Please Help! My book and instructor are no help. Please help me understand why I need to make a scatter diagram, when dealing with bivariate data, and how to find SS(x), SS(y), SS(xy), and r. I have all the formulas, but don't understand. Thanks!
I don't know that the diagram is required to be drawn, if you have all the xi, yi.
See http://www.ms.unimelb.edu.au/~s620202/EMS05ch7.pdf
Using Excel, there is a method of doing the scatter plot, then a fitted line, and from that, determining SS.
So, I really guess I cant offer much help....is there another student in your class you can team up with...that is my recommendation.
Why a scatter diagram?
Well, to repeat a phrase, a picture is worth a 1000 words. With a scatter diagram, one can usually see, at an instant, any correlation between the variables x and y.
SS(x), SS(y), and SS(xy) are, by themselves, not very interesting or informative. However, they are components of calculations that are very helpful. So, dividing SS(x) by n-1 gives the sample variance of the x value; a very helpful statistic.
Combining the three statistics, according to your formula, gives r, which is a correlation statistic. r provides a measure of the linear relationship (correlation) between x and y. With a high degree of correlation, one can use values of x to predict values of y -- Which is what regression analyses is all about.
I hope this helps.
1 answer