ABSTRACT
In Chapter 11 we saw that correlation is a measure of the relationship between X and Y. Like correlation, regression is based on a data set of (X, Y) pairs, but it’s different from correlation in that it gives an estimate of Y for a value of X that we choose. So correlation is a number that summarizes, overall, how X and Y relate, whereas regression takes a chosen single value of X and provides an estimate of Y for that X. Recall that Figure 11.1 was a scatterplot of Well-being (the Y variable) and Body Satisfaction (the X variable), for 106 college students. Figure 11.21 showed the separate scatterplots for women and men. Suppose Daniel scores X = 3.0 for Body Satisfaction: What Well-being score would we expect for him, assuming he comes from the same population of college students? We can use regression to estimate Y (Daniel’s Well-being score) for X = 3.0. There are two steps:
1. Calculate from the data the regression line for Y on X. 2. Use that line to calculate an estimate of Y for X = 3.0.
