So, remember, in order to do regression, we must have a scatter plot that looks somewhat football-shaped. It turns out, in that case, not only can we compute a regression line, but we can actually also do normal approximation and say a little bit more about the y-values. For a given value of X, we can predict by simply by looking at the point that falls onto the line. But it turns out that the y-values of pairs, which are near that x-value, actually follow the normal curve. So, there's a normal curve for the y-values around that point. That means we can use normal approximation for those y-values. Remember, in order to do normal approximation, we need to know two things. We need to know the center and the scale of the normal curve in order to standardize. In the case of regression, these two numbers are given by the predicted value y-hat. And the scale is given by the formula square root 1 - r squared times the standard deviation of y. Let's do an example. Among the students who scored around 41 on the midterm, what percentage scored above 60 on the final? We already computed that the predicted value for a student who scores 41 on the midterm is 62.5. That means that the normal curve is centered at 62.5. So, the percentage of students who scored above 60 on the final follows a normal curve that is centered at 62.5. And we want to figure out what percentage scored above 60. So that would be that shaded area. So, now we simply standardize 60. We take 60, subtract off 62.5, and we divide by that formula over here. So, that's square root 1 - 0.67 squared times the standard deviation of the final exam scores, which is 11.8, and we find -0.29. So then, we get a standard normal curve, and we have to figure out the area to the right of -0.29. And if you look up software, you'll find that this area is 61.4%. So, the answer would be that among all the students who scored around 41 on the midterm, about 61% scored above 60 on the final exam.