Estimation with the Pollination Data

R packages

We’ll use the following R packages in this chapter:

• ggplot2
• Sleuth3
• SMPracticals
• arm

Introduction to estimation

Statistics is all about signal and noise. In the analysis of Darwin’s maize data, we have focused initially on description: we quantified the signal in Darwin’s maize data in the form of measures of central tendency (mean and median values), and the noise using the standard deviations. But what Darwin wanted to know was whether pollination affects fitness— specifically, whether self-pollination is detrimental. In other words, our question concerns the differences in height—something we haven’t analyzed yet. In this lesson, we’ll estimate the differences in height and test if the pollination treatments were affected. Our goal is to estimate the mean heights of the two treatments in Darwin’s experiment and the difference between them, and to calculate various measures that quantify our confidence (and the flip side, uncertainty) in the estimates, which we can use to judge whether they appear to be different or not.