BIO361H5 Chapter Notes - Chapter 4: Microwave Popcorn, Pop Secret, Dependent And Independent Variables
Document Summary
Factorial designs: any statistical designs that are structures to use factors (explanatory variables) to organize groups or treatment conditions. A two-sample t-test is a special case of factorial design that has one factor and two levels (all other factors would need to be held constant) In any study, factor-level combinations are called conditions: they are called treatments in experiments. An experiment imposes a treatment on experimental units while an observational study collects data under varying conditions without imposing changes. Well-designed experiments are conducted under controlled studies to make it easier to isolate the impact of each treatment conditions. In observational studies unknown factors may bias the results. Unlike an observational study, it is easier to show cause and effect in a well-designed experiment because conditions are controlled. Chapter example: microwave popcorn: test: difference in quality between brands, quality = percentage of popped kernels per bag, 3 factors of interest: Factor 1: popcorn brand at 2 levels, fastco and pop.
