The 2008 Joint Meeting of the Society for Range Management and the America Forage and Grassland Council. |
Wednesday, January 30, 2008
42
Systems, ANOVA and Risk Analyses
William M. Clapham and James M. Fedders. USDA-ARS, 1224 Airport Rd., Beaver, WV 25813
Analysis of variance (ANOVA) is one of our most powerful tools for analyzing and interpreting experimental data. Risk analysis is another tool that can be used for assessing differences among treatments or processes. Risk analysis generates probabilities of success and failure of meeting objectives. These two methods approach data in different ways. ANOVA partitions sources of variation to increase the resolution to differentiate one or more treatments from others. Risk analysis, on the other hand, begins with the specification of your objective function and then guides the interpretation to determine the probability of meeting that goal. In this study, we used a simulated data set representing yields of four forages. The data were analyzed by both ANOVA and risk analysis. Combined, the two approaches provided a much richer view of the data than when analyzed separately. The data show that the forage with the highest average yield was also the riskiest. Given our objective function of 2.5 t/ha yield, two of the other forages with lower average yield met our goals with higher probabilities of success. Interpreting data from the point of view of risk not only is extremely informative, it is very useful to producers as they attempt to meet enterprise goals.