Techniques for exploratory analysis
We will be studying the following techniques:
The stem-and-leaf plot
Letter values
Data re-expression
Bagplot—a bivariate boxplot
Resistant line
Smoothing data
Median polish
The stem-and-leaf plot
The stem-and-leaf plot is considered as one of the seven important tools of Statistical Process Control (SPC); refer to Montgomery (2005). It is a bit similar in nature to the histogram plot.
The stem-and-leaf plot is an effective method of displaying data in a (partial) tree form. Here, each datum is split into two parts: the stem part and the leaf part. In general, the last digit of a datum forms the leaf part; the rest form the stem. Now, consider a datum 235. If the split criteria is the units place, the stem and leaf parts here will be respectively 23 and 5; if it is tens, then 2 and 3; and finally if it is hundreds, it will be 0 and 2. The left-hand side of the split datum is called the leading digits and the right-hand side the trailing digits.
In the next step,...
