Sandi Bowman
June 3, 2007, 09:02 AM
Hi, Ankesh,
Sorry I took so long to reply but been up and down lately. Let me give your question a shot at least. Understand I am not expert at this and a lot of my knowledge comes about indirectly.
The way it was explained to me was that, in a simple analysis, your greater numbers equal greater validity argument is the way it normally goes. However, when dealing on primary (individual) predictability, one deals with a great many variables. These variables each have statistical significance of varying degrees (and these are closely held secrets by those who deal with such things since it is essentially their stock in trade). Knowing what weight to assign to each variable and then totaling the result gives the answer. There's probably more to it but this is the simplified answer I was given.
How valid the result depends a lot upon the validity of the originating data upon which the variables are determined, the proper formatting of the questions, and the collection methods incorporated.
Hope this helps answer your question.
Sandi Bowman
Sorry I took so long to reply but been up and down lately. Let me give your question a shot at least. Understand I am not expert at this and a lot of my knowledge comes about indirectly.
The way it was explained to me was that, in a simple analysis, your greater numbers equal greater validity argument is the way it normally goes. However, when dealing on primary (individual) predictability, one deals with a great many variables. These variables each have statistical significance of varying degrees (and these are closely held secrets by those who deal with such things since it is essentially their stock in trade). Knowing what weight to assign to each variable and then totaling the result gives the answer. There's probably more to it but this is the simplified answer I was given.
How valid the result depends a lot upon the validity of the originating data upon which the variables are determined, the proper formatting of the questions, and the collection methods incorporated.
Hope this helps answer your question.
Sandi Bowman
