input levels v/s contributions
From: josh milkwood (lordsimplex_at_yahoo.com)
Date: 01/17/05
- Next message: Jose Blas Navarro Pastor: "All possible subsets regression"
- Previous message: joeu2004_at_hotmail.com: "Re: from odds to probs"
- Next in thread: Mark: "Re: input levels v/s contributions"
- Reply: Mark: "Re: input levels v/s contributions"
- Messages sorted by: [ date ] [ thread ]
Date: 17 Jan 2005 03:46:42 -0800
Hi
I have been using L9 orthogonal array for a 4 factor, 3-level process
(with Design-Expert) and worked out contributions of each factor.
However the input signal for Factor A is in log scale whereas input
signals for B, C & D are doubled over each level (geometric). To make
things clear I reproducing my data here.
Factor Levels Level steps Contributions obtained
A 1, 10, 100 Log increment 40
B 5, 10, 20 Geometric 30
C 1, 24, 48 Geometric 20
D 24, 48 ,72 Geometric 10
My question is: Does the fact that factor A input is incremented on a
log scale distort factor contribution and/or skew my main effects plot.
Hope I am clear
Many Thanks, josh
- Next message: Jose Blas Navarro Pastor: "All possible subsets regression"
- Previous message: joeu2004_at_hotmail.com: "Re: from odds to probs"
- Next in thread: Mark: "Re: input levels v/s contributions"
- Reply: Mark: "Re: input levels v/s contributions"
- Messages sorted by: [ date ] [ thread ]