Question

The following table gives the thermoconductivity of the washers. You will note that they don’t follow the nice -1, 0, +1 format of experimental designs. You decide what to use, and if it is important in you final analysis of data. Also, because they don’t fit the nice pattern, you will need to calculate the values of your coded variables For example, your real values might be -0.89, 0, 1.2; or you may chose to set the ends at -1 and +1, and you midpoint may not be zero.

Material	Thermal Conductivity (Btu / ft h °F)
Stainless Steel	9.4
Wrought Iron	34.6
Brass	60
Aluminum	119
Copper	218

Answer 1

If you take thermal conductivity as dependent variable in anova set up, if it is to be taken as categorical variable you can create 4 dummies which will take value 1 if it is stainless steel or wrought iron, brass, Aliminium excetra. Else if conductivity of stainless steel is taken as standard for comparison, then conductivity of iron, brass, aluminium copper are approximately 3,6,13,26 times that of stainless steel. Hence a variable coded as 3,6,13,26 can be used as proxy for thermal conductivity.