BIOL 2000H Lecture Notes - Lecture 16: Statistical Unit, Sampling Error, Anthology Series
Document Summary
Sampling error is when you have sub-sampled part of the population and your results are not consistent with the true mean there is a difference between the estimate and the true mean. If you preferentially choose only tall trees in a sample, this would be bias. By not randomly sampling, your ability to test the hypothesis is restricted. Scenario part 3 1 bee is taken from 4 young and from 4 old hives each bee is homogenized and one subsample is analyzed for protein age there are still not enough bees. It the difference between hives that is of interest and not between subsamples. Experimental unit the lowest unit to which a treatment can be applied. Unit of replication would probably be a tube, based on the type of experiment. You are looking at the effects of prozac on tumour growth in mice.