Date: Tue, 11 Apr 2000 13:46:22 -0400
Reply-To: Arthur J Kendall <kendalla.hehs@GAO.GOV>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: Arthur J Kendall <kendalla.hehs@GAO.GOV>
Subject: Re: point-biserial
Content-Type: text/plain; charset=US-ASCII
You could just use CORRELATIONS. I one variable is a dichotomy and one
continuous the Pearson correlation is called a point-biserial.
The following applies, iff I am reading between the lines correctly.
You seem to be interested in item-total correlations. If so, take a look at the
RELIABILITY procedure. Be sure that all of your items are scored in the same
direction.
One of the outputs is the "corrected item-total correlation" i.e., the
correlation of each item with the sum of the other items. It also gives you the
squared multiple correlation of each item with the set of other items, and the
alpha for the scale if the particular item were omitted from the scale. It has
other info on how good a scale you have.
____________________Reply Separator____________________
Subject: point-biserial
Author: abdalla alsmadi <alsmadi@MUTAH.EDU.JO>
Date: 04/11/2000 5:24 PM
Hi evrybody
I have 75 binary items(0,1), and a Total score. I need to compute the point
biserial for each item, I couldn't figure it out, is it because I have the
student version of spss???? can you help? Thanks
Abdalla Alsmadi,Ph.D
Dep.of Psychology, College of Educational Sciences
p.o.Box 7
M'utah University - Karak - JORDAN
| 
