Often in science it is important to address the question of whether mean responses differ from one another between groups. For example, average anthropometric measurements may differ between organisms that come from two differing environmental backgrounds. When one is interested in the difference between means of strictly two groups, the most common statistical procedure in this situation is known as the t-test. When there are two or more groups the typical approach is to employ the analysis of variance (ANOVA).
This course will review the concepts behind two-sample and paired t tests and one and two-way ANOVA models. Course attendees will learn the basic motivation and assumptions of each method. Hypothesis tests, confidence intervals, and measures of effect size will also be described for each approach.
To exemplify the t-test and one-way ANOVA, the Egyptian Skulls data set (http://www.dm.unibo.it/~simoncin/EgyptianSkulls.html) will be utilized, which consists of measurements of male Egyptian skulls from five different time periods.
For a better understanding of the two-way ANOVA, we will incorporate analysis of a data set from a lung cancer study at the VA (http://biostat.mc.vanderbilt.edu/wiki/Main/DataSets).
Attendees are invited to bring their laptops with JMP preinstalled if they would like to practice these examples during the course. Laptops/JMP are supplemental and not a requirement for this course.
LISA Short Course: T-tests & ANOVA from LISA on Vimeo.