Elsmar Cove Forums Index - Lies, Darned Lies, and Statistics

Pages: 1

Lies, Darned Lies, and Statistics
(View the original discussion thread with full colors, attachments and images)

Posted by: Steve Prevette

As a follow-on to last month's "Liars" article, here is a more recent Power Point presentation I developed for the local ASSE chapter. You will see some commonality with last month's paper.

This is an "official" Fluor Hanford document, and yes, they accepted the title, even though this fileserver won't.

Posted by: Jim Howe

Quote:

Originally Posted by Steve Prevette

As a follow-on to last month's "Liars" article, here is a more recent Power Point presentation I developed for the local ASSE chapter. You will see some commonality with last month's paper.

This is an "official" Fluor Hanford document, and yes, they accepted the title, even though this fileserver won't.

Steve, Once I got into the show I could not exit (could not get back to your post), nothing worked to let me out so I had to x-out (go off line) and then come back to leave this message. maybe I did something wrong

As for the show itself it seemed well composed but most shows usually are not complete without the narrator and unfortunately i don't have a sound card on my work computer. Is there a sound track?

Posted by: Steve Prevette

Nope, no sound track. You'll just have to send me an airline ticket

If you compare last month's Liar's document, you'll see some of the thoughts behind this presentation. If anyone is trying to convince their management of the benefits of looking at your data instead of misusing your data, I hope this may be a useful start for you.

It is a power point file, so I would assume it should load up okay and release.

Posted by: Jim Howe

Quote:

Originally Posted by Steve Prevette

Nope, no sound track. You'll just have to send me an airline ticket

Well I wonder what I am doing wrong then. Usually the ESC button will end the show but it had no effect. clicking the mouse sent me back to the beginning and started over. Maybe a glitch for Marc to look at.

i got it Steve, hit the back button!

Posted by: M Greenaway

Good topic.

I try to explain this concept all the time, almost without exception to blank faces and looks of utter bewilderment

Posted by: Steve Prevette

Quote:

Originally Posted by M Greenaway

Good topic.

I try to explain this concept all the time, almost without exception to blank faces and looks of utter bewilderment

It t'ain't easy. After 12 years at it, I finally have most people eating out of my hand, instead of biting my hand . . .

Posted by: Marc

Quote:

Originally Posted by Steve Prevette

...and yes, they accepted the title, even though this fileserver won't.

I'm following the basic "...can my child see this..." 'test' we established for the site, sorry. The file will download as you titled it but 'certain' words in the forums views will not display.... Just trying to keep things 'nice'...

Posted by: Govind

Thanks for sharing this presentation with us. I found this both practical and informative.

I found that some books are published in this subject as well.
How to Lie With Statistics by Darrell Huff, Irving Geis ISBN: 0393310728
The other book is ISBN: 0520219783
I have not read these books far. If any one did, let me know if it is worth buying.
Regards,
Govind.

Posted by: Marc

I enjoyed it as well. Thanks for posting it!

Posted by: Steve Prevette

Quote:

Originally Posted by Marc

Understand. Just having a little fun, and it is such a well-known quote.

Posted by: engjane

I would love to hear the whole thing too...and I totally agree witht he title. Ive even had QA Managers who say that to me when I try and use some stats based tool to show things.

Posted by: Steve Prevette

Quote:

Originally Posted by engjane

I would love to hear the whole thing too...and I totally agree witht he title. Ive even had QA Managers who say that to me when I try and use some stats based tool to show things.

One thing I have come to realize that "what" the numbers are is not so important as "how" you use them. Ethics ought to be a required course for statisticians, and managers ought to have a basic "driver's license" prior to using numbers.