Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site watarts.UUCP
Path: utzoo!watmath!watarts!csgro
From: csgro@watarts.UUCP (Carlo Sgro)
Newsgroups: net.singles
Subject: Virginity Survey IV (Help for V)
Message-ID: <8341@watarts.UUCP>
Date: Tue, 19-Mar-85 00:56:47 EST
Article-I.D.: watarts.8341
Posted: Tue Mar 19 00:56:47 1985
Date-Received: Tue, 19-Mar-85 06:09:43 EST
Reply-To: csgro@watarts.UUCP (Carlo Sgro)
Organization: U of Waterloo, Ontario
Lines: 18


An introduction to correlation and significance levels:
	A correlation is a measure of how closely changes in the value of 
one variable corresponds to a change in another variable.  Correlational
values range from -1.00 to +1.00.  -1.00 indicates that one variable 
decreases when the other increases. +1.00 indicates that one variable 
increases when the other increases.  0.00 indicates no relationship.  
Thus, the relationship between age and number of grandchildren would be a 
positive correlation (0.70?) while the relationship between age and number 
of living grandparents would be a negative correlation (0.50?) and the 
relationship between income level and number of living grandparents would
be near 0.  The correlation coefficient used was the Spearman Rank Correlation
Coefficient (if that means anything to anyone).  The strength of a correlation
is measured by its significance level (p).  Significance level indicates how
likely the results obtained are to be obtained by chance.  A value of p < .05
means that the chance of the results occurring by chance was less than 5%.
A lower significance level indicates a stronger result.  Usually, p < .05 is 
accepted in statistical and other circles as being strong.