<table cellspacing="0" cellpadding="0" border="0" ><tr><td valign="top" style="font: inherit;"><DIV>I thought that each subscale is a dimension. Thereby, there are 4 dimensions since the entire instrument has 4 subscales. I dont know if this thinking is right.</DIV>
<DIV><BR><BR>--- On <B>Thu, 1/22/09, Stephanou, Andrew <I><Stephanou@acer.edu.au></I></B> wrote:<BR></DIV>
<BLOCKQUOTE style="PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: rgb(16,16,255) 2px solid">From: Stephanou, Andrew <Stephanou@acer.edu.au><BR>Subject: RE: [Rasch] 4 vs 1 dimension<BR>To: firstname.lastname@example.org<BR>Date: Thursday, 22 January, 2009, 5:39 AM<BR><BR><PRE>"Though I know that the instrument has 4 dimensions"
How do you know? Isn't this only an assumption?
From: email@example.com on behalf of Juanito Talili
Sent: Wed 1/21/2009 5:10 PM
Subject: [Rasch] 4 vs 1 dimension
The 50-item instrument has 4 subscales where each subscale has 2 to 15 items
quantified using 4-point ordinal scale coded 1=strongly disagree to 4=strongly
agree. Though I know that the instrument has 4 dimensions, I've tried to
test the unidimensionality of the overall 50 items. Using Winsteps, the % of
variance explained is somewhat acceptable (>75%) for a unidimensional
assumption. Am I wrong if I do Rasch analysis for the entire instrument rather
than Rasch analysis for each domain?
: Faster, More Secure,Customizable and FREE.
<hr size=1> <a href="http://ph.answers.yahoo.com/question/index?qid=20070903121838AAsHifH">What are the top 10 Pinoy street foods?</a> <br>Yummy! Find out on Yahoo! Answers