Since there is a strong correlation, it is reasonable to use this equation to estimate other values.
For example, using your equation $x=0.351y-12.6$, estimate the hours of study per week for a student who:
has an average test score of 50%. (to 3 significant figures)
has an average test score of 85%. (to 3 significant figures)
has an average test score of 10%. (to 3 significant figures)
hours
hours
hours
Clearly, the last estimate is unrealistic and unreliable because $y=10$ lies far outside the data points.
This is called