Re: Calculator policy

In article <[EMAIL PROTECTED]>, [EMAIL PROTECTED] says... > >I am trying to formulate a calculator policy in a department that currently >allows any calculator except "those capable of storing text". That rules >out all of the graphing calculators since they have alphanumeric >capability. I use

Re: Calculator policy

Brian E. Smith wrote on 3/23/00 1:01 AM: >I am trying to formulate a calculator policy in a department that currently >allows any calculator except "those capable of storing text". That rules >out all of the graphing calculators since they have alphanumeric >capability. I use a TI-83 or TI-86 i

Calculator policy

I am trying to formulate a calculator policy in a department that currently allows any calculator except "those capable of storing text". That rules out all of the graphing calculators since they have alphanumeric capability. I use a TI-83 or TI-86 in my statistics class but under the current po

Reference for "regression discontinuity"

Hi, Carl ---   If you still have your copy of  Introduction to Linear Models (Ward & Jennings) you will find many examples in Chapters 10 and 11.   An interesting example is on paged 217, 11.9 Discontinuity Between Two Second-Degree Polynomials.   With facility to create linear models appr

Re: do we need huge sample sizes?

[EMAIL PROTECTED] wrote: : had forgotten what a p-value is. I find it helpful to explain : significance tests and the outcome of a study : as follows: : : The difference is either : : Statistically significant and big enough to be of practical importance. : Statistically not significant and

Re: Sample size: way tooo big?

Robert McGrath ([EMAIL PROTECTED]) wrote: : : If what you mean is that really large samples can lead to distorted : results of significance tests, I would disagree. The problem is not : that the sample is too big, but that Significance tests are interpreted in : inappropritae ways when readers as

Re: Sample size: way tooo big?

I am really puzzled at this idea that a test can be 'too significant'. All that a test does is to give a measure of the weight of the statistical evidence provided by the data - within the scope of the models used. Something like: 'It is highly likely that the alternative model is preferable to t

cites for regression discontinuity

Carl...even though brief, Pedhazur and Schmelkin (1991). Measurement, design, and analysis. Hillsdale, NJ: Lawrence Erlbaum. do a nice job discussing this subjecton pages 296-301...as well as Cook & Campbell (1979). Quasi-experimentation. Hougton Mifflin (pp.137-146 and 202-205)...h

Re: Sample size: way tooo big?

- I agree with a couple of others, DeLa presents a bad idea. The price being the same, MORE information is BETTER - On Wed, 22 Mar 2000 13:59:19 GMT, "DeLa" <[EMAIL PROTECTED]> wrote: > Well, I suppose that when the sample is too big almost every > relation will prove to be "significant". A lo

New Jersey Short courses

Five New Jersey Area announcements: ===( Announcement #1: Short Course )=== The New Jersey and New York City Metro Chapters Present: An American Statistical Association Short Course, An Introduction to Logistic Regression Stanley Lemeshow, Ph.D.

Re: Sample size: way tooo big?

On Wed, 22 Mar 2000, DeLa wrote: > I have been trying to explain to some co-workers that a sample > can be too big. > That is not very easy because ... ... as everyone knows, more is better. > [... and] because it is contradictory to what > intuition says. Only untutored intuit

Re: do we need huge sample sizes?

On Wed, 22 Mar 2000, dennis roberts wrote in part: > larger sample sizes (random i hope) get us closer to the answer we want > ... what is the parameter or ... what is it NOT > > in this context, you can't have a sample that is too large since, it > will systematically get you closer and close

Teachers this is for you! U4all.com...

u4all.com is currently looking for you to join us and begin sharing your knowledge by being an instructor on our site. u4all.com allows you to lead a live virtual classroom joined by other students. You and your students are able to interact with each other making u4all.com’s concept not only edu

request for help from bangladesh

folks, i am copying to the edstat list this message from a fellow who asks for help for what appears to be a worthy cause. i have edited somewhat for reading ease. he is looking for advice on designing a survey in re "Hence, to monitor the progress of the project a it has been decided that a ba

Re: Sample size: way tooo big?

When we focus on estimates of effect sizes and the stability of those estimates, we are delighted to have a huge sample. Don't focus on statistical significance. === This list is open to everyone. Occasionally, less thou

Re: conficenc inverval

On Mon, 20 Mar 2000 09:28:58 -0500, Taweewan Sidthidet <[EMAIL PROTECTED]> wrote: > Can anyone give me the answer this question: > Will the confidence interval for an unconditional forecast be wider or > narrower if the exogenous variable has greater variation in sample? why? "greater variation

Cambridge University student film reviews

Cambridge University student films review - visit www.worldreviews.com === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about th

No Subject

Will someone give me a (readable) reference for "regression discontinuity"? Thanks in advance. Carl Huberty === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO

Re: Sample size: way tooo big?

DeLa wrote: > Well, I suppose that when the sample is too big almost every > relation will prove to be "significant". A lot of > pseudo-relations will occur. It will become difficult to detect > intermediate(1) variables or neutralise them because there will > be many candidates - if not all the

When can a sample size be too large?

When can a sample be too large? As an experimental psychologist, I explain to my students that in an experiment in which you expect your treatments to modify your sample means, there are always three reasons why one can get a significant result of a statistical test. (i.e.. a result in the a

Re: Your message to hdelagrange@msgto.com

MsgTo.com Spam Protection TestI recently posted a message to Edstat-L and received, shortly afterwards, a message asking me to click on a word to prove that I was human, so that my message might be forwarded to the subscriber [EMAIL PROTECTED] Without thinking, I did so. However, a moment

RE: Sample size: way tooo big?

>I have been trying to explain to some co-workers >that a sample can be too big. That is not very >easy because it is contradictory to what intuition >says. > >Can someone point me to some good arguments or >literature? Or correct me if my assumption is >wrong? Suppose you are taking a random sam

Re: Statistics Software

I recommend MINITAB. I have also had good use from SigmaStat with SigmaPlot for plotting. In article , "news.cwcom.net" <[EMAIL PROTECTED]> wrote: > > > I am looking for a stastical software package I have only used the JMP > > > software (mostly for Design of

RE: do we need huge sample sizes?

One reason why a large sample is not necessarily preferable to a small sample is that sampling error is not the sole source of errors in a survey. For example, it may be better to conduct a small survey and use resources to reduce non-response than to conduct a larger survey and ignore n

Re: Sample size: way tooo big?

Well, I suppose that when the sample is too big almost every relation will prove to be "significant". A lot of pseudo-relations will occur. It will become difficult to detect intermediate(1) variables or neutralise them because there will be many candidates - if not all the variables will be inte

Re: Sample size: way tooo big?

- Original Message - From: DeLa <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Wednesday, March 22, 2000 6:27 AM Subject: Sample size: way tooo big? > I have been trying to explain to some co-workers that a sample > can be too big. > That is not very easy because it is contratictory

Re: Sample size: way tooo big?

If what you mean is that really large samples can lead to distorted results of significance tests, I would disagree. The problem is not that the sample is too big, but that Significance tests are interpreted in inappropritae ways when readers assume statistical significance equals clinical signif

Re: Sample size: way tooo big?

1. If you use a sample larger than necessary, the costs for collecting, processing and analyzing the data are unnecessarily increased. 2. If you take a mechanical approach to statistics, such as looking only at "statistical significance" and not at practical significance, then you may end up bei

do we need huge sample sizes?

the purpose and any inferential statistical procedure is to either answer the question: what is the parameter, or ... to test some specific hypothesis ABOUT a parameter ... thus, the goal of inferential statistics IS finding the parameter. now, significance is nothing more than asking what is

Re: Sample size: way tooo big?

DeLa wrote: > I have been trying to explain to some co-workers that a sample > can be too big. > That is not very easy because it is contratictory to what > intuition says. > > Can someone point me to some good arguments or literature? > Or correct me if my assumption is wrong? I can see that hu

Re: Sample size: way tooo big?

In article , [EMAIL PROTECTED] says... > I have been trying to explain to some co-workers that a sample > can be too big. > That is not very easy because it is contratictory to what > intuition says. > > Can someone point me to some good arguments or literature? > Or

Sample size: way tooo big?

I have been trying to explain to some co-workers that a sample can be too big. That is not very easy because it is contratictory to what intuition says. Can someone point me to some good arguments or literature? Or correct me if my assumption is wrong? --DeLa

Re: Statistics Software

> Shana Mueller wrote: > > I am looking for a stastical software package I have only used the JMP > > software (mostly for Design of Experiments) in the past but am looking > > for software that will not only help in planning experiments, but also > > one in which I can input my data from Excel

Help! Survival curve plot (Kaplan-Meier) with absolute numbers - which

Dear Statisticians, Could somebody please give me an advice on WHICH SOFTWARE CAN CREATE KAPLAN-MEIER SURVIVAL CURVE GRAPHS WITH ABSOLUTE NUMBERS ADDED UNDER THE X-axis? (as it is done in this example http://users.ox.ac.uk/~sg/survival_curve_example.gif where I added numbers at the bottom manual