In a scatter plot, the closer the clustered data points are close to the best line of fit, the greater the correlation that would exist between the two variables.
If we are to draw a best line of fit in the scatter plot that is shown above, the closest data points amongst data points A, B, C, D, and E, that would be close to the best line of fit are data points A and D.
Therefore, removing data point A and point D would cause the correlation to decrease the most.
