02-09-2017 03:18 PM
I have data of two methods 1) Non-destructive and destructive modes of estimating the total and corn ear components moisture content.
Considering that I have: Husk Differential = Total Ear Moisture – (Kernel + Cob Moisture)
How can I deduce that husk may be influencing the total moisture of a corn ear?
I guess, by analyzing (plotting: The differential on the y-axis and husk moisture on the x-axis) the means of the square of the difference between the total ear moisture and Kernel+cob moisture, can provide some information!!!!
My question is why to square the differences? Then how to analyze it? Any suggestion (with sas codes, if possible) would be highly appreciated.
02-09-2017 05:42 PM
The differences are squared because if you take the simple differences from the mean and add them up you'll get zero (+ or - some rounding error)
You may be looking for a regression model where the dependend is variable expressing the right side of your equation and the independent variable the husk measure.
Use a BY variable with a variable that indicates the testing method to see if the results are the same.
You may want to standarize the units of measure (since you didn't mention them) since one expects more moisture by absolute measure (mass, weight or volume) for larger ears than smaller.
02-16-2017 09:57 PM
Thank you very much for your suggestion. I tried doing a regression between the difference in husk moisture (obtained: ((Total corn ear moisture - (kernel+cob moisture)) and kernel moisture. There is no linear relationship between them. I will start working on the problem in two weeks and will update you if I find anything different. Thank you.