Comments on: Create an overall variable / rating from a number of ratings / variables http://www.spsslog.com/2006/09/29/create-an-overall-variable-rating-from-a-number-of-ratings-variables/ From users, for users Wed, 07 Jan 2009 13:06:40 +0000 http://wordpress.org/?v=2.7 hourly 1 By: Yves http://www.spsslog.com/2006/09/29/create-an-overall-variable-rating-from-a-number-of-ratings-variables/comment-page-1/#comment-126 Yves Tue, 21 Nov 2006 19:41:43 +0000 http://www.spsslog.com/2006/09/29/create-an-overall-variable-rating-from-a-number-of-ratings-variables/#comment-126 In order to have a correct analysis, you must be sure your data is structured the same way. An example: Q1: I think public service delivers good work: 1->7 Q2: I think public service delivers bad work: 1->7 These 2 questions actually have the same meaning, but there noses point into another direction... To have a good analysis, you must rescale Q2: from 7->1 Another way to comprehend the underlying feeling people have in regard of public survice, is factor analysis... In order to have a correct analysis, you must be sure your data is structured the same way. An example:

Q1: I think public service delivers good work: 1->7
Q2: I think public service delivers bad work: 1->7

These 2 questions actually have the same meaning, but there noses point into another direction… To have a good analysis, you must rescale Q2: from 7->1

Another way to comprehend the underlying feeling people have in regard of public survice, is factor analysis…

]]>