The diagram illustrates the effect of outliers on the correlation coefficient, the SD-line, and the regression line determined by data points in a scatter diagram. As the y-value corresponding to the x-value 2 moves from 0 to 7, we can see the correlation coefficient r first increase and then decrease, and the effects on the SD-line (in pink) and the regression line (in green) are evident. (The regression line is discussed in the next unit; it is included here for future reference.) Of course, there are only 5 data points, so a change in one of them has more effect than if there were many.