Charlie used a regression calculator to generate the
equation f(x) = -0.15x + 20.1 for the ordered pairs (2,
15), (4, 21), (6, 26), (8, 20), and (10, 14).
Is a linear representation the best way to represent the
data? If it is, explain why. If not, explain why and suggest
a better alternative.

Answer below, let me know if it didn't make sense.  

Step-by-step explanation:

I attached a picture of the scatterplot with a line connecting the points.  There is a definite upside down u type shape, and it doesn't look close to a line at all, so linear is not a good choice.  

an upside down u shape (or right side up u shape for that matter) is exactly what a quadratic is, or any polynomial if you do it right, but quadratic is the simplest, so let's go with that.