We covered a few of the more practical aspects of data collecting for correlation analyses, such as that one should eyeball the scatterplot initially to see if it is even worth applying a full-blown correlation test - if the data is all over the place, it's generally not worth embarking on the math. We also talked about the fact that the data must be paired data, and that it must be collected in a non-biased manner. We then studied the actual formula for the linear correlation coefficient for a sample that is denoted by the letter 'r'. We took the example of the study on foot length vs. height to explore how the formula works, by calculating x^2, y^2, and xy, and then calculating the sums of these, and plugging them into the formula.