The closer the set of data is to a line the closer the correlation coefficient is to 1
in a you can see somewhat of a line so you should pick a value between 0.5 and 0.7
in b you can see a line, the value you should estimate should be greater than 0.9
In c you kinda maybe imagine a line, but worse than a. so you should pick a number between 0.2 and 0.4