ENV330H5 Lecture Notes - Lecture 5: Observational Error, Response Bias, Sampling Frame
Document Summary
Allocation of sampling effort (partitioning of variance) Split plot: when the spatial area/logistics/expense of setting up treatments for one factor is larger than for another area. Split the plot by levels of factor a; levels of b are applied within the different plots. E. g. factor a = species of plant (corn & soybean) Experimental unit the smallest unit to which a treatment is applied. E. g. a replicate plot that receives a fertilizer treatment = experimental unit. What are common objectives of a sampling design in. To identify the location of hot spots (areas with high levels of contamination) To characterize the nature and extent of contamination at a site. To monitor trends in environmental conditions or indicators of health. Experimental units receiving treatments are never identical. We expect our treatments to produce responses (signal), but we expect there will be variability (noise) Variability is noise, if we want signal we want to get rid of noise.
