BIO360H5 Lecture Notes - Lecture 17: Type I And Type Ii Errors, Statistical Hypothesis Testing, Test Statistic
Document Summary
Methods: testing means with parametric and non- parametric methods. Methods: categorical data analysis, power analysis and experimental design. What should we expect to find? (weeks 5 6) How strong is the evidence? (weeks 7 8) Today"s program: correlation inference, hypothesis testing with regression analysis, residual analysis, see also: lab 9, tutorial: decision tree practice (dr. hossam abdel moniem) The di erence between correlation analysis and regression analysis is how we hypothesize these variables, to be in uencing each other. In regression we hypothesize a directed relationship where x a ects y, Regression correlation we don"t say that one is a ecting the other one, it could be both interact back and forth, or it could be that they both kind of depend on a same other variable. Biological hypothesis claims that there is an association between y1 and y2. Note: we hypothesize a causal relationship, even though the study design may not allow us to infer causality!