NDACAN Logo

National Data Archive on Child Abuse and Neglect -Title Banner
RE: sample size and sig levels
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: sample size and sig levels



Bill,

The usual adjustment when working with very large samples is to choose a
smaller value for alpha.  Something like .01 or .001 rather than the
usual .05 level.

Mark Schmitz

-----Original Message-----
From: owner-CHILD-MALTREATMENT-RESEARCH-L@xxxxxxxxxxx
[mailto:owner-CHILD-MALTREATMENT-RESEARCH-L@xxxxxxxxxxx] On Behalf Of
bill higgins
Sent: Thursday, May 08, 2003 5:55 PM
To: Child Maltreatment Researchers
Subject: sample size and sig levels

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





[ Home | About NDACAN | Datasets | User Support | Contribute Data | Summer Research Institute ]
[ CMRL List Serve | Bibliography | Measures Index | Useful Links | Search ]

Copyright © 1996-2012 National Data Archive on Child Abuse and Neglect