Measuring Treatment Effects Through Comparisons Along Plot Boundaries
In many experimental situations, conventional methods for removing variation from treatment comparisons are inadequate. This problem is illustrated by data from a pine seed-source planting on a flatwoods site in which significant treatment differences could not be detected using an analysis of variance appropriate for the randomized block design. Treatment differences are apparent when comparisons are made along plot boundaries. Such analysis may prove useful in other situations. Forest Sci. 26:704-709.
No Supplementary Data
No Article Media