[EX03-027] Ronald Fisher, an English statistician

(1890–1962), collected measurements for a sample of 150 irises. Of concern were five variables:

species, petal width (PW), petal length (PL), sepal width (SW), and sepal length (SL) (all in mm). Sepals are the outermost leaves that encase the flower before it has opened. The goal of Fisher’s experiment was to produce a simple function that could be used to classify flowers correctly.

A random sample of his complete data set is given in the accompanying table.

Type PW PL SW SL Type PW PL SW SL 0 2 15 35 52 1 24 51 28 58 2 18 48 32 59 1 19 50 25 63 1 19 51 27 58 0 1 15 31 49 0 3 13 35 50 1 23 59 32 68 0 3 15 38 51 2 13 44 23 63 2 12 44 26 55 2 15 42 30 59 1 20 64 38 79 1 25 57 33 67 2 15 49 31 69 1 21 57 33 67 2 15 45 29 60 0 2 15 37 54 2 12 39 27 58 1 18 49 27 63 1 22 56 28 64 1 17 45 25 49 1 13 52 30 67 1 24 56 34 63 0 2 14 29 44 0 2 14 36 50 2 16 51 27 60 2 10 50 22 60 0 5 17 33 51 0 2 12 32 50

a. Construct a scatter diagram of petal length, x, and petal width, y. Use different symbols to represent the three species.*

b. Construct a scatter diagram of sepal length, x, and sepal width, y. Use different symbols to represent the three species.

c. Explain what the scatter diagrams in parts a and b portray.
Let’s see how well a random sample represents the data from which it was selected.

d. Complete parts a and b using all 150 of Fisher’s data on your CD. [DS03-027]

e. Aside from the fact that the scatter diagrams in parts a and b have fewer data, comment on the similarities and differences between the distributions shown for 150 data and for the 30 randomly selected data.
*In addition to using the commands on page 155, use:
For MINITAB: Data display: For each: Select:
Group Group variable: Select: Type For TI-83-84: Enter different groups into separate x, y columns. Use a separate Stat Plot and “Mark” for each group.