Address typo in AFCARS Federal Register Notice

["Elliott G. Smith" <egs1@xxxxxxxxxxx>]

A recent post to the CMRL gave links to a Federal Register announcement
asking for comments regarding ways to improve the AFCARS (Adoption and
Foster Care Analysis and Reporting System). The e-mail address published
in the Federal Register contains a typographical error. The correct
address is AFCARS_Project@acf.hhs.gov (There is an underline _ between
AFCARS and Project. 

Comments will be accepted until June 27, 2003. 

Thanks,
Elliott 


> Below are the links
> to the Federal Register Announcement asking for comments from the field
> on AFCARS.   We would like to get wide distribution of this so
> please send it to any list serves you are on.
>  
> (PDF file)
> http://a257.g.akamaitech.net/7/257/2422/14mar20010800/edocket.access.gpo.gov/2003/pdf/03-10294.pdf
> (TEXT file)
> http://a257.g.akamaitech.net/7/257/2422/14mar20010800/edocket.access.gpo.gov/2003/03-10294.htm
>  
> If you have trouble following the above links, go to
> http://www.access.gpo.gov/nara/
> browse the FR issues by date (4/28/03) and look for the posting under " Children and Families Administration." 
>

------------------------
Elliott G. Smith, Ph.D.
Associate Director
National Data Archive on Child Abuse and Neglect
Family Life Development Center
Cornell University, Ithaca, NY  14853
607.255.8104   egs1@cornell.edu   www.ndacan.cornell.edu From ???@??? Fri May 09 13:01:41 2003 Status: U Return-Path: Received: from elist02.mail.cornell.edu (elist02.mail.cornell.edu [132.236.56.15]) by postoffice.mail.cornell.edu (8.9.3p2/8.9.3) with ESMTP id NAA12145; Fri, 9 May 2003 13:01:20 -0400 (EDT) Received: (from daemon@localhost) by elist02.mail.cornell.edu (8.9.3p2/8.9.3) id NAA21027; Fri, 9 May 2003 13:01:11 -0400 (EDT) Received: from elist02.mail.cornell.edu (localhost [127.0.0.1]) by elist02.mail.cornell.edu (8.9.3p2/8.9.3) with SMTP id NAA21001; Fri, 9 May 2003 13:00:59 -0400 (EDT) Received: from mailhub2.mail.cornell.edu (mailhub2.mail.cornell.edu [132.236.56.26]) by elist02.mail.cornell.edu (8.9.3p2/8.9.3) with ESMTP id MAA14533 for ; Fri, 9 May 2003 12:42:33 -0400 (EDT) X-PH: V4.1@elist02 (Cornell Modified) From: GTooman@AmericanHumane.org Received: from mailhub2.mail.cornell.edu (localhost [127.0.0.1]) by mailhub2.mail.cornell.edu (8.12.9/8.12.6) with ESMTP id h49GgXPC024506 for ; Fri, 9 May 2003 12:42:33 -0400 (EDT) Received: (from daemon@localhost) by mailhub2.mail.cornell.edu (8.12.9/8.12.6/Submit) id h49GgXsj024503 for CHILD-MALTREATMENT-RESEARCH-L@elist02.mail.cornell.edu; Fri, 9 May 2003 12:42:33 -0400 (EDT) Received: from exch.americanhumane.org (static-209-150-220-107.dsl.customer.eschelon.com [209.150.220.107]) by mailhub2.mail.cornell.edu (8.12.9/8.12.6) with ESMTP id h49GgVPC024470 for ; Fri, 9 May 2003 12:42:31 -0400 (EDT) Received: by EXCH with Internet Mail Service (5.5.2653.19) id ; Fri, 9 May 2003 10:42:23 -0600 Message-ID: X-PH: V4.1@mailhub2 To: Child Maltreatment Researchers Subject: RE: sample size and sig levels Date: Fri, 9 May 2003 10:42:22 -0600 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain; charset="iso-8859-1" Message-Tag: 6123 Reply-To: CHILD-MALTREATMENT-RESEARCH-L@cornell.edu Sender: owner-CHILD-MALTREATMENT-RESEARCH-L@cornell.edu X-Listprocessor-Version: 8.2.09.cu.03/021023/11:56 -- ListProc(tm) by CREN It's good to keep in mind what significance levels are for: to determine whether the same results could be arrived at via chance. The likelihood that a result could be arrived at by chance 5% of the time is considered a "gold standard" of social science research, but may not always be the most important statistic to look at. Effect size with large cohorts is, I think, more important, given that many frequently used statistics are sensitive to sample size. The converse is also true: a pronounced effect, significant at only, say, p = .10 may be considered important if the effect size (variance accounted for by the effect) is high -- sometimes, this is referred to as the "medical effect size," because often in medical studies with small samples the effect is so pronounced that trials are suspended so that the control group can receive the treatment immediately -- such was the case with the initial aspirin trials for prevention of subsequent heart attack, if I remember right. Statistical significance is a good tool in the toolbox. But only one, and has limited value. Gregory Tooman Research Associate, Children's Services American Humane 63 Inverness Drive East Englewood, CO 80112 p: 303-925-9471; fax: 303-792-5333 http://www.americanhumane.org > -----Original Message----- > From: John M Price PhD [mailto:jmprice@calweb.com] > Sent: Friday, May 09, 2003 8:52 AM > To: Child Maltreatment Researchers > Subject: Re: sample size and sig levels > > > > One should look also to the effect size to see if it is > meaningful in any > way. When you have a large sample, you will find the small > parts to any > relationship, but one has to evaluate that in light of the whole > issue/system under study. An r of .05 can be significant, > indicating the > covariance is reliable, but it only accounts for one quarter of one > percent (.0025, .25%) of the variance in the two variables. > > In preventing sudden failure of a jet engine on a commercial > airline, that > might be important. In preventing the same failue in an auto > engine where > gravity isn't such a killer, well, maybe not. > > On Thu, 8 May 2003, bill higgins wrote: > > > I recently read an article that had an extremely large > > sample (population study of health care). The > > differences looked small but were statistically > > significant. should there be adjustments made for > > extremely large samples? > > > > bill > > > > > > __________________________________ > > Do you Yahoo!? > > The New Yahoo! Search - Faster. Easier. Bingo. > > http://search.yahoo.com > > > > -- > John M. Price, PhD > jmprice@calweb.com > Life: Chemistry, but with feeling! | PGP Key on > request or FTP! > Comoderator: sci.psychology.psychotherapy.moderated > Atheist# 683 >