BIOL 3P96 Lecture Notes - Lecture 12: Standard Error, Standard Deviation, Null Hypothesis
Document Summary
Biol 3p96 lecture 12: comparing two means. Tests with one categorical and one numerical variable. Goal is to compare the mean of a numerical variable for different groups. Two-sample: each treatment is composed of an independent, random sample of units. Paired: both treatments are applied to every sampled unit. Two treatments: clear cutting & no clear cutting. Two sample design: random sample of forest plots with clear-cut or no clear-cut treatment randomly assigned to each plot. Paired design: random sample of forest plots, with half of each plot subject to clear-cut and the other half untouched. Paired comparisons control for a lot of extraneous variation between plots or sampling units. Reduces the noise and allows you to more clearly see the effects of the treatment. Example: adjacent plot-halves in the forest are likely to be similar in soil, water, sunlight & other conditions.