## Step 3: Explore the data - Graphs

Click this link to open the data file for a follow-up study on 40 other students.

Important Note: For convenience, we have defined the number of Facebook friends in terms of "hundreds of friends." So a result of 2.18 means that student indicated 218 friends. The brain density measurement is in "arbitrary units."

With two quantitative variables, the appropriate graph to examine is a scatterplot

 Follow this link to open the Correlation/Regression applet in its own window. You will keep this window open as you do other things. On the datafile page, highlight the two columns (e.g., ctrl-a) to select them and then copy to your clipboard (e.g., ctrl-c). In the applet, press Clear and then click inside the Sample data window and paste the columns of data into the text box (e.g., ctrol-v). Note: The applet assumes you are entering the explanatory variable column first. This is how it was saved in the Excel file, but in the future you may need to press the grey (Explanatory, Response) button to indicate the other ordering or select the Explanatory and Response variables from the pull-down menus before the data widnow. Press Use Data. Confirm that the variable labels have been added to the two axes. If not, make sure you have copied in the column names from the data file.

For each observational unit, a dot is placed at the intersection of the density value and the number of Facebook friends value.  Our convention is to call this the plot of density vs. friends (y vs. x). Take a screen capture and paste a copy of this graph in your report.

(b) In describing scatterplots, our goal is to describe the association between the two variables.  To do so, we focus on three things: direction, strength, and form (but you may want to continue on for now and finish this part later).

• Is the direction of the association as expected by the researchers (clarify what was expected)?
• Would you consider it a strong association? (In other words, does knowing a personâ€™s brain density in this region help you predict the number of Facebook friends?)
• Does the pattern of the association seem reasonably well modeled by a line? (or do you see curvature or some other prominent pattern in the graph?)