Viewing 7 posts - 16 through 21 (of 21 total)
You must be logged in to reply to this topic. Login to reply