|
northcentral iowa | LSD is an abreviation for "least significant difference" The FIRST data listed LSD at the 90% confidence interval. If you assume that the statistical analysis was done correctly (which is frequently not the case) the LSD value is a tool to use in assessing the data. Your approach is certainly valid in that yield differences that are less than the LSD are statistically the same at a 90% confidence. That doesn't mean that you shouldn't take notice of differences less than the LSD but just understand that error increases as you evaluate yield that is less than the LSD. Replicated small plots have served us all very well because this is the only way that they major seed companies evaluate hybrids through the development process so don't be too critical of a small plot. If they are executed with the utmost skill and precision the information is better than a larger area because the soil/environmental variability is easier to manage. As mentioned above most of the data and plots that we producers see and evaluate are poorly managed with resulting data value comprimised. Good luck | |
|
I need help interpretting test plot results
