Steve Krivit sent me the following message. He told me he has cut me
off with the "Delete Before Read" feature, so there is no point to
responding to him. If his mysterious friend (Mitchell Swartz, I'll
bet) communicates with him again, I suggest you tell him to buy the
book "How to Lie with Statistics."
Whether he intended to deceive of not, he did use a stupid trick
described in the book. Anyone familiar with this book or with
statistics and mathematical presentations will get the impression
that Krivit either lied or he doesn't know the first thing about how
to present data. So he is not doing himself any favors.
Perhaps someone he has not Cast into Darkness should relay this message him.
- Jed
- - - - - - - - - - - - - - - - -
From: Steve Krivit <[email protected]>
Subject: Rothwell's Character Attack on Krivit
Dear friend,
Thanks for the heads-up about Jed Rothwell's character attack on me
(copied below) on the publicly-accessible Vortex Internet list.
Here is what you need to know:
1.
<http://newenergytimes.com/v2/library/2010/2010KrivitS-ACS.pdf>These
slides are simplified so people can get the larger picture - which
is, in fact, depicted correctly in this slide; the predicted value
was shifted significantly. I simplified the slide because otherwise
people will get lost in the details and they won't get the big
picture. That's why, for precision, I started that sequence on slide
28 with a note at the bottom that says "See New Energy Times Issue 34
for Full Investigation."
2. In the full investigation, slide
<http://newenergytimes.com/v2/news/2010/34/4HeRetention-Slide3-100.shtml>#3
in NET 34 has fully labeled axes, which Jed appears to be complaining
about (but he's actually confused, see #6 below.)
4. However, you don't put error bars on predicted values, only
experimental values. The deception by McKubre and Hagelstein has
occurred through the misrepresentation of the predicted values. It is
the alleged proximity to the predicted values that has been sold in a
misleading way, not the measured values.
5. If Jed looked carefully at slide #30, he would actually find a
real typographical error I made, but that error is inconsequential to
the point of the slide. Plus, Jed has not spent the time to
understand the experiment. I have. All of it.
http://newenergytimes.com/v2/news/2010/34/4HeRetention-Slide8-100.shtml
6. Lastly, it was not experiment M4 that Jed was discussing error
bars and zero lines with me, it was experiment HH.
Therefore, it is Jed Rothwell, not I, who has engaged in dishonest
debate and played tricks. Rothwell's association of my character to
"How to Lie With Statistics" is without basis and slanderous.
I am copying Rothwell on this. If he has any integrity, which I think
he does, he will post this message to the Vortex list (I'm sure with
some explanation for his misunderstandings) and apologize not only to
me, but also to other Vortex readers for misleading them. Please let
me know if he does so.
If Jed wishes to repair the damage he has done to our relationship, I
welcome his phone call.
I will not respond to Vortex comments/discussion on this matter any
further with anyone. I have real work to do.
Steve
At 06:09 AM 3/30/2010, you wrote:
[Vo]:Krivit again uses annoying trick
Jed Rothwell
Wed, 24 Mar 2010 07:03:59 -0700
Here again are the slides I discussed yesterday:
http://newenergytimes.com/v2/library/2010/2010KrivitS-ACS.pdf
A few weeks ago, Krivit discussed here the graph shown on Slide 30.
I pointed out that it should show the zero line, and it should
include error bars. In the presentation Krivit put it back the way
it was, and for good measure he removed the Y-axis labels.
This annoying trick is described on p. 62 of a marvelous little book
by Darrell Huff, "How to Lie with Statistics" (1954, now in its 39th
printing). He describes a graph showing a 10% increase in national income:
