ANOVA demonstration

Download: ANOVA_demo_init.xlsx

Introduction

When you collect a series of measurements you often have two sources of variability:

• uncertainty introduced in the measurement process
• variability due to changes in the value being measured

These two sources of variability can be separated and quantified by Analysis of Variance (ANOVA). In this screencast we will build a simulation of ANOVA-type data that we will analyze later.

Tasks

• Generate the true means for each group; the function should be volatile.
• Use rerrorV to generate measurements on each group
• Plot the data on an XY scatterplot, with the measurement values on the $x$ axis. Get rid of gridlines.
• Add the jittered data underneath the plot and then lose the $y$ axis
• Create and add error bars that stretch from the minimum to the maximum values in each group

When you are finished your spreadsheet should look like the image below.