🚀 go-pugleaf

This is a multi-part message in MIME format.

------=_NextPart_000_0310_01C799A0.8C287FF0
Content-Type: text/plain;
	charset="utf-8"
Content-Transfer-Encoding: quoted-printable

I found this discussion very interesting. Make up your own mind:         =


http://scienceblogs.com/deltoid/2007/04/a_new_flavour_of_global_warmin.ph=
p

=20

Basically, Sinclair Davidson, a Professor of Economics, and Alex Robson, =
a Lecturer in Economics, wrote an article, part of which was published =
in the Australian Financial Review, criticising the Intergovernmental =
Panel on Climate Change (IPCC) and including this statement:

=E2=80=9CAs an elementary textbook of statistics reminds its readers, =
with large data sets, confidence intervals have to be increased, so a 90 =
per cent confidence level would not then be valid -- the hypothesis is =
falsified.=E2=80=9D

Blogger Tim Lambert dissects their article and dismisses the above =
claim:=20

=20

"No, elementary textbooks of statistics do not say that. Mainly because =
it's not true. A larger data set means that  smaller effects will be =
statistically significant, so it is possibly that an effect could be =
statistically significant but so small that it is not practically =
significant. Furthermore, if a test does not achieve a specified =
confidence level, it does not falsify the hypothesis -- all you can say =
is that there is insufficient data to draw a conclusion."

=20

concluding

=20

"It is worrying that economists at Australia's universities have such a =
poor understanding of basic statistics."

=20

The economists respond, and some of their arguments are remarkable.

=20

Firstly, they repeatedly cite Lindley's paradox as a reason for doubting =
the IPCC=E2=80=99s conclusions, e.g.:

=20

"The IPCC claim to have a large data set, and if they do then Lindley's =
paradox might be a problem. Now do they have a lerge data set or don't =
they? If they do then Lindleys paradox is a problem, if they don't and =
Lindley isn't a problem then they don't have a large sample."

=20

but Wikipedia states that Lindley's paradox only applies to hypothesis =
testing problems in cases where

=20

" the prior distribution is the sum of a sharp peak at H0 with =
probability p and a broad distribution with the rest of the probability =
1-p. It is a result of the prior having a sharp feature at H0 and no =
sharp features anywhere else."

=20

=20

They also hammer the point that the IPCC=E2=80=99s Summary for =
Policymakers does not explain exactly how they arrived at their figure =
of  >90% for the assessed likelihood that most of the observed increase =
in globally averaged temperatures since the mid-20th century is due to =
the increase in anthropogenic greenhouse gas concentrations.=20

=20

"So the basic question we raise still remains unanswered: =E2=80=A6 =
where does their 90% figure come from?"

"The point is though, we don't know how data the IPCC have because they =
don't reveal how they come to their 90% estimate."

"How is the 90% value derived? If they were transparent as they are =
required to be, then we wouldn't be having this debate."

"They have reported a number (indeed several numbers) and all I'm asking =
for is an explanation of where that number came from."

=20

However, when asked "whom you attempted to contact that is affiliated =
with the IPCC that was unwilling to answer your question(s)", the reply =
was:

=20

"I am unable to contact the IPCC directly because I don't have ethics =
approval to collect data from human subjects." !!

=20

=20

Finally, challenged to name one elementary textbook of statistics which =
proves their assertion that:

=E2=80=9CAs an elementary textbook of statistics reminds its readers, =
with large data sets, confidence intervals have to be increased, so a 90 =
per cent confidence level would not then be valid -- the hypothesis is =
falsified.=E2=80=9D,

Professor Davidson tries to dodge the question with truisms like: "I =
suspect all of them will have section on hypothesis testing. Most =
chapters include a discussion of confidence levels, Type I and Type II =
errors and the like", then changes the subject.

=20

This exchange follows. (I=E2=80=99ve extracted it from all the other =
debate going on in the thread):

=20

Lambert:  "If your statement about stats texts is true, it would be easy =
for you to provide just one example. Why can't you do this? Is it =
because your claim is a fabrication?"

=20

Davidson:  "Fabrication in what sense?"

=20

Lambert:  "Fabrication in the sense that elementary texts do not say =
that "with large data sets, confidence intervals have to be increased" =
but rather you and Alex Robson made that part up. It is trivial to prove =
that you didn't -- just cite one elementary text that says it. How about =
it?"

=20

Davidson:  "That's a very agressive stance you're taking Tim. You're =
leaving no wiggle room for yourself. I'm wondering what you're going to =
do when I name a book?

=20

I'll give you a clue. While some hardcover versions did come out, it's a =
big fat blue paperback with a spine that cracks easily."

=20

Lambert:  "Sinclair, the normal practice when citing a book is to give =
the author and title. I'm afraid that "big fat blue paperback" isn't a =
very useful way of citing."

=20

Davidson:  "Tim, the book is Judge etal. Intro to practice of =
econometrics (title from memory) it's out of print now. Another is by =
Zellner that's an intro to Baysian stats (probably of out of print too). =
Leamer wrote about this as well in one of his books (probably not an =
intro, from memory)."

Lambert:  "Thanks for finally providing a proper cite, Sinclair. The =
UNSW library has a copy of "Introduction to the Theory and Practice of =
Econometrics" (2nd Edition) by Judge et al. In section 3.5 Hypothesis =
Testing I find (page 98):

As a final note on the power of a test, note that for a given level of =
significance of the test, =CE=B1, as the sample size increases (to =
infinity) the power of the test increases (to one) for every =CE=B2 =
=E2=89=A0 1. The intuition for this is based in the fact that the test =
is constructed around a consistent estimator. As the sample size =
increases, the sampling distribution of the estimator collapses about =
the true parameter value, and thus the probability that we would reject =
a false hypothesis goes to unity. Tests with this property are called =
consistent tests.

Judge provides no support (in over a thousand pages) for Sinclair's =
claim:

As an elementary textbook of statistics reminds its readers, with large =
data sets, confidence intervals have to be increased, so a 90 per cent =
confidence level would not then be valid -- the hypothesis is falsified.

I also went to the UNSW book store and looked at all the elementary =
textbooks on statistics and econometrics. None of them said what =
Sinclair claimed they did. None of them.

Sinclair, I think you need to send a correction for your article to IPA =
Review."

Davidson:  "Tim, eye on the ball. I also provided Zellner and Leamer as =
refs. I have others too, there is a nice example in Walpole, =
Introduction to Statistics (but he doesn't expand on it), and Rao. But =
unless you want to argue I fabricated Lindleys paradox (I'm sure I would =
have called it 'Davidson's paradox'), and the formulas that underlie =
hypothesis testing (especially the bit about dividing through by the =
square root of n) I'm not sure where this leaves your argument about the =
IPCC."

=20

Robson and Davidson then reply with quotes from stats textbooks, =
claiming that these quotes support their position, as indeed they appear =
to do. However, Lambert then shows that the quotes have been edited. Try =
reading the quotes below, skipping the sections in bold, then read them =
again with the bold text included.

Lambert:  "Elementary stats texts do not say that you have to increase =
significance levels with large samples and they must definitely do not =
say that you should reject the alternative hypothesis if the specified =
level of significance is not achieved. Alex and Sinclair have now =
resorted to quote doctoring to try to make it appear that they do. In =
the quotes below I have added in bold the words they left out from their =
quotes.

Sinclair's quote of Judge et al

"For a given significance level the probability of a Type I error is =
held constant whatever the sample size, and the probability of a Type II =
error declines as T increases. This situation favors H0 when T is small =
and favors H1 when T is large. In contrast, the Bayesian approach treats =
both errors symmetrically and favors neither hypothesis as T increases. =
For the sampling theory effect to be achieved within the Bayesian =
framework, we would need to make the prior odds ration a decreasing =
function of sample size. Similarly, for the Bayesian effect to be =
achieved within the sampling theory framework, the significance level =
would need to be a decreasing function of sample size".

Judge is contrasting Bayesian and frequentist hypothesis testing. He is =
not, as Sinclair tried to make you think with his quote doctoring, =
advocating that the significance level needs to decrease with bigger =
samples. Compare the original with Sinclair's version and ask yourself =
if he was being honest.

Next we have Alex's quote of Kennedy. Bolded are the bits he left out =
without even an ellipsis:

"One of the more interesting problems is the fact that almost any =
parameter can be found to be significantly different from zero if the =
sample size is sufficiently large. (Almost every relevant independent =
variable will have some influence, however small, on a dependent =
variable; increasing the sample size will reduce the variance and =
eventually make this influence statistically significant.) Thus, =
although a researcher wants a large sample size to generate more =
accurate estimates, too large a sample size might cause difficulties in =
interpreting the usual tests of significance. McCloskey and Ziliak =
(1996) look carefully at a large number of empirical studies in =
economics and conclude that researchers seem not to appreciate that =
statistical significance does not imply economic significance. One must =
ask if the magnitude of the coefficient in question is large enough for =
its explanatory variable to have a meaningful (as opposed to =
"significant") influence on the dependent variable. This is called the =
too-large sample size problem. It is suggested that the significance =
level be adjusted downwards as the sample size grows; for a =
formalization see Leamer (1978). See also Attfield (1982) Leamer would =
also argue that this problem would be resolved if researchers recognized =
that genuinely interesting hypotheses are neighbourhoods, not points.

He says (in the bit Alex omitted) "that statistical significance does =
not imply economic significance". But Sinclair claims that Kennedy is =
not writing about "Practical v statistical significance". Note further =
that he states the problem does not apply if the hypothesis is a =
neighbourhood, which is, of course, the case here. (Hypothesis being =
"most of recent warming is man-made".)=20

Last, we have Alex's quote of Maddala. Again, bold is the bit he left =
out.

The usual procedure that is suggested (which is called the =
Neyman-Pearson approach) is to fix =CE=B1 at a certain level and =
minimize =CE=B2, that is choose the test statistic that has the most =
power. In practice, the tests we use, such as the t, =CF=872, and the F =
tests.

There have been some statisticians who disagree with the ideas of the =
Neyman-Pearson theory. ... There are some statisticians who think that =
the significance level used should depend on the sample size.

"The problem with a preassigned significance level is that if the sample =
size is large enough, we can reject every null hypothesis. This is often =
the experience of those who use large cross-sectional data sets with =
thousands of observations. Almost every coefficient is significant at =
the 5% level. Lindley ["A Statistical Paradox", Biometrika, 1957: =
187-192] argues that for large samples one should use lower significance =
levels and for smaller samples higher significance levels. Leamer =
["Specifiation Searches: Ad Hoc Inference with Non-Experimental Data", =
New York: Wiley, 1978] derives significance levels in the case of =
regression models for different sample sizes that show how significance =
levels should be much higher than 5% for small sample sizes, and much =
lower for large sample sizes."

Alex made it appear that Maddala said that you should use lower =
significance levels for large samples when, in fact, he said that the =
usual procedure was to use fixed levels.

To summarize, none of the text cited by Alex and Sinclair supported =
their claim that:

As an elementary textbook of statistics reminds its readers, with large =
data sets, confidence intervals have to be increased, so a 90 per cent =
confidence level would not then be valid -- the hypothesis is falsified.

So Alex and Sinclair resorted to quote doctoring to make it look like =
that they did."

Davidson:  "Tim, quote doctoring is a serious allegation. Can you make =
an argument for how our quoting changes the meaning of the original =
texts, or have you just found that we quoted the important aspects of =
the text? Afterall I quoted nothing from the previous 305 pages of Judge =
etal."

Lambert:  "I already explained how you changed the meaning. Here's waht =
I wrote about your doctored quote from Judge:

Judge is contrasting Bayesian and frequentist hypothesis testing. He is =
not, as Sinclair tried to make you think with his quote doctoring, =
advocating that the significance level needs to decrease with bigger =
samples. Compare the original with Sinclair's version and ask yourself =
if he was being honest.

Do you have any explanation to offer for your misconduct?"

Davidson:  "Now you laying a charge of academic misconduct? Please, go =
ahead.=20

So what have we got? Fabrication and misconduct. I think you've been =
shown up Tim. You challenged me and the best you can come up with is =
this? Vague allegations of misconduct? Tsk, tsk."

Lambert (to someone else):  "I'm hoping he can explain why he thinks =
that was honest quoting, but he keeps being evasive."

Davidson:  "Tim, I still waiting for amplification of the academic =
misconduct charges you've laid against me. You wouldn't want to deny me =
natural justice, now would you? Facts, please."

=20

Lambert shows again how sections were left out of quotes so as to change =
their meaning, then:

 "I have repeatedly asked you to explain yourself. You have not done so. =
Yet again, do have any explanation to offer for your actions?"

Davidson:  "Tim, I stand by my interpretation of the literature. Do you =
stand by your allegation of academic misconduct?"

Lambert:  "Sinclair, you don't have any evidence. Which is why you =
fabricated it by doctoring a quote from Judge. You haven't offered any =
defence at all against this charge. I'll let the jury decide whether you =
are guilty or not."

It goes on, with the professor refusing to admit that his selective =
quoting of references altered their meaning, saying things like:  "That =
is a very serious allegation of misconduct that you keep making", and =
"That's on page 5, Tim. For the record I'm doctoring this quote by not =
quoting the 4 pages before, or the the rest of page 5, or the remaining =
12 or so pages" , but I think that=E2=80=99s enough for you to get the =
picture.

=20

You have to wonder whether John Howard is getting advice on climate =
change from people like this.



------=_NextPart_000_0310_01C799A0.8C287FF0
Content-Type: text/html;
	charset="utf-8"
Content-Transfer-Encoding: quoted-printable

=EF=BB=BF<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML xmlns:o =3D "urn:schemas-microsoft-com:office:office"><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; charset=3Dutf-8">
<META content=3D"MSHTML 6.00.2800.1400" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=3D#ffffff><FONT face=3DArial><FONT size=3D2>
<DIV>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">I found this discussion =
very=20
interesting. Make up your own mind:<SPAN=20
style=3D"mso-tab-count: =
1">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
</SPAN><o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">http://scienceblogs.com/deltoid/2007/04/a_new_flavour_of_global_wa=
rmin.php<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Basically, Sinclair =
Davidson, a=20
Professor of Economics, and Alex Robson, a Lecturer in Economics, wrote =
an=20
article, part of which was published in the Australian Financial Review, =

criticising the Intergovernmental Panel on Climate Change (IPCC) and =
including=20
this statement:<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 36pt; mso-pagination: none; mso-layout-grid-align: =
none"><FONT=20
face=3D"Times New Roman"><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">=E2=80=9CAs an =
elementary textbook of=20
statistics reminds its readers, with large data sets, confidence =
intervals have=20
to be increased, so a 90 per cent confidence level would not then be =
valid --=20
the hypothesis is falsified.=E2=80=9D</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Blogger Tim Lambert =
dissects their=20
article and dismisses the above claim: <o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT face=3D"Times New Roman">"No, elementary =
textbooks=20
of statistics do not say that. Mainly because it's not true. A larger =
data set=20
means that<SPAN style=3D"mso-spacerun: yes">&nbsp; </SPAN>smaller =
effects will be=20
statistically significant, so it is possibly that an effect could be=20
statistically significant but so small that it is not practically =
significant.=20
Furthermore, if a test does not achieve a specified confidence level, it =
does=20
not falsify the hypothesis -- all you can say is that there is =
insufficient data=20
to draw a conclusion."<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">concluding<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT=20
face=3D"Times New Roman">&nbsp;<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT face=3D"Times New Roman">"It is worrying =
that=20
economists at Australia's universities have such a poor understanding of =
basic=20
statistics."<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT=20
face=3D"Times New Roman">&nbsp;<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">The economists respond, =
and some of=20
their arguments are remarkable.<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Firstly, they repeatedly =
cite=20
Lindley's paradox as a reason for doubting the IPCC=E2=80=99s =
conclusions,=20
e.g.:<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT=20
face=3D"Times New Roman">&nbsp;<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT face=3D"Times New Roman">"The IPCC claim =
to have a=20
large data set, and if they do then Lindley's paradox might be a =
problem. Now do=20
they have a lerge data set or don't they? If they do then Lindleys =
paradox is a=20
problem, if they don't and Lindley isn't a problem then they don't have =
a large=20
sample."<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT=20
face=3D"Times New Roman">&nbsp;<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">but Wikipedia states that =
Lindley's=20
paradox only applies to</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial; mso-ansi-language: EN"> =
<I><SPAN=20
lang=3DEN>hypothesis testing </SPAN></I><SPAN =
lang=3DEN>problems</SPAN></SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial"> in cases=20
where<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT=20
face=3D"Times New Roman">&nbsp;<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT face=3D"Times New Roman">" the prior =
distribution is=20
the sum of a sharp peak at H0 with probability p and a broad =
distribution with=20
the rest of the probability 1-p. It is a result of the prior having a =
sharp=20
feature at H0 and no sharp features anywhere =
else."<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT=20
face=3D"Times New Roman">&nbsp;<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">They also hammer the point =
that the=20
IPCC=E2=80=99s </SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial; mso-ansi-language: =
EN">Summary for=20
Policymakers does not explain exactly how they arrived at their figure =
of<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN>&gt;90% for the assessed =
likelihood that=20
most of the observed increase in globally averaged temperatures since =
the=20
mid-20th century is due to the increase in anthropogenic greenhouse gas=20
concentrations. <o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT=20
face=3D"Times New Roman">&nbsp;<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">So the basic question =
we raise=20
still remains unanswered: =E2=80=A6 where does their 90% figure come =
from?</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">The point is though, we =
don't=20
know how data the IPCC have because they don't reveal how they come to =
their 90%=20
estimate.</SPAN><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">How is the 90% value =
derived? If=20
they were transparent as they are required to be, then we wouldn't be =
having=20
this debate.</SPAN><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">They have reported a =
number=20
(indeed several numbers) and all I'm asking for is an explanation of =
where that=20
number came from."<o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">&nbsp;<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">However, when =
asked</SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt; =
mso-ansi-language: EN">=20
</SPAN><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">whom you attempted to =
contact=20
that is affiliated with the IPCC that was unwilling to answer your=20
question(s)</SPAN><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN></FONT><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">,</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial; mso-ansi-language: EN"> =
</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">the reply was:</SPAN><SPAN =

style=3D"FONT-SIZE: 10pt"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">&nbsp;<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT face=3D"Times New Roman">"I am unable to =
contact the=20
IPCC directly because I don't have ethics approval to collect data from =
human=20
subjects." !!<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Finally, challenged to =
name one=20
elementary textbook of statistics which proves their assertion=20
that:<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 36pt; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">=E2=80=9CAs an elementary textbook of =
statistics reminds its=20
readers, with large data sets, confidence intervals have to be =
increased, so a=20
90 per cent confidence level would not then be valid -- the hypothesis =
is=20
falsified.=E2=80=9D</FONT></SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial; mso-ansi-language: =
EN">,</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Professor Davidson tries =
to dodge=20
the question with truisms like: </SPAN><SPAN style=3D"FONT-SIZE: =
10pt"><FONT=20
face=3D"Times New Roman">"I suspect all of them will have section on =
hypothesis=20
testing. Most chapters include a discussion of confidence levels, Type I =
and=20
Type II errors and the like", </FONT></SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">then changes the=20
subject.<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">This exchange follows. =
(I=E2=80=99ve=20
extracted it from all the other debate going on in the=20
thread):<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Lambert:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN></SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"If your =
statement about=20
stats texts is true, it would be easy for you to provide just one =
example. Why=20
can't you do this? Is it because your claim is a =
fabrication?"</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Davidson:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN></SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"Fabrication in =
what=20
sense?"</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Lambert:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN></SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"Fabrication in =
the sense=20
that elementary texts do not say that "with large data sets, confidence=20
intervals have to be increased" but rather you and Alex Robson made that =
part=20
up. It is trivial to prove that you didn't -- just cite one elementary =
text that=20
says it. How about it?"</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Davidson:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN></SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT face=3D"Times New Roman">"That's a very =
agressive=20
stance you're taking Tim. You're leaving no wiggle room for yourself. =
I'm=20
wondering what you're going to do when I name a=20
book?<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT=20
face=3D"Times New Roman">&nbsp;<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">I'll give you a =
clue. While=20
some hardcover versions did come out, it's a big fat blue paperback with =
a spine=20
that cracks easily."</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Lambert</SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN>"Sinclair, the normal practice =
when=20
citing a book is to give the author and title. I'm afraid that "big fat =
blue=20
paperback" isn't a very useful way of citing."</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Davidson:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN></SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"Tim, the book =
is Judge=20
etal. Intro to practice of econometrics (title from memory) it's out of =
print=20
now. Another is by Zellner that's an intro to Baysian stats (probably of =
out of=20
print too). Leamer wrote about this as well in one of his books =
(probably not an=20
intro, from memory)."</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Lambert</SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN>"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">Thanks for finally =
providing a=20
proper cite, Sinclair. The UNSW library has a copy of "Introduction to =
the=20
Theory and Practice of Econometrics" (2nd Edition) by Judge et al. In =
section=20
3.5 Hypothesis Testing I find (page 98):<o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 36pt; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">As a final note on the power of a test, note =
that for a=20
given level of significance of the test, =CE=B1, as the sample size =
increases (to=20
infinity) the power of the test increases (to one) for every =CE=B2 =
=E2=89=A0 1. The=20
intuition for this is based in the fact that the test is constructed =
around a=20
consistent estimator. As the sample size increases, the sampling =
distribution of=20
the estimator collapses about the true parameter value, and thus the =
probability=20
that we would reject a false hypothesis goes to unity. Tests with this =
property=20
are called <I>consistent tests</I>.<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">Judge provides no support (in over a thousand =
pages) for=20
Sinclair's claim:<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 36pt; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">As an elementary textbook of statistics reminds =
its=20
readers, with large data sets, confidence intervals have to be =
increased, so a=20
90 per cent confidence level would not then be valid -- the hypothesis =
is=20
falsified.<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">I also went to the UNSW book store and looked =
at all the=20
elementary textbooks on statistics and econometrics. None of them said =
what=20
Sinclair claimed they did. None of them.<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><FONT=20
face=3D"Times New Roman"><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">Sinclair, I think you =
need to=20
send a correction for your article to IPA Review.</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Davidson:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN></SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">Tim, eye on the ball. I =
also=20
provided Zellner and Leamer as refs. I have others too, there is a nice =
example=20
in Walpole, Introduction to Statistics (but he doesn't expand on it), =
and Rao.=20
But unless you want to argue I fabricated Lindleys paradox (I'm sure I =
would=20
have called it 'Davidson's paradox'), and the formulas that underlie =
hypothesis=20
testing (especially the bit about dividing through by the square root of =
n) I'm=20
not sure where this leaves your argument about the IPCC.</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Robson and Davidson then =
reply with=20
quotes from stats textbooks, claiming that these quotes support their =
position,=20
as indeed they appear to do. However, Lambert then shows that <I>the =
quotes have=20
been edited</I>. Try reading the quotes below, skipping the sections in =
bold,=20
then read them again with the bold text included.<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Lambert</SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN>"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">Elementary stats texts =
do not say=20
that you have to increase significance levels with large samples and =
they must=20
definitely do not say that you should reject the alternative hypothesis =
if the=20
specified level of significance is not achieved. Alex and Sinclair have =
now=20
resorted to quote doctoring to try to make it appear that they do. In =
the quotes=20
below I have added in bold the words they left out from their=20
quotes.<o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">Sinclair's quote of Judge et=20
al<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 36pt; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">"For a given significance level the probability =
of a Type=20
I error is held constant whatever the sample size, and the probability =
of a Type=20
II error declines as T increases. This situation favors H0 when T is =
small and=20
favors H1 when T is large. <B>In contrast, the Bayesian approach treats =
both=20
errors symmetrically and favors neither hypothesis as T increases. For =
the=20
sampling theory effect to be achieved within the Bayesian framework, we =
would=20
need to make the prior odds ration a decreasing function of sample size. =

Similarly, for the Bayesian effect to be achieved within the sampling =
theory=20
framework,</B> the significance level would need to be a decreasing =
function of=20
sample size".<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">Judge is contrasting Bayesian and frequentist =
hypothesis=20
testing. He is not, as Sinclair tried to make you think with his quote=20
doctoring, advocating that the significance level needs to decrease with =
bigger=20
samples. Compare the original with Sinclair's version and ask yourself =
if he was=20
being honest.<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">Next we have Alex's quote of Kennedy. Bolded =
are the bits=20
he left out without even an ellipsis:<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 36pt; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">"One of the more interesting problems is the =
fact that=20
almost any parameter can be found to be significantly different from =
zero if the=20
sample size is sufficiently large. <B>(Almost every relevant independent =

variable will have some influence, however small, on a dependent =
variable;=20
increasing the sample size will reduce the variance and eventually make =
this=20
influence statistically significant.)</B> Thus, although a researcher =
wants a=20
large sample size to generate more accurate estimates, too large a =
sample size=20
might cause difficulties in interpreting the usual tests of =
significance.=20
<B>McCloskey and Ziliak (1996) look carefully at a large number of =
empirical=20
studies in economics and conclude that researchers seem not to =
appreciate that=20
statistical significance does not imply economic significance. One must =
ask if=20
the magnitude of the coefficient in question is large enough for its =
explanatory=20
variable to have a meaningful (as opposed to "significant") influence on =
the=20
dependent variable.</B> This is called the too-large sample size =
problem. It is=20
suggested that the significance level be adjusted downwards as the =
sample size=20
grows<B>; for a formalization see Leamer (1978). See also Attfield =
(1982) Leamer=20
would also argue that this problem would be resolved if researchers =
recognized=20
that genuinely interesting hypotheses are neighbourhoods, not=20
points.</B><o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">He says (in the bit Alex omitted) "that =
statistical=20
significance does not imply economic significance". But Sinclair claims =
that=20
Kennedy is not writing about "Practical v statistical significance". =
Note=20
further that he states the problem does not apply if the hypothesis is a =

neighbourhood, which is, of course, the case here. (Hypothesis being =
"most of=20
recent warming is man-made".) <o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">Last, we have Alex's quote of Maddala. Again, =
bold is the=20
bit he left out.<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 36pt; mso-pagination: none; mso-layout-grid-align: =
none"><FONT=20
face=3D"Times New Roman"><B><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">The usual procedure =
that is=20
suggested (which is called the Neyman-Pearson approach) is to fix =CE=B1 =
at a certain=20
level and minimize =CE=B2, that is choose the test statistic that has =
the most power.=20
In practice, the tests we use, such as the <I>t</I>, =CF=87<SUP>2</SUP>, =
and the=20
<I>F</I> tests.</SPAN></B><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 36pt; mso-pagination: none; mso-layout-grid-align: =
none"><FONT=20
face=3D"Times New Roman"><B><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">There have been some=20
statisticians who disagree with the ideas of the Neyman-Pearson theory. =
...=20
There are some statisticians who think that the significance level used =
should=20
depend on the sample size.</SPAN></B><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 36pt; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">"The problem with a preassigned significance =
level is=20
that if the sample size is large enough, we can reject every null =
hypothesis.=20
This is often the experience of those who use large cross-sectional data =
sets=20
with thousands of observations. Almost every coefficient is significant =
at the=20
5% level. Lindley ["A Statistical Paradox", Biometrika, 1957: 187-192] =
argues=20
that for large samples one should use lower significance levels and for =
smaller=20
samples higher significance levels. Leamer ["Specifiation Searches: Ad =
Hoc=20
Inference with Non-Experimental Data", New York: Wiley, 1978] derives=20
significance levels in the case of regression models for different =
sample sizes=20
that show how significance levels should be much higher than 5% for =
small sample=20
sizes, and much lower for large sample =
sizes."<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">Alex made it appear that Maddala said that you =
should use=20
lower significance levels for large samples when, in fact, he said that =
the=20
usual procedure was to use fixed levels.<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">To summarize, none of the text cited by Alex =
and Sinclair=20
supported their claim that:<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 36pt; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">As an elementary textbook of statistics reminds =
its=20
readers, with large data sets, confidence intervals have to be =
increased, so a=20
90 per cent confidence level would not then be valid -- the hypothesis =
is=20
falsified.<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><FONT=20
face=3D"Times New Roman"><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">So Alex and Sinclair =
resorted to=20
quote doctoring to make it look like that they did.</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Davidson:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN></SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">Tim, quote doctoring is =
a serious=20
allegation. Can you make an argument for how our quoting changes the =
meaning of=20
the original texts, or have you just found that we quoted the important =
aspects=20
of the text? Afterall I quoted nothing from the previous 305 pages of =
Judge=20
etal.</SPAN><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Lambert</SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN>"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">I already explained how =
you=20
changed the meaning. Here's waht I wrote about your doctored quote from=20
Judge:<o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 36pt; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">Judge is contrasting Bayesian and frequentist =
hypothesis=20
testing. He is not, as Sinclair tried to make you think with his quote=20
doctoring, advocating that the significance level needs to decrease with =
bigger=20
samples. Compare the original with Sinclair's version and ask yourself =
if he was=20
being honest.<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><FONT=20
face=3D"Times New Roman"><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">Do you have any =
explanation to=20
offer for your misconduct?</SPAN><SPAN style=3D"FONT-SIZE: =
10pt">"</SPAN><SPAN=20
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Davidson:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN></SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">Now you laying a charge =
of=20
academic misconduct? Please, go ahead. <o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><FONT=20
face=3D"Times New Roman"><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">So what have we got? =
Fabrication=20
<I>and</I> misconduct. I think you've been shown up Tim. You challenged =
me and=20
the best you can come up with is this? Vague allegations of misconduct? =
Tsk,=20
tsk.</SPAN><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Lambert (to someone=20
else)</SPAN><FONT face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: =
10pt">:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN>"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">I'm hoping he can =
explain why he=20
thinks that was honest quoting, but he keeps being evasive.</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Davidson:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN></SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">Tim, I still waiting =
for=20
amplification of the academic misconduct charges you've laid against me. =
You=20
wouldn't want to deny me natural justice, now would you? Facts,=20
please.</SPAN><SPAN style=3D"FONT-SIZE: =
10pt">"<o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><FONT=20
face=3D"Times New Roman">&nbsp;<o:p></o:p></FONT></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Lambert</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt"><FONT face=3D"Times New Roman"> =
</FONT></SPAN><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial; =
mso-ansi-language: EN">shows=20
again how sections were left out of quotes so as to change their =
meaning,=20
then:<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><FONT=20
face=3D"Times New Roman"><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><SPAN=20
style=3D"mso-spacerun: yes">&nbsp;</SPAN></SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">I have repeatedly asked =
you to=20
explain yourself. You have not done so. Yet again, do have any =
explanation to=20
offer for your actions?</SPAN><SPAN style=3D"FONT-SIZE: =
10pt">"</SPAN><SPAN=20
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 0cm 0cm 0pt; mso-pagination: none; =
mso-layout-grid-align: none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Davidson:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN></SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">Tim, I stand by my =
interpretation=20
of the literature. Do you stand by your allegation of academic=20
misconduct?</SPAN><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Lambert</SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN>"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">Sinclair, you don't =
have any=20
evidence. Which is why you fabricated it by doctoring a quote from =
Judge. You=20
haven't offered any defence at all against this charge. I'll let the =
jury decide=20
whether you are guilty or not.</SPAN><SPAN style=3D"FONT-SIZE: =
10pt">"</SPAN><SPAN=20
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></FONT></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial; =
mso-ansi-language: EN">It=20
goes on, with the professor </SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">refusing to admit that his =
selective=20
quoting of references altered their meaning, saying things like:<SPAN=20
style=3D"mso-spacerun: yes">&nbsp; </SPAN></SPAN><FONT=20
face=3D"Times New Roman"><SPAN style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">That is a very serious =
allegation=20
of misconduct that you keep making</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">, </SPAN></FONT><SPAN =
lang=3DEN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial; mso-ansi-language: =
EN">and</SPAN><FONT=20
face=3D"Times New Roman"><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"> </SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt">"</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN">That's on page 5, Tim. =
For the=20
record I'm doctoring this quote by not quoting the 4 pages before, or =
the the=20
rest of page 5, or the remaining 12 or so pages</SPAN><SPAN=20
style=3D"FONT-SIZE: 10pt">" </SPAN></FONT><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">, but I think =
that=E2=80=99s enough for you=20
to get the picture.</SPAN><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
Arial">&nbsp;<o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><SPAN=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">You have to wonder whether =
John=20
Howard is getting advice on climate change from people like =
this.</SPAN><SPAN=20
lang=3DEN style=3D"FONT-SIZE: 10pt; mso-ansi-language: =
EN"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal=20
style=3D"MARGIN: 5pt 0cm; mso-pagination: none; mso-layout-grid-align: =
none"><FONT=20
face=3D"Times New Roman"><SPAN lang=3DEN=20
style=3D"FONT-SIZE: 10pt; mso-ansi-language: EN"><SPAN=20
style=3D"mso-spacerun: =
yes"></SPAN></SPAN></FONT>&nbsp;</P></DIV></FONT></FONT></BODY></HTML>

------=_NextPart_000_0310_01C799A0.8C287FF0--