Re: Asked a commercial cabinet shop about their SawStop(s)

Joe Bemier wrote:
USA Today likes to print fancy looking charts and graphs
to go with a particular story. However, try to find the confidence
interval or standard deviation to go with the data. The charts are
nice, but unless we can see how the data was prepared and the
variable, it means very little.

Not only that-- the way they scale their charts is almost always
extremely misleading. Say, for example, they've got some stat that's
gone from 1012 up to 1097 over some period of time. That's under 8.5%,
not very impressive really. So they start the Y axis of the damned
chart at 1000 instead of zero and end at 1100. Now the number goes from
near the bottom to near the top, and it looks like whatever they're
reporting has octupled. It drives me crazy.

.

Relevant Pages

  • Re: Non-standard control chart?
    ... control charts for "averages and ranges" and charts for attribute data ... Then do it with the process average somewhat ... that will pass before you detect the average is not aimed to target. ... The formula for thye standard deviation ...
    (sci.stat.consult)
  • Re: OT : Arctic Ice returns
    ... "This week's ice coverage charts, from the people who really know the ... It'd be better if they used a chart that showed the standard deviation along ... which makes claims shown to be patently false by the charts presented. ... To make a global warming or anti global warming case. ...
    (sci.electronics.design)
  • Re: OT : Arctic Ice returns
    ... "This week's ice coverage charts, from the people who really know the ... It'd be better if they used a chart that showed the standard deviation along ... There are lots of charts and data available through the website. ...
    (sci.electronics.design)
  • Re: OT : Arctic Ice returns
    ... "This week's ice coverage charts, from the people who really know the ... It'd be better if they used a chart that showed the standard deviation along ... UN-funded gravy train interested only in another agenda? ... which makes claims shown to be patently false by the charts presented. ...
    (sci.electronics.design)
  • Re: Better control of axis ranges
    ... For some complex charts, I use a process that completely rescales ... in the worksheet to give me ultimate control. ... I start by determining min and max values I'll need on an axis. ...
    (microsoft.public.excel.charting)