BIO375H5 Study Guide - Final Guide: Financial Analyst, Insider Trading, Blind Experiment
Document Summary
Both can be observed: with regression one variable is used to predict the other. One variable affects the other (unidirectional: you would manipulate one variable in a study and look at the effect on the other variable. If it is anticipated, it fails the novelty test. Using vitamin c in oil: composition of matter that has a novel use, things found in nature cannot use patent. Ims is predictive of pml: time consuming to test for ims ims is correlated and causally related to pml but not practical to test ims. It is measurable but not practical to test: what is predictive of ims and tells us risk of pml. Ims is surrogate endpoint: pml is the true endpoint. Is there relationship with ims and the other variables: regression gives predictive relationship, effecting one variable causes other to be effected as well, unidirectional, causal relationships not predicted with correlation. Ims can be argued to be surrogate because it predicts pml.
