Analyzing Relationships with Scatterplots

1
Example 2.7
Analyzing Relationships with Scatterplots
2
ACTORS.XLS

• Recall that this data set contains information on
  66 movie stars including their, Gender, Domestic
  Gross, Foreign Gross and Salary.
• We might guess that stars whose movies gross
  large amounts have the largest salaries. Is this
  true?

3
Scatterplots

• We are often interested in the relationship
  between two variables.
• A useful way to picture this relationship is to
  plot a point for each observation, where the
  coordinates of the point represent the values of
  the two variables.
• The resulting graph is a scatterplot.

4
Scatterplots -- continued

• After constructing a scatterplot, we can examine
  the scatter of points.
• We can usually see whether there is any
  relationship between the two variables, and, if
  so, what type of relationship it is.
• We will use a scatterplot to determine if the
  statement about the relationship between movie
  gross and salaries is true.

5
Creating a Scatterplot

• To analyze this we need to plot each stars
  salary on the vertical axis and the corresponding
  domestic gross on the horizontal axis.
• To create this scatterplot
• Place the cursor anywhere in the data set.
• Select the StatPro/Charts/Scatterplot menu item.
• Select Salary as the Y variable, Domestic Gross
  as the X variable and give the chart sheet a name
  such as Scatter.

6
The Resulting Scatterplot
7
Analyzing the Scatterplot

• The message is clear. The points tend to move up
  and to the right. This means that stars in films
  with large domestic grosses tend to make the
  largest salaries.
• The correlation of 0.61 shown in the chart
  supports this conclusion. As we will see in
  Chapter 3, this implies a reasonably strong
  positive linear relationship between the two
  variables.