sampling methods would not be suitable for real-time
applications, since there is no bound on how many points may be
rejected before one is accepted, and hence no bound on the time
required to generate a random normal variate.
Radford Neal
coefficients obtained from this
model (if one could fit it, which one can't, of course) could well be
quite different from those obtained on the whole data set.
Radford Neal
Radford M. Neal
result (after
finding the distribution of the test statistic, and calculating
p-values on that basis) is no different from the usual t test.
Perhaps the same is the case for a two-sample test, which would
explain why no one talks about the possibility of doing it this way.
Radford Neal
-values if the null is true will be close to
correct when using the pooled estimate (apart from the use of a normal
approximation, etc.)
Radford Neal
Radford M. Neal [EMAIL PROTECTED
that
methods which are invalid in this situaton (where the residuals are
not independent of the predictors) must also be invalid in other
situations where the predictors are stochastic, even when they are
independent of the residuals.
Radford Neal
Radford Neal:
Coefficients:
Estimate Std. Error t value Pr(|t|)
(Intercept) 15.510240.62466 24.83 2e-16 ***
x0.408630.01898 21.52 2e-16 ***
---
Signif. codes: 0 `***' 0.001 `**' 0.01 `*' 0.05 `.' 0.1 ` ' 1
Vadim and Oxana
, but they may perhaps form a stationary process, in
which case adjustment for autocorrelation will work fine.)
Radford Neal
Radford M. Neal [EMAIL PROTECTED]
Dept. of Statistics and Dept
Radford Neal [EMAIL PROTECTED]:
In any case, the original poster explicitly claimed that regression
with an explanatory variable that was generated by a non-stationary
process was invalid even if the residuals of the regression are
independent. I claim that this is not true.
David B [EMAIL
a claim that is somewhat like this
in some reference book, and have mis-interpreted it.
Radford Neal
Radford M. Neal [EMAIL PROTECTED]
Dept. of Statistics and Dept. of Computer
that the model will not make good predictions for new data from
the same source? If so, I think you are wrong. Or are you saying
that you won't be able to make conclusions about causal influences?
That might well be, but for that, it's not really just a matter of
fixed versus stochastic.
Radford Neal
fit a regression
model, it will be obvious that the residuals are autocorrelated, and
you need to adjust for this in doing your significance test.
Radford Neal
Radford M. Neal [EMAIL
it might be.
Radford Neal
=
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=
intervals. You can't
interpret them that way, though, if you're abiding by the orthodox
frequentist philosophy.
Radford Neal
Radford M. Neal [EMAIL PROTECTED]
Dept. of Statistics and Dept
observation. But a confidence interval is NOT a probability
statement concerning the unknown parameter. In the frequentist
statistical framework in which confidence intervals exists,
probability statements about unknown parameters are not considered to
be meaningful.
Radford Neal
before and a treatment/control indicator as explanatory variables, and
judging the effectiveness of the treatment by looking at the coefficient for
the treatment/control variable. Or is the actual proposal something else?
Radford Neal
In article [EMAIL PROTECTED],
Dennis Roberts [EMAIL PROTECTED] wrote:
the basic idea is to be able to explain the post score variance in terms
of something ELSE ... that is, for example ... we know that some of the
variance in pain is due to one's TOLERANCE for PAIN ... thus, if we can
remove
ne is to have such a statistic, and will realize
that one shouldn't expect that to happen all the time. (Well, the
really good ones might realize all this.)
Radford Neal
=
Instructions for joining and leaving this list
ulty based on their
actual achievements, but rather on the basis of some estimate of their
ability, disregarding "random factors". That's an interesting
opinion, but would a policy of paying based on actual achievement (or
a noisy estimate of actual achievement) constitute discriminati
The reason why one should NOT do a significance test on this data, at
any level, and regardless of how much power the test would have, was
explained by me a while ago in the post I have repeated below.
If you think there is something wrong with my reasoning, I suggest you
explain the flaw.
Ra
Radford Neal wrote:
... the median is probably not
the best thing to look at, whatever you might have been taught. What
to look at depends not just on the shape of the distribution, but also
on what your purpose is. Ask yourself whether there are very many
purposes for which it would make
hypothesis is true, which it isn't.
Radford Neal
Radford M. Neal [EMAIL PROTECTED]
Dept. of Statistics and Dept. of Computer Science [EMAIL PROTECTED]
University of Toronto
* measure of the strength of evidence against the null.
Use something else if you like. But if you're going to use p-values,
what could you possibly think you are doing if you *aren't* going to
interpret the result as being stronger evidence against the null the
smaller the p-value is?
Radford
subtle patterns to stock prices that haven't
been noticed by the many thousands of people looking for such
patterns, discovering them will take more subtle tools that ARIMA.
Radford Neal
Radford M. Neal
Responding to my posting below:
I concluded that one would expect the division to be *closer* to
1/2 each than for the clearly marked ballots. If you mix a certain
number of valid ballots with 1/3 - 2/3 proportions with another group
of invalid ballots, in which votes have been
In article [EMAIL PROTECTED],
Thom Baguley [EMAIL PROTECTED] wrote:
I admit to being confused by Radford's analysis. Radford, is your point that
if there are a large proportion of "empty/no vote" ballots then we would
expect them to be decided close to 50:50 and hence shift the balance from
In article [EMAIL PROTECTED],
Robert J. MacG. Dawson [EMAIL PROTECTED] wrote:
I think what Radford Neal is saying is that *some* dimpled ballots were
dimpled by a process other than an attempt to vote, and these will
divide 50-50. This should be easily determined by comparing the number
In article 90dpba$47r$[EMAIL PROTECTED],
Gene Gallagher [EMAIL PROTECTED] wrote:
Hengartner's simple analysis shows that after the machine recount in
Broward County, the percentage of Bush to Gore votes was 31% to 69%.
The additional votes, after the hand recount and assignment of some
of the
In article [EMAIL PROTECTED],
[EMAIL PROTECTED] (Radford Neal) wrote:
Of course, there surely must have been some legitimate votes missed in
the machine recount. But there also surely must be some "votes" being
counted that do not reflect the actual intent of the voter, and it
s
"Type III" errors seems to me to be quite on
a par with Type I and Type II errors.
Radford Neal
=
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES
statement
was meant to imply that support for Gore *causes* illiteracy, which
would be the only reason to take offense. The reverse causation, or
some common cause, would be the more reasonable explanation for an
association of this so
I may not be the only one confused on what these confidence intervals
mean. In the above press release, the Gallup organization provides this
description of what their +/- 2% means:
"For results based on the total sample of likely voters, one can say
with 95% confidence that the margin of
d be responsible for ensuring that the grades are fair.
If that person is you, then do it. If it's not you, then ask the
person whose responsibility it is. If someone is supposed to be
responsible, but is not doing their job, then you have a difficult
problem tha
ect, but the benefit if it does
might be so large that we'll use it on the off-chance that it does
have an effect. But if you're using the black-and-white language of
"significant" versus "not significant", it makes no sense to say that
an effect is "imp
Thom Baguley [EMAIL PROTECTED] wrote:
You can get important significant effects, unimportant significant
effects, important non-significant effects and unimportant
non-significant effects.
Radford Neal wrote:
I'll go for three out of four of these. But "important non-significant
ef
Radford Neal wrote:
I presume that the people making such models are interested in whether
or not the poor or good performance of a country might be due to
controllable factors such as organization, training facilities, etc.
In other words, they want to know if they could be doing better
o it makes very little
difference if a student whose true ability is slightly above the
cut-off scores slightly below the cut-off on the test. (Assuming, of
course, that the cut-off is set at the right level.)
Radford Neal
=
In
In article [EMAIL PROTECTED],
Radford Neal [EMAIL PROTECTED] wrote:
The following is a question posed by a faculty member in the social
sciences. He has done a logistic regression finding near significance.
But his Betas are zero. What does this mean? Is it just that the accuracy
is only
degrees of
freedom times the variance of the measurements divided by n-1. From
this fact it's possible to get a confidence interval for your estimate
of the standard deviation of the measurements.
Radford Neal
o-sum game, as the
return to be expected over such a short time period is very small.
Radford Neal
===
This list is open to everyone. Occasionally, less thoughtful
people send inappropriate messages. Please DO NO
in looking at it is.
Radford Neal
===
This list is open to everyone. Occasionally, less thoughtful
people send inappropriate messages. Please DO NOT COMPLAIN TO
THE POSTMASTER about these messages because the postmaster
In article 0srD4.219697$[EMAIL PROTECTED],
Barry Richards [EMAIL PROTECTED] wrote:
I have a friend with a statistics problem who needs some assistance. She
has performed a multiple regression on each of two sets of data and so has
two R squared values. I think one is about 90% and the other is
elease of S-Plus - which
is up to SEVENTEEN times slower than the previous release.
Finally, I doubt very much that the "C" language stands for "computer".
What would it's predecessor language, called "B", have stood for?
Radford Neal
.
Radford Neal
===
This list is open to everyone. Occasionally, people lacking respect
for other members of the list send messages that are inappropriate
or unrelated to the list's discussion topics. Please just
m carrying it. The
result is that the sperm carrying the X chromosome kill off the sperm
carrying Y chromosome, leading to all his children being female. Such
meiotic drive genes have been observed in various animals, though not
yet in humans, to my knowledge.
Ra
a non-blinded study. Subjects are likely to think (even if not told)
that self-defense training should lead to high self-confidence, making
any apparent difference suspect. Administering a test to measure
self-confidence beforehand would make this problem worse.
Radford Neal
45 matches
Mail list logo
